Description:FIELD OF THE INVENTION
The present invention relates to a method for localizing epileptogenic zone using HFO analysis in SEEG and a system thereof. More particularly, the present invention discloses a method and a system thereof for mapping physiological HFO rates of the cerebral cortex of the human brain to enable advanced region-specific HFO analysis for improved epileptogenic zone delineation in stereo-EEG in patients with drug-refractory epilepsy.

BACKGROUND OF THE INVENTION
Epilepsy is a chronic neurological disorder marked by recurrent, unprovoked seizures, stemming from abnormal brain electrical activity. These "electrical storms" vary widely in manifestation, from subtle confusion to full convulsions. For patients with drug-resistant epilepsy, surgical intervention is a key treatment. The success of this surgery hinges on precisely identifying and removing the epileptogenic zone (EZ)—the specific brain region where seizures originate. Accurate EZ localization is critical for achieving seizure freedom and minimizing neurological deficits, guiding the entire surgical strategy. While the seizure onset zone (SOZ) is currently derived from visual analysis of SEEG, ongoing research aims to improve EZ delineation for better surgical outcomes.

A significant fraction of individuals with drug-resistant epilepsy (DRE) ultimately requires invasive EEG monitoring when noninvasive modalities fail (in more than 50% of DRE cases) to confidently identify the EZ. In these circumstances the presumptive cortical regions, as inferred from the preliminary presurgical assessment are targeted with intracranial electrodes. Both ictal and interictal stereoelectroencephalography (SEEG) recordings are then utilized to better characterize the EZ, following which epilepsy surgery decisions aimed at reaching seizure freedom are taken.
A number of literature have been published including patents and non-patents documents in said domain.

A non-patent literature by Jayabal, V. et al. titled, “Role of magnetoencephalography and stereo-electroencephalography in the presurgical evaluation in patients with drug-resistant epilepsy”, published in Neurol India in 2017, suggests that for those who are otherwise destined to suffer from uncontrolled seizures and their consequences, SEEG guided ES is an effective option. It explains that SEEG is capable of mapping the entire epileptic network and spatiotemporal relationships across even distant structures. However, determining the precise onset of seizures in SEEG often proves challenging, particularly in instances of neocortical seizures that propagate swiftly through tightly connected microcircuits. Given the challenges of interpreting pre-ictal and ictal phase SEEG, a complementary approach of examining interictal activity may be of value. Consequently, there is a need for complementary methodologies to more accurately characterize the EZ from SEEG recordings.

High-frequency oscillation (HFO) analysis has been extensively studied as a potential biomarker in this regard, supplementing the visual analysis of interictal and ictal discharges. Interictal HFOs, detectable in SEEG at frequencies exceeding 80 Hz, are categorized into two distinct types based on their frequency ranges: ripples (80–250 Hz) and fast ripples (250-500 Hz).
Several studies such as a non-patent literature by Jacobs, J. et al. titled, “High-frequency electroencephalographic oscillations correlate with outcome of epilepsy surgery”, published in Ann Neurol in 2010 report an elevated rate of HFOs within the EZ compared to other brain regions, with a positive correlation observed between removal of cortex with higher HFO rates and successful epilepsy surgery outcomes. However, HFOs, recorded at frequencies above 80Hz in SEEG, are often pronounced in, but not unique to the epileptogenic zone (EZ) when compared to normal cortex. Therefore, HFOs are not exclusive to the EZ and are also known to be generated during complex cognitive processes such as language, motor planning, and memory consolidation in healthy cerebral cortex. Therefore, it is imperative to distinguish pathological from physiological HFOs.

A patent document US20250057484A1, titled “Objective and Training-Free Detection of High Frequency Oscillations in The Epileptic Brain”, discloses a method of identifying high frequency oscillations (HFOs) in neural signal data involves detrending the neural signal data, and identifying HFOs through one or more objective and training-free strike tests selected from the group consisting of amplitude, rhythmicity, and ringing. Numerous other studies have endeavored to distinguish pathological HFOs from physiological HFOs by examining the association of HFOs with spikes, slow waves, and spindles, by clustering HFOs based on features like frequency, duration, and amplitude, or by characterizing task-induced HFOs. Despite the promising nature of these results, the precise identification of pathological HFOs remains an area needing further refinement for its clinical application.

In recent years, the integration of SEEG and advanced computational methods has catalyzed the creation of automated, precise tools for analyzing HFOs, a task previously reliant on manual filter adjustments. A non-patent literature by Zelmann, R. et al., titled “Automatic detector of high frequency oscillations for human recordings with macroelectrodes”, published ion Annu Int Conf IEEE Eng Med Biol Soc in 2010, relates to an automatic detector that detects HFOs by incorporating information of previously detected baselines. However, computational intensity is a significant limitation for certain methods, potentially hindering the speed of analysis. With many different approaches currently in use, establishing a universally agreed-upon definition for a quality HFO sample remains a considerable challenge.

Therefore, there is a need for precise, accurate and reliable method that is computationally efficient for mapping region-specific physiological HFO rates of the cerebral cortex of human brain for improved epileptogenic zone delineation from stereo-EEG in patients with drug-refractory epilepsy.

OBJECT OF THE INVENTION
In order to overcome the shortcomings in the existing state of the art the main object of the present invention is to provide a method for localizing epileptogenic zone using HFO analysis in SEEG in patients with drug-refractory epilepsy.

Another objective of the present invention is to provide a method for deriving unique HFO rate ratio for identifying the EZ using region-specific interictal HFO rates in SEEG.

Another objective of the present invention is to estimate the optimal threshold for the HFO rate ratio for both ripples and fast ripples to localize the EZ.

Another objective of the present invention is to provide a complementary approach of examining interictal activity to accurately characterize the EZ from SEEG recordings.

Yet another objective of this invention is to provide a region-specific normative HFO database using interictal physiological HFO detected from non-resected brain.

Another objective of the present invention is to distinguish pathological from physiological HFOs and identify pathological HFO rates in Interictal HFOs, detectable in SEEG.

Another objective of the present invention is to provide a system for localizing epileptogenic zone using HFO analysis in SEEG using a unique HFO rate ratio and a region-specific interictal HFO rates in patients with drug-refractory epilepsy.

Yet another object is to provide a precise, accurate and reliable method that is computationally efficient for mapping region-specific physiological HFO rates of the cerebral cortex of the human brain for improved epileptogenic zone delineation from stereo-EEG in patients with drug-refractory epilepsy.

SUMMARY OF THE INVENTION
Accordingly, the present invention provides a method for localizing epileptogenic zone using HFO analysis in SEEG and a system thereof. The invention discloses a method and a system for mapping physiological HFO rates of the cerebral cortex of the human brain to enable advanced region-specific HFO analysis for improved epileptogenic zone delineation in stereo-EEG in patients with drug-refractory epilepsy.

The present invention uses the rate of occurrence of HFO and a normative HFO database to identify the epileptogenic zone (EZ) by distinguishing it from physiological HFO rates detected in interictal SEEG recordings. The invention introduces a novel, region-specific HFO rate ratio to localize the EZ.

The method of the present invention comprises of steps of building region specific normative HFO database, implementing the region-specific HFO rate ratio and detector in a graphical user interface (GUI), storing it as a customized EZ delineation tool for automation and easy usability and localization of the EZ in patients using HFO rate ratio. The process of localization of the EZ in patients comprises of acquiring of intracranial SEEG recordings of the patient to be studied, detecting HFOs, mapping the SEEG contacts to their corresponding cortical location as per a brain atlas, calculating the region specific HFO rates for the patient for comparison with region specific normative HFO database, calculating HFO rate ratio and analysing HFO rate ratio by comparing it with values above specific threshold to localize the epileptogenic zone in the patient brain.

The system comprises of but not limited to a data acquisition module for acquiring data from a group of epilepsy patients who have undergone resective surgery or from patient to be studied, one processing module for processing of acquired data from data acquisition module, one analysis and display module etc., for estimating region specific HFO rates database and subsequently facilitating delineation of EZ in patients under study.

As part of the invention the SEEG data from both seizure-free and non-seizure free patients following a brain intervention (resection, ablation or occasional subhemispheric disconnection) were examined to explore the validity of the method of the present invention in predicting the EZ. The method of the present invention accurately localized the EZ in 92% of seizure-free patients and identified pathological HFO rates in 80% non-seizure-free patients, both in 19% within the resected areas and outside in 35% the resected areas. The method of the invention provides an optimal threshold for the ripple HFO rate ratio to accurately localize the EZ as 5.8 times the normative value and 2.7 times for fast ripples, with sensitivity and specificity above 85%. The invention provides clear evidence that differentiating pathological and physiological HFO rates are helpful in localizing the EZ with potential utility in patient-specific surgical planning.

Accordingly, the present invention provides a precise, accurate and reliable method that is computationally efficient for mapping region-specific physiological HFO rates of the cerebral cortex of human brain for improved epileptogenic zone delineation from stereo-EEG in patients with drug-refractory epilepsy.

BRIEF DESCRIPTION OF DRAWINGS
Figure 1 displays the flow diagram of patient selection and creation of region-specific database, A: flow diagram of patient selection, B and C: flow diagram for creation of region-specific database and localization of the EZ.
Figure 2 displays the system for localizing epileptogenic zone using HFO analysis in SEEG, A: building region specific normative HFO database, B: localization of EZ using region specific normative HFO database, C: delineated epileptogenic zone (EZ).
Figure 3 displays the mapping SEEG contacts to corresponding cortical location using the Desikan–Killiany atlas, A: the Desikan–Killiany atlas labelled on 3D anatomical brain model of a patient using FreeSurfer, B and D: images of the patient's brain model coregistered with post-operative MRI and post-implantation CT scans. This coregistration aids in localizing SEEG electrodes within the surgical resection area. SEEG electrodes are indicated by green, while contacts within the resection cavity are marked with red dots, B: provides an axial view, while C and D offer lateral views of the patient's 3D brain model (Patient P27 in table 3), with B and D captured after cortical resection.
Figure 4 displays the frequency domain analysis of HFO, ripple A: and fast ripple B: detected using the MNI HFO detector, C: Periodogram and D: Fourier transformation of interictal SEEG, E: Logarithmic PSD plot shows the strength of the variations (energy) as a function of frequency.
Figure 5 displays the cortical distribution of SEEG electrodes contacts. This depiction shows all cortical points sampled by SEEG electrode contacts that were deemed to be physiological, mapped in MNI space. Each red dot represents an SEEG electrode contact. The views include A: Dorsal, B: Ventral, C: right lateral, D: left lateral, E: Anterior and F: Posterior views of the 3D brain model.
Figure 6 displays the distribution of physiological HFO Rates in the Cortex: This illustration shows the average rates of High-Frequency Oscillations (HFOs) obtained from cortical regions outside the resection zone in patients who are seizure-free. Views in A display the brain model with region- specific Ripple HFO rates, while views in B show the brain model with region- specific Fast Ripple HFO rates. The color maps indicate the HFO rates (blue = 0 and red = 5 HFOs per minute).
Figure 7 displays a graph showing the Receiver Operating Curve (ROC) for determining the best HFO rate ratio to identify the Epileptogenic Zone (EZ). The optimal balance between sensitivity and specificity is depicted in (A, D), with a specificity of 95% displayed in (B, E), and a sensitivity of 95% shown in (C, F) for ripples and fast ripples.
Figure 8 displays the diagram illustrating the classification of patients based on the predicted epileptogenic zone (EZ) by HFO analysis compared to the actual resection area, A: Seizure-free patients and B: Patients non-seizure-free.

DETAILED DESCRIPTION OF THE INVENTION WITH ILLUSTRATIONS AND EXAMPLES
While the invention has been disclosed with reference to certain embodiments, it will be understood by those skilled in the art that various changes may be made, and equivalents may be substituted without departing from the scope of the invention. In addition, many modifications may be made to adapt to a particular situation or material to the teachings of the invention without departing from its scope.

Throughout the specification and claims, the following terms take the meanings explicitly associated herein unless the context clearly dictates otherwise. The meaning of “a”, “an”, and “the” include plural references. Additionally, a reference to the singular includes a reference to the plural unless otherwise stated or inconsistent with the disclosure herein. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this disclosure belongs. The system, methods, and examples provided herein are only illustrative and not intended to be limiting.

The abbreviations used in the invention are represented in table 1 as below:
Table 1: Legend of abbreviations
S.no. Particulars Legend
1 Stereoelectroencephalography SEEG
2 High Frequency Oscillations HFO
3 Electroencephalography EEG
4 Epileptogenic zone EZ
5 Drug-resistant epilepsy DRE
6 Magnetic resonance imaging MRI
7 Deep Brain Stimulation DBS
8 European Data Format EDF
9 Montreal Neurological Institute MNI
10 Fluid-Attenuated Inversion Recovery FLAIR
11 Computed Tomography CT
12 Statistical Parametric Mapping SPM
13 Percentage of Resection PR
14 Desikan–Killiany atlas DK Atlas
15 Receiver operating characteristic curve ROC curve
16 Seizure onset zone SOZ
17 Graphical user interface GUI

Some of the technical terms used in the specification are elaborated as below:
Epileptogenic Zone- The term "Epileptogenic Zone" (EZ) is used to denote the minimum area of cortex that is necessary for the initiation and propagation of seizures, the surgical removal or disconnection of which results in the complete cessation of epileptic episodes.
Drug-refractory epilepsy – The term Drug refractory epilepsy also known as Drug-resistant epilepsy (DRE) is a condition where seizures persist despite adequate trials of two or more different anti-seizure medications (ASMs), either as single drugs or in combination, used at appropriate doses. Essentially, it means the seizures are not controlled by standard medication regimens.
High-frequency oscillations- In brain, high-frequency oscillations (HFOs) refer to rapid, rhythmic brain activity detected through invasive electroencephalography (EEG), typically with frequencies above 80 Hz. These oscillations are of particular interest in neurology, especially epilepsy, as they are thought to be potential biomarkers for identifying seizure activity and the epileptogenic zone.
Physiological HFOs- The terminology Physiological HFOs refer to naturally occurring brain activity within the high-frequency range (typically 80 Hz and above) that are associated with normal brain functions, as opposed to being indicators of pathological conditions like epilepsy. These HFOs, which include ripples (80-250 Hz) and fast ripples (250-500 Hz), are observed in various brain regions and are involved in processes like motor planning, language processing and memory consolidation.
Pathological HFOs- Pathological high-frequency oscillations (pHFOs) are a type of brainwave activity, typically occurring above 80 Hz, that are associated with epilepsy and are often found in the seizure onset zone. They are considered pathological because they are linked to the abnormal electrical activity that causes seizures, unlike physiological HFOs which can occur normally in the brain.
Seizure-free patients- A seizure-free patient is generally defined as someone who has not experienced a seizure for a specified period, often determined by a multiple of their previous seizure frequency. The International League Against Epilepsy (ILAE) suggests a common operational definition where a patient is considered seizure-free if they have been without seizures for at least three times the duration of their longest pre-intervention seizure interval in the preceding 12 months.
Non-seizure-free patients - "Non-seizure free patients" refers to patients who continue to experience seizures despite treatment, whether through anti-seizure medications or interventions such as epilepsy surgery (ES).
Pre-ictal- The term "preictal" refers to the period immediately before a seizure. This phase can vary in length, sometimes lasting minutes.
Ictal phase- The ictal phase of a seizure is the period when clinical symptoms are evident. It's the time between the first observable symptom (including an aura) and the end of seizure activity. This phase correlates with the intense electrical activity happening in the brain during a seizure.
Interictal activity- Interictal activity, in the context of epilepsy, refers to abnormal brain activity that occurs between seizures. It's the period of time that separates one seizure from the next. This activity is often characterized by interictal spikes or other abnormal electrical patterns on an EEG (electroencephalogram).
Intracranial EEG- Intracranial EEG (iEEG) is a diagnostic tool used in epilepsy to locate the source of seizures in the brain by recording electrical activity directly from the brain's surface or within its structures. It is a specialized type of EEG where electrodes are surgically implanted directly onto or within the brain.
Stereo EEG- Stereo-EEG (stereoelectroencephalography) is a minimally invasive surgical procedure used to locate the source of seizures in individuals with drug-resistant epilepsy. It involves inserting thin electrodes deep into the brain to record electrical activity and identify the seizure onset zone (SOZ), which is the area where seizures begin.
Epilepsy surgery- Epilepsy surgery is a neurosurgical procedure performed to reduce or eliminate seizures in individuals whose seizures are not adequately controlled by medication. It involves removing or disconnecting or ablating the area of the brain causing seizures.
Neocortical seizures- Neocortical seizures are a type of epileptic seizure that originate in the neocortex, which is the outer layer of the brain involved in higher-level cognitive functions. These seizures can be challenging to localize because they often spread rapidly and widely and may not have a clearly defined seizure onset zone.
Parcellation: Cortical parcellation in a medical context refers to the process of dividing the cerebral cortex into distinct, non-overlapping regions based on anatomical features. The goal is to create a meaningful and standardized way to study and compare different areas of the brain.
Engel- The Engel classification is a system used to categorize the outcome of epilepsy surgery. It assesses seizure freedom or improvement based on the frequency and disabling nature of seizures following surgery. The classification ranges from Class I to Class IV as explained further. Class I: Seizure-free or only experiencing a few early, non-disabling seizures, or seizures only upon drug withdrawal. Class II: Disabling seizures occur rarely, with disabling seizures potentially being more frequent soon after surgery, or nocturnal seizures only. Class III: Worthwhile improvement with a significant reduction in seizure frequency for a prolonged period. Class IV: No worthwhile improvement or worsening of seizures.

The ideal localization of the EZ is integral to achieving favorable epilepsy surgery outcomes. When non-invasive methods can't localize the epileptogenic zone (EZ) in people with drug-resistant epilepsy (DRE), a significant number of them will eventually need invasive EEG monitoring. In these situations, doctors implant intracranial electrodes into the brain regions suspected of being the source of seizures, based on initial pre-surgical assessments. Both ictal (during seizures) and interictal (between seizures) stereoelectroencephalography (SEEG) recordings are then used to better define the EZ.

Localizing the precise onset of seizures using SEEG presents substantial challenges, notably with neocortical seizures that rapidly propagate across interconnected microcircuits. This highlights the imperative for additional methodologies that augment the accuracy of EZ delineation via SEEG. HFO analysis stands out as a promising biomarker in this realm, offering a complementary perspective to the conventional visual interpretation of interictal and ictal phases. SEEG's ability to capture interictal HFOs at frequencies exceeding 80 Hz, categorized into ripples and fast ripples, enriches our understanding of epileptic network behavior. Interictal pathological HFO rate detection from SEEG aids in delineating the EZ, thereby enabling tailored surgical planning for individual patients. The application of computational tools for HFO detection and rate estimation can enhance the accuracy of EZ mapping, thus contributing to improved surgical results

The present invention discloses a methodology that focusses on the inclusion of region-specific interictal HFO rates as an additional component of SEEG analysis. The invention facilitates the creation of a large-scale region-specific physiological HFO rate database that shall serve as a benchmark for patients undergoing SEEG evaluation, thereby assisting the characterization of the EZ in SEEG-based epilepsy surgery.

STATEMENT OF INVENTION
The present invention discloses a method for localizing epileptogenic zone using HFO analysis in SEEG and a system thereof. The invention discloses a method for mapping physiological HFO rates of the cerebral cortex of human brain to enable advanced region-specific HFO analysis and a system thereof for improved epileptogenic zone delineation in stereo-EEG in patients with drug-refractory epilepsy.
The reference numerals used in the present invention are tabulated below in table 2.
Table 2: Legend of reference numerals
Ser no. Item description Reference numerals
1 System S

2 Patient(s) Pa

3 Data acquisition module D
SEEG submodule D1
CT submodule D2
MRI submodule D3
Medical records submodule D4

4 Processing module P
HFO detector submodule P1
Brain atlas tool P2
Tool for analysis and visualization P3
Processing tool P4
Mapping tool P5
Memory and storage unit P6

5 Analysis and display module A
Computing unit A1
Customized EZ delineation tool A2
Display unit A3

The method for localizing epileptogenic zone using HFO analysis in SEEG comprises of steps of building region specific normative HFO database, implementing the region-specific HFO rate ratio and detector in a graphical user interface (GUI), storing it as a customized EZ delineation tool (A2) for automation and easy usability and localization of the Epileptogenic Zone (EZ) in patients (Pa) using HFO rate ratio.

As per an embodiment of the present invention the flow chart of the method of the present invention is shown in figure 1.

Building region specific normative HFO database comprises of steps as described in the following paragraphs. A group of patients (Pa) who had undergone epilepsy surgery to include seizure-free and non-seizure free patients (Pa) are selected based on the availability of postoperative volumetric MRI scans and interictal High-Frequency Oscillation (HFO) recordings for acquiring of data.

Further, data of the group of patients (Pa) selected is acquired from various sources as described further. The medical history concerning epilepsy resective surgery of the selected patients (Pa) to include baseline characteristics, pre and post - surgical or operative data to include MRI scans, histopathology, surgical outcomes of the selected patients (Pa) etc. are collected retrospectively from electronic medical records. Volumetric T1- weighted and FLAIR MRI, of selected patients (Pa), also referred to as preimplantation MRI scans, are acquired. The MRI scans, specifically the T1 weighted MRI scans are employed to construct three-dimensional representations of individual patient’s (Pa) brain surfaces. SEEG recordings, also referred to as SEEG signal, of the selected patients (Pa) for analysis, are acquired by implanting SEEG electrodes intracranially, and selecting a 30-minute interictal SEEG recording during non-rapid eye movement sleep in each patient (Pa). These recordings are independently reviewed by experienced epileptologist and are exported in EDF format for subsequent analysis. The post-implantation CT scans of individual patients (Pa) are also acquired for the purpose of building the region specific normative HFO database.

Next HFOs are detected and ripples in the range of 80–250 Hz and fast ripples in the range of 250-500 Hz are identified from the selected SEEG recordings of individual patients (Pa).

The SEEG signal of the patient (Pa) is mapped to brain structures using computational methods described as follows. Cortical parcellations are conducted using the three-dimensional representations of an individual patient’s (Pa) brain surfaces by leveraging a brain atlas that provides a comprehensive cortical and subcortical map of individual patient’s (Pa) brains, facilitated by an analysis and visualization tool. Co-registering of post-implantation CT scans, pre-implantation MRI scans, and the cortical parcellations of the individual patients’ (Pa) brains is performed using a processing tool. The post-implantation CT scans are segmented by partitioning the images into multiple segments, each corresponding to a different anatomical structure of the brain. The patient (Pa) SEEG contacts from SEEG recordings are mapped to their corresponding cortical location as per the brain atlas.

Subsequently, SEEG contacts located in the EZ are identified in the following way. The post-implantation CT scans are co-registered with the postoperative MRI using a mapping tool. The SEEG contacts located within the resection cavity which defines the EZ in seizure-free patients (Pa) are identified by careful visual examination of the co-registered images. Subsequently, the HFOs identified from seizure free patients (Pa) (Engel 1) are classified based on whether HFOs identified are in the resection cavity or in non-EZ areas as EZ-HFOs, and non-EZ HFOs respectively. The Percentage of Resection (PR) is estimated by measuring the calculated EZ that was actually resected compared to the cortical resection performed.

Next, the region-specific physiological (non-epileptogenic) HFO rates and database are estimated in the following way. Normal brain area is defined as encompassing any non-lesional, non-epileptogenic, and non-irritative part of the cerebral cortex. The physiological HFO rates in the non-EZ areas are estimated using SEEG signals recorded from the normal brain area. Region-specific normal HFO database are estimated by computing physiological HFO rates from respective anatomical electrode locations across all Engel 1 patients (Pa). Parameters like standard deviation, and percentile are computed for each cortical region in the brain atlas to build a region-specific normal (physiological) HFO rate database.

The defining of normal brain area is based as cortical regions in seizure-free patients (Pa) that fulfill the criteria of being identified as non-lesional cortex on visual inspection of preoperative MRI by an experienced neuroradiologist, the corresponding SEEG contact not falling within the irritative zone with interictal discharges and it not overlapping with the resection cavity on postoperative MRI.

Subsequent to the estimation of region specific physiological (non-epileptogenic) HFO rates and database, the region-specific HFO rate thresholds are defined for the purpose of localization of EZ in the following ways. The EZ for each patient (Pa) is determined as the SEEG contacts that showed a statistically significant deviation from the normative data. A parameter or quantity called the HFO rate ratio is defined and calculated as the quotient of a patient’s (Pa) region-specific HFO rate divided by the normal physiological HFO rates at the 95th percentile value, to accurately localize the EZ. An optimum ratio for each seizure-free patient (Pa) estimated by testing ratios at each level, at increments of 0.1 starting from 1, to detect pathological HFO rates. The optimum ratio is selected by finding the ratio that best matches the resection volume which defines the EZ in the seizure-free patients (Pa) and these optimum ratios are used to localize the EZ in seizure-free and non-seizure free patients (Pa).

Statistical analysis is also performed to estimate the optimum ratios for localizing the EZ that includes plotting a receiver operating characteristic (ROC) curve and testing whether the pathological HFOs rates accurately localize in seizure free patients (Pa), testing median statistical difference between the HFO rates in the EZ and non-EZ channels etc..

The region-specific HFO rate ratio database thus obtained and the detection tool of the present invention was implemented in a graphical user interface (GUI) and stored as a customized EZ delineation tool (A2) for automation and easy usability.

Finally, the localization of the Epileptogenic Zone (EZ) in any subject/patient (Pa) to be studied as a candidate for epilepsy surgery using region specific HFO rate ratio and database comprises of following steps. The intracranial SEEG recordings of the patient (Pa) to be studied are acquired and a 30-minute SEEG recording during non-rapid eye movement sleep is selected for analysis. The HFOs are detected followed by identification of ripples and fast ripples from the selected SEEG recording of the patient (Pa). The patient (Pa) SEEG contacts from SEEG recordings are mapped to their corresponding cortical location as per the brain atlas. The region specific HFO rates for the patient (Pa) are then calculated and compared with region specific normative HFO database. The HFO rate ratio is calculated as the patient’s (Pa) region specific HFO rate divided by normal physiological region specific HFO rate at 95th percentile. The HFO rate ratio is analysed by comparing it with values above specific threshold to localize the epileptogenic zone in the patient (Pa) brain.

The present invention discloses a system (S) for localizing epileptogenic zone using HFO analysis in SEEG that comprises of one data acquisition module (D) for acquiring data from a group of epilepsy patients (Pa) who have undergone epilepsy surgery for preparation of region specific HFO rate database and subsequently for delineation of EZ in a patient (Pa) under study who are candidates for epilepsy surgery, one processing module (P) for processing of acquired data from data acquisition module (D) and one analysis and display module (A) for estimating region specific HFO rates database and subsequently facilitating delineation of EZ in patients (Pa) under study.

As per an embodiment of the present invention the block diagram illustrating the system of the invention is as shown in figures 2 A, B and C.

The data acquisition module (D) in turn comprises of one SEEG submodule (D1) for acquiring SEEG recordings of patient(s) (Pa) comprising of a SEEG system, one CT submodule (D2) for acquiring post implantation of SEEG- CT scans of patient(s) (Pa) comprising of a CT machine system, one MRI submodule (D3) for acquiring preimplantation of SEEG, volumetric T1- weighted and FLAIR MRI scans comprising of an MRI system and one medical records submodule (D4) for acquiring medical history of patients (Pa) comprising of medical records in health care systems.

The processing module (P) for processing of acquired data from data acquisition module (D) in turn comprises of one HFO detector submodule (P1) for detecting HFOs from SEEG recordings comprising of HFO detector system, one brain atlas tool (P2) that provides a comprehensive cortical and subcortical map of patient’s (Pa) brains and thus provides a spatial framework and standardized labels for brain regions, facilitating processing , research, analysis etc., one tool for analysis and visualization (P3) that facilitates usage of brain atlas for cortical parcellations, one processing tool (P4) that facilitates co-registering of post-implantation CT scans, preimplantation MRI scans, and the cortical parcellations for enabling spatial alignment across different imaging modalities, one mapping tool (P5) that facilitates co registering of post-implantation CT scans with post operative MRI scans that facilitates identification of the SEEG contacts in the epileptogenic zone and one memory and storage unit (P6) for storage of acquired or processed data.

The analysis and display module (A) comprises of one computing unit (A1) for housing the various tools and units of the system, one customized EZ delineation tool (A2) that facilitates derivation of region specific HFO rates and database, HFO rate ratios, detection of pathological HFO rates, selection of optimum HFO rate ratios, delineation of EZ etc., and one display unit (A3) with graphical user interface (GUI) for visualization and communication of data and results.

The system (S) of the present invention as described above is characterized with the following features that provide the distinct technical advantage over the prior arts.

The system facilities use of HFO rates to identify the epileptogenic zone (EZ) by distinguishing it from physiological HFO rates detected in interictal SEEG recordings. The system (S) provides a region-specific normative HFO database that is retrospectively constructed using interictal physiological HFO detected from non-resected brain areas of a group of patients (Pa) who remained seizure-free after epilepsy surgery. The system (S) provides and uses region-specific HFO rate ratio and specific thresholds in a method to aid in the detection of region-specific pathological HFOs to localize the EZ. The acquiring of SEEG recordings is done by implanting SEEG electrodes intracranially and selecting a 30-minute SEEG recording during non-rapid eye movement sleep in each patient (Pa). The system (S) facilitates delineation of EZ in a patient (Pa) under study by detecting HFOs from the SEEG recordings and comparing the patient’s (Pa) HFO rate with region specific normative database, deriving the HFO rate ratio and delineation of EZ based on the ratios exceeding specific threshold. Therefore, the system of the present invention as described above enables precise, accurate, reliable and computationally efficient mapping of region-specific physiological HFO rates of the cerebral cortex of the human brain for improved epileptogenic zone delineation from stereo-EEG in patients (Pa) with drug-refractory epilepsy.

The SEEG system of the data acquisition module (D) comprises of SEEG electrodes intracranially implanted using preferably a SEEG robotic or frame-based technique and the SEEG recordings are undertaken using systems selected from Natus EEG Monitoring System, or any clinically validated EEG Monitoring System etc. The CT machine systems are selected from Siemens, GE etc. The MRI systems are selected from Siemens Biograph mMR system, GE Discovery MR 750W or any 3T MRI preferably Siemens Biograph mMR system etc. The brain atlas of the processing module (P) is selected from Desikan–Killiany atlas, VEP atlas, preferably Desikan–Killiany atlas etc. The HFO detector system is selected from Montreal Neurological Institute (MNI) HFO detector HFOApp, pyHFO, Spiking Neural Network (SNN) based system, any HFO detector systems etc The analysis and visualization tool is selected from FreeSurfer version 6.3, 7 etc. The processing tool is selected from any version of Gardel tool etc. The mapping tool is selected from Statistical Parametric Mapping (SPM) software version 8, 12etc.

The system (S) finds application in the application in the domain of neurology, specifically within the field of epilepsy for diagnosis, treatment planning (including surgery), and understanding the mechanisms of epileptic seizures. etc.

EXAMPLES
The present invention shall now be explained with accompanying examples. These examples are non-limiting in nature and are provided only by way of representation. While certain language has been used to describe the disclosure, any limitations arising on account of the same are not intended. As would be seeming to a person skilled in the art, various working alterations may be made to the method in order to implement the inventive concept as taught herein. The figures and the preceding description give examples of embodiments. Those skilled in the art will appreciate that one or more of the described elements may well be combined into a single functional element. Alternatively, certain elements may be split into multiple functional elements. Elements from one embodiment may be added to another embodiment. For example, order of steps of method or processes of data flow described herein may be changed and is not limited to the manner described herein. Moreover, the actions of any flow diagram need not be implemented in the order shown; nor do all of the acts need to be necessarily performed. Also, those acts that are not dependent on other acts may be performed in parallel with the other acts. The scope of embodiments is by no means limited by these specific examples.

For the purpose of development of the present invention the experimental work included all Patients (Pa) who underwent resective surgery following SEEG from January 2017 to December 2022 at the Amrita Advanced Centre for Epilepsy who had post-operative MRI and at least 12 months of clinical follow-up. From a pool of 148 patients who underwent SEEG during this period, 18 patients were not considered for surgical intervention due to inconclusive SEEG findings. Among the remaining 130 SEEG patients, 66 patients with Engel 1, 31 with Engel 2, and 23 with Engel 3 or 4 outcomes were included. Two patients received thalamic Deep Brain Stimulation (DBS) implants, while eight patients were lost to follow-up. For the study, 37 patients were selected from the Engel 1 outcome group, 9 from Engel 2, 13 from Engel 3, and 4 from Engel 4 based on the availability of postoperative volumetric MRI scans as shown in figure 1. Baseline characteristics, presurgical data, histopathology, and surgical outcomes were gathered retrospectively from electronic medical records. The retrospective study received approval from the institutional ethics committee.

SEEG electrodes (PMT Corporation, USA) were intracranially implanted using a SEEG robotic or frame-based technique. The SEEG recordings were undertaken using the Natus EEG Monitoring System, with signals recorded at a sampling frequency of 2048 Hz. A 30-minute SEEG recording during non-rapid eye movement sleep was selected for analysis in each Patient (Pa). The recordings were exported in EDF format for subsequent analysis.

The Montreal Neurological Institute (MNI) HFO detector was utilized to identify ripples and fast ripples in the development of the present invention. Only HFOs with strength 7 or above were considered for the analysis as this threshold better distinguished HFOs from artifacts. The number of HFO events per minute was termed as the HFO rate.

Prior to the procedure of electrode implantation, each Patient (Pa) underwent both volumetric T1- and FLAIR Magnetic Resonance Imaging (MRI, Siemens Biograph mMR system or GE Discovery MR 750W). Specifically, the T1-weighted MRI scans were employed to construct three-dimensional representations of individual patient’s brain surfaces. These detailed renderings served as the basis for subsequent cortical parcellations. The Desikan–Killiany atlas, a widely used brain atlas that provides a comprehensive cortical and subcortical map was utilized to conduct these parcellations. The parcellation process was facilitated by FreeSurfer software suite (http://surfer.mgh.harvard.edu/ ) known for its capacity to analyze and visualize neuroimaging data.

Post-implantation Computed Tomography (CT) scans, pre-implantation MRI scans and the cortical parcellations were then co-registered using the Gardel tool. This process of co-registration ensures precise spatial alignment across different imaging modalities, thereby enabling a more accurate comparison and analysis of the imaging data. Following the co-registration, the post-implantation CT images were segmented, a process which involves partitioning the image into multiple segments, each corresponding to a different anatomical structure. The SEEG contacts were then mapped to their corresponding cortical location as per the Desikan–Killiany atlas. This mapping process further refines the spatial accuracy of the SEEG contact points within the context of cortical anatomy, thereby contributing to a more precise interpretation of the SEEG data.

All Patients (Pa) included in the analysis underwent volumetric MRI within 4-6 months post-epilepsy surgery. To identify the SEEG contacts that sampled the epileptogenic zone, the post-implantation CT was co-registered with the postoperative MRI using Statistical Parametric Mapping (SPM) software version 12. Through the careful visual examination of the co-registered images, it was possible to identify the SEEG contacts located within the resection cavity which defines the EZ in seizure-free patients as shown in figure 3. Following a brain intervention (resection, ablation or occasional sub hemispheric disconnection), tissue displacement can occur and be detected in postoperative MRI scans. To account for this shift an additional 5 mm margin was added around the resection cavity. This modification assists in accurately locating the SEEG contacts within the cavity, acknowledging the chronic brain movement in response to the intervention.

The Percentage of Resection (PR) was also estimated by measuring how much of the calculated EZ was actually resected compared to the cortical resection performed. A PR of 100% indicates a perfect match between the calculated EZ and the actual cortical resection. A PR below 100% means the resection was insufficient to cover the calculated EZ, while a PR above 100% indicates that the resection included additional brain areas beyond the calculated EZ.

According to an embodiment of the present invention, the normal brain was defined based on cortical regions in seizure-free patients fulfilling these criteria of firstly non-lesional cortex on visual inspection of preoperative MRI by an experienced neuroradiologist, secondly the corresponding SEEG contact did not fall within the irritative zone with interictal discharges and lastly ensuring it did not overlap with the resection cavity on postoperative MRI. Therefore, the normal brain area encompasses any non-lesional, non-epileptogenic, and non-irritative part of the cerebral cortex. The physiological HFO rates in the non-EZ areas were estimated using SEEG signals recorded from this normal brain region. The region-specific normal HFO database was estimated by computing HFO rates from respective anatomical electrode locations across all Engel 1 patients (Pa) included in the study. The interhemispheric HFO rates were averaged since they were derived from a relatively small dataset. Mean, standard deviation, and percentile were computed for each cortical region in the DK atlas to build a region-specific normal (physiological) HFO rate database. The region-specific HFO rate ratio and detector were implemented in a graphical user interface (GUI) for automation and easy usability. The customized EZ delineation tool is available at https://github.com/Brain-Mapping/HFORateRatio .

After creating the region-specific physiological HFO database, the EZ was determined for each Patient (Pa) as the contacts that showed a statistically significant deviation from the normative data. To accurately localize the EZ, a parameter or quantity called the HFO rate ratio was defined. The HFO rate ratio was calculated as the quotient of a patient’s region-specific HFO rate divided by the normal physiological HFO rates at the 95th percentile value.

〖HFO rate ratio〗_( ripples)= (〖Patient’s region specific HFO rate〗_( ripples) )/〖Normal physiological HFO rate at 95^th percentile〗_( ripples)
……………. 1

〖HFO rate ratio 〗_(fast ripples)= (〖Patient’s region specific HFO rate〗_( fast ripples) )/〖Normal physiological HFO rate at 95^th percentile〗_( fast ripples)
…………….2

To detect pathological HFO rates, an optimum ratio was estimated for each seizure-free Patient (Pa) by testing ratios at each level, at increments of 0.1 starting from 1. The optimum ratio was then selected by finding the ratio that best matched the resection volume which defines the EZ in seizure free Patients (Pa). Furthermore, the optimum ratios were used to localize the EZ in seizure-free and non-seizure free patients (Pa) included in the study.

To estimate the optimum ratio for localizing the EZ, a receiver operating characteristic (ROC) curve was plotted. This was done by estimating the sensitivity and specificity at different ratios by testing whether the pathological HFOs rates accurately localize in seizure free patients. The Mann-Whitney U-test was also used to test the median statistical difference between the HFO rates in the EZ and non-EZ channels.

The research work for the present invention included a total of 7512 bipolar SEEG recordings from 63 patients (31 males and 32 females, mean age 25 years (range 8-53)) with a mean surgical outcome follow-up of 27 ± 14.7 months (range 12.1 – 59.5). Of 63 patients, 54 underwent cortical resection, 7 received stereotactic radiofrequency thermal ablation, and 2 had sub-lobar disconnection as the surgical interventions. The region-specific HFO database was created from 3762 bipolar SEEG recordings of 37 seizure-free patients. The patient demographics, pre-surgical data, histopathology, and surgical outcomes are detailed table 3.

Using the MNI HFO detector, HFOs were effectively detected in both ripple and fast ripple frequency bands as shown in figure 4. From 1,890 minutes of SEEG recordings, 317,874 ripples and 166,038 fast ripples were detected in both the EZ and non-EZ brain regions. Region-specific HFO rate analysis successfully localized the EZ in 34 (92%) out of the 37 seizure-free patients. Additionally, data from 26 non-seizure free patients were analyzed.

According to an embodiment of the invention SEEG data gathered from normal brain regions was utilized to derive the region-specific physiological HFO rate data base by the method as described in the preceding paragraphs. The implantation schema of these patients (Pa) sampled almost all gyri of the brain and it was possible to map it to Desikan–Killiany atlas. Distribution of SEEG electrodes on brain structures were given in figure 5. The Montreal Neurological Institute's HFO detector efficiently identified HFOs in SEEG recordings of all patients (Pa) included in the work of the present invention. An analysis focusing on the frequency domain of these HFOs is provided in figure 4.

The present invention identified an increased occurrence of ripple oscillations across various areas of the cerebral cortex, including both eloquent and non-eloquent zones. Specifically, these oscillations were noted in the grey matter of the lingual gyrus, paracentral gyrus, and superior, parietal lobule, as well as in the white matter of the cuneus, medial orbitofrontal, gyrus , middle temporal , gyrus, parahippocampal gyrus, and precentral gyrus as seen from table 4 and figure 6. However, overall fast ripple oscillations were observed less frequently in these cortical regions. When detected, fast ripples were primarily found in the grey matter of Precentral gyrus, Lingual gyrus, Isthmus of cingulate and in the white matter of the Parahippocampal gyrus, Medial orbitofrontal gyrus, and Precentral gyrus as seen from table 4 and figure 6.

Additionally, in the normal cerebral cortex, it was also found that fast ripples occurred significantly less often (average rate of 0.34, standard deviation 0.92, and median rate 0.1) compared to ripple oscillations, which showed an average rate of 0.97, standard deviation 2.25, and a median rate of 0.18.

The HFOs identified from seizure free patients (Engel 1) were subsequently classified into two categories: 1) HFOs identified in the resection cavity as EZ-HFO, and 2) HFOs identified in non-EZ areas as non-EZ HFOs. It was evident that the rates of ripples and fast ripples were significantly higher in EZ contacts when compared to non-EZ. Statistical analysis revealed a significant difference (p <0.001) in HFO rates from EZ areas when compared to normal non-epileptogenic cortex. In EZ channels, the ripples were detected at a mean rate of 2.92 (± 6.29) per minute and a median rate of 0.73 per minute, the fast ripples were observed at a mean rate of 2.09 (± 6.05) per minute and a median rate of 0.23 per minute. In contrast, non-EZ channels exhibited ripples at a mean rate of 1.12 (± 3.34) per minute and a median rate of 0.20 per minute; fast ripples had a mean rate of 0.33 (± 0.76) per minute and a median rate of 0.1 per minute. All measurements are expressed in rates.

Upon examination of physiological HFO rates (non-EZ HFOs) across patients (Pa), it was observed to be a skewed distribution as contrasting to a normal distribution, prompting the use of percentiles for subsequent ratio estimation. Further we used percentile for estimating optimum HFO rate ratio for ripples and fast ripples in order to effectively localize the EZ. Using ROC analysis, the optimum ratio for ripple HFO rates in EZ localization was identified as 5.8 times the value of the 95th percentile, yielding a sensitivity of 83.5% and specificity of 86.4% . For fast ripples, the optimum ratio was 2.7 times the value of the 95th percentile, resulting in a sensitivity of 89.5% and specificity of 85.3 % as seen from figure 5. A 95% specificity can be achieved by setting the optimum ratio to 7.8 for ripples and 4.3 for fast ripples, respectively. Similarly, to achieve a 95% sensitivity, the optimum ratio should be set to 2.9 for ripples and 1.9 for fast ripples as referred to figure 7.

Further, the HFO rate ratio tool, also referred to as customized EZ delineation tool, was tested on seizure free and non-seizure free patients. The EZ in these patients (Pa) was determined by comparison of the patient's individual HFO rates at a given cortical region against the database of region-specific normal HFO rates. Out of the 37 seizure-free patients included in the study, successful EZ localization was achieved in 34 patients, representing 92% of cases as seen in table 5.

The Percentage of Resection (PR) was estimated by measuring how much of the calculated EZ was actually resected compared to the cortical resection performed. In seizure-free patients, the mean PR was 135 % for ripples and 144% for fast ripples as seen in table 6.

The HFO rate ratio was further analyzed in a group of 26 patients who did not achieve seizure freedom, including nine Engel-2, thirteen Engel-3, and four Engel-4 patients.

Among the 9 Engel-2 patients (P38-P46): in 4 patients, the predicted epileptogenic zone (EZ) matched the resection cavity; in 1 patient, the predicted EZ extended to both the resection cavity and other brain areas; in 2 patients, the predicted EZ did not match the resection cavity; and in 2 patients, a broadly distributed EZ was suggested as almost all SEEG contacts were classified as EZ as shown in figure 8B and table 6).

Among the 13 Engel-3 patients (P47-P59): in 1 patient, the predicted EZ matched the resection cavity; in 4 patients, the predicted EZ was found in both the resection cavity and other brain areas; in 5 patients, the predicted EZ did not align with the resection cavity; in 2 patients, no pathological HFOs were detected; and in 1 patient, a diffuse EZ was indicated as almost all SEEG contacts were identified as EZ as shown in figure 8B and table 6.

Among the 4 Engel-4 patients (P60-P63): in none of them did the predicted EZ match the resection cavity; in 3 patients, the predicted EZ was localized to both the resection cavity and adjacent non-resected cortex; in 1 patient, the predicted EZ did not align with the resection cavity; and none had cases where no pathological HFOs were detected or where all SEEG contacts were classified as EZ as shown in figure 8B and table 6. We also estimated the PR for non-seizure-free patients and found that the mean PR was 42% for ripples and 48% for fast ripples, which is lower compared to seizure-free patients as seen in table 6.

The present invention endeavors to use the rate of occurrence of HFO to identify the epileptogenic zone (EZ) by distinguishing it from physiological HFO rates detected in interictal SEEG recordings. The approach centers on a novel, region-specific HFO rate ratio to localize the EZ and its inplications are discussed in the following paragraphs.
The task of selecting SEEG contacts to categorize as non-EZ HFOs is particularly intricate, given that it involves implanting and recording SEEG electrodes in patients with epilepsy. These patients (Pa) typically have electrodes implanted in cortical regions encompassing primary, secondary, and tertiary areas to test hypotheses regarding the EZ. Additionally, electrodes are placed adjacent to EZ to confirm resection boundaries. Once the EZ of a Patient (Pa) is corroborated by both an epileptologist and a surgeon, a tailored resection of the cortex is performed. Identifying the EZ electrode in these patients (Pa) is possible by co-registering the post-implantation images with the post-resection MRI images. Any remaining electrodes, which are in the non-lesional, non-epileptogenic cortex in seizure free patients can then be classified as physiological.

Prior studies have demonstrated that physiological HFOs are generated in the cortex during the preparation and execution of various cortical functions, such as language processing, motor control, and memory. The work on the present invention has presented that different cortical regions and tissue types appear to generate varying baseline rates of HFOs as suggested by prior research. Therefore, the findings of the present invention align well with this previous observation. Specifically, it is observed that ripples and fast ripples in non-EZ regions were predominantly found in the occipital, mesiotemporal, and frontal lobes, which resonates with the observations made by prior research in the field. Consistent with prior research, the findings of the present invention also revealed significantly higher mean rates and amplitudes of HFO in the seizure onset zone (SOZ) compared to non-SOZ areas, with statistical significance of p < 0.001. Another nuance observed was that the incidence of fast ripples was comparatively less frequent than that of ripples and therefore their high rate of occurrence adds to the specificity to the seizure onset zone.
Integrating SEEG with advanced computational methods has recently automated and refined HFO analysis, a process that used to depend on manual filter adjustments. The Montreal Neurological Institute's HFO detector is known to have high sensitivity and accuracy in HFO detection. The routine integration of HFO analysis in SEEG practice shall help differentiate between physiological and pathological HFOs for localizing the epileptogenic zone.

Given the diverse cognitive functions regulated by the cerebral cortex, it is natural for HFO rates to vary substantially across different cortical regions. This complexity has traditionally presented a significant challenge to relying solely on global HFO rates to differentiate the epileptogenic zone (EZ) from healthy brain regions. Despite the diverse methodologies attempted in the prior arts, the existing body of research has yet to provide a clinically useful method for delineating pathological HFOs and use in SEEG-guided surgical management of epilepsy. The present invention therefore introduces a new perspective that utilizes region-specific HFO rate ratio to discriminate EZ HFOs from physiological HFOs.
The present invention focus is on constructing a region-specific physiological HFO database that encompasses the cerebral cortex, as outlined in Table 4, and validating the HFO rate ratio threshold.

The ideal region-specific HFO rate ratio for EZ detection was calculated to be 5.8 times and 2.7 times the 95th percentile values for ripples and fast ripples, respectively as shown in figure 7. Applying this optimal HFO rate ratio to 37 seizure-free patients yielded promising results: the EZ was accurately localized in 34 patients (92%), but not in 3 patients (8%). This failure in EZ localization for 8% of seizure-free patients suggests that, despite adequate implantation (i.e., the EZ was resected and the patients were seizure-free), the pathological high-frequency oscillations (HFOs) were either insufficiently generated or diffuse. Another explanation is that, under certain conditions, the rates and amplitudes of HFOs in both pathological and healthy tissues may appear identical, rendering differentiation impossible.

It was also observed that HFO rate analysis was unable to localize the hypothesized EZ in 5 non-seizure-free patients (out of 26), and only partially localized it in 21 non-seizure-free patients (out of 26). This suggests that either the resection did not completely remove the EZ, or the implantation did not adequately cover the EZ. Clearly, this aspect of the study warrants further investigation, ideally with a larger cohort of non-seizure-free patients and a broader range of seizure etiologies.
In the research work for the present invention, the PR was measured as quantity of the calculated EZ that was actually resected compared to the cortical resection performed. A PR of 100% indicates a perfect match between the calculated EZ and the actual cortical resection. A PR below 100% means the resection was insufficient to cover the calculated EZ, while a PR above 100% indicates that the resection included additional brain areas beyond the calculated EZ. In seizure-free patients, the mean PR was 135% for ripples and 144% for fast ripples. Conversely, in non-seizure-free patients, the mean PR was 42% for ripples and 48 % for fast ripples. Therefore, a relatively higher incidence of incomplete resection of cortex with pathological HFO rates was associated with failure to achieve seizure freedom.

It is well documented that patients (Pa) with drug-resistant epilepsy (DRE) and normal high resolution MRI results often undergo SEEG evaluation to accurately map the epileptogenic zone (EZ). However, even with SEEG evaluation in MRI-negative epilepsy, post-surgical seizure freedom is only attained less than 60% of cases. This suboptimal outcome is due to the possible causes to include insufficient EZ sampling, incomplete EZ resection, or the existence of a secondary epileptogenic zone. Therefore, there is a requirement of tests that can confirm the adequacy of EZ sampling during SEEG evaluation. In this context, if the HFO rate analysis of a given SEEG indicates only physiological rates, then there is a possibility that the EZ was missed in the SEEG sampling. Thus, HFO rate analysis can additionally serve as a valuable tool to confirm whether the SEEG implantation has successfully sampled the epileptogenic zone.

Table 3. The patient demographics, pre-surgical data, histopathology, and surgical outcomes. HS = Hippocampal Sclerosis, FCD = Focal cortical dysplasia, MTS = Mesial temporal sclerosis, ATLAH = anterior temporal lobectomy with amygdalohippocampectomy.
Patient no. Age MRI findings Epilepsy Surgery Type of epilepsy Pathology Seizure freedom in Engel scale Follow up in months
P1 37 Normal, subtle blurring at the right orbitofrontal region. HV: normal ATLAH + Neocortical resection Fronto temporal Hippocampal Sclerosis Blumke's Class I, No Evidence of any dysplasia seen in the Orbito Frontal Specimens. 1 14.4
P2 26 Normal Neocortical resection Parietal No evidence 1 48.9
P3 28 Subtle FCD in right pre frontal Neocortical resection Frontal Focal Cortical Dysplasia Taylor Type IIB 1 12.3
P4 15 Left HS ATLAH Temporal Blumcke type I B 1 16.3
P5 26 Normal ATLAH Temporal HS 1 32.5
P6 21 Normal Neocortical resection Frontal FCD 2a 1 45.5
P7 24 Normal Neocortical resection Occipital Focal Cortical Dysplasia Taylor Type IIB. 1 21.8
P8 23 Normal Neocortical resection Frontal FCD type 2a 1 25.1
P9 12 Normal, subtle volume loss in distal body and tail of left Hippocampus. Neocortical resection Frontal FCD 2b 1 47.8
P10 34 Mild left HS ATLAH Temporal Hippocampal Sclerosis and Amygdala: Focal Cortical Dysplasia Type IIA 1 32.7
P11 27 Normal Neocortical resection Frontal Inconclusive 1 23.0
P12 37 Normal Neocortical resection Parietal MTS Type 1a 1 47.9
P13 22 Subependymal heterotopias - both sides, loss of internal architecture and mild hyperintensity of the left hippocampus, small hyperintensity of GM - in the left parietal - superior parietal ATLAH + Neocortical resection Temporo occipital Hippocampus - Mesial Temporal Sclerosis. Heterotropia Noted 1 14.6
P14 31 Left anterior temporal /middle cranial arachnoid cyst. Neocortical resection Parietal - 1 15.9
P15 33 Normal Radiofrequency ablation occipital - 1 23.0
P16 37 Right MTS ATLAH Temporal HS 1 40.4
P17 13 Left hippocampus hyperintensity, T2 Relaxometry 123 on left and 113 on right, normal hippocampal volumetry, malrotated bilateral hippocampus, Doubtful thickening of left posterior insula, normal amygdala ATLAH Temporal Hippocampal Sclerosis _MTS Blumke's Type 1 C 1 12.4
P18 41 Mild cortical atrophy, slightly brighter hippocampus on the left side; otherwise no overt HS. ATLAH Fronto temporal MTS 1 19.1
P19 25 Right HS ATLAH Temporal Hippocampal Sclerosis-Blumke's MTS Type IB 1 21.5
P20 28 Normal Neocortical resection Frontal Reactive Gliosis. 1 14.2
P21 17 Right amygdala and temporal pole thickening of cortex and blurring of grey white junction ATLAH Temporal Amygdyla- Dysplasia 1 52.2
P22 32 Subtle alteration in the left hippocampus internal architecture, T2 relaxometry - 98 right / 93 left; MB/Fornix are symmetric, FLAIR hippocampal hyperintensity can be seen visually, no significant volume loss, no encephalocoele , HV : normal ATLAH Temporal Hippocampal Sclerosis-ILAE Type 3 1 12.3
P23 19 Normal Neocortical resection Frontal FCD 2a 1 54.3
P24 28 Normal Neocortical resection Frontal FCD type 1 1 24.4
P25 30 Overall volume loss over the right posterior hemisphere. Ulegyria int he right lateral parietal lobe, temporal and right perisylvian and temporal neocortical volume loss. Mesiotemporal volume loss with fornix atrophy. ATLAH Temporal Hippocampus : reactive gliosis, no neuronal loss, Amygdala : Suspected dysmorphic neurons 1 16.3
P26 8 Bilateral gliotic foci in the subcentral gyri left > right with contracted adjacent overcool-insulae. Remaining brain ~ normal. Neocortical resection Fronto-Parietal Biopsy Fronto-Parietal Opercular Lesion: Consistent with Reactive Gliosis. 1 39.9
P27 15 Normal Neocortical resection Frontal Normal tissue 1 40.6
P28 29 Normal Neocortical resection Frontal Inconclusive 1 27.4
P29 12 Normal Neocortical resection Frontal Reactive gliosis 1 51.6
P30 16 Normal Neocortical resection Frontal Focal Cortical Dysplasia Taylor Type 2A 1 12.9
P31 18 Gliosis involving the left superior parietal lobule and posterior temporal and basifrontal lobe. Neocortical resection Frontal Reactive gliosis 1 39.1
P32 40 Normal Neocortical resection Frontal Reactive gliosis 1 52.8
P33 29 Thickening of the grey matter of the entorhinal cortex / temporal pole
secondary TLE changes are present: fornix/MB atrophy ATLAH Temporal MTS - Blumcke classification - type III 1 17.3
P34 20 Normal Radiofrequency ablation Frontal - 1 12.6
P35 13 Mild right HS Neocortical resection Frontal Reactive gliosis 1 44.7
P36 23 Normal ATLAH Temporal Hippocampal Sclerosis Blumke's Classification MTS Type 2, Shows Dyslamination (FCD Type 1) 1 39.6
P37 21 FLAIR hyperintensities with white matter volume loss in bilateral parietal regions with prominent occipital horns of lateral ventricle - Features could represent chronic PVL changes. Neocortical resection Parietal - 1 17.8
P38 30 Left cerebral hemiatrophy with gliosis involving white matter Neocortical resection Frontal Reactive Gliosis 2 59.5
P39 27 Gliosis involving the right superior and middle temporal gyrus, insular cortex, inferior frontal gyrus, frontal and temporal operculum. Neocortical resection Frontal, temporal - 2 13.9
P40 41 Normal ATLAH Temporal Focal Cortical Dysplasia Taylor Type II 2 36.3
P41 15 Left area 6 premotor cortex FCD Neocortical resection Fronto-Parietal Focal Cortical Dysplasia-Type IIA 2 12.1
P42 24 Normal ATLAH Temporal Hippocampal Sclerosis Blumcke Classification Type 1 2 12.4
P43 28 Normal Neocortical resection Frontal Reactive gliosis. 2 51.9
P44 17 Cystic encephalomalacia/gliotic changes noted involving bilateral basifrontal and bilateral anteroinferior temporal lobe. Frontopolar and anterior cingulate disconnection with anterior callosotomy Frontal Reactive gliosis 2 19.3
P45 36 Normal Radio frequency ablation Frontal - 2 13.8
P46 53 Normal Radio frequency ablation Frontal Cortical Dyslamination cannot be commented . -However there are no Dysmorphic neurons seen.No Balloon Cells seen. -No evidence of any Malignancy seen. 2 15.0
P47 31 Normal Neocortical resection Parieto-occipital Focal Cortical Dysplasia Taylor Type II A 3 12.9
P48 37 Normal Neocortical resection Parieto-occipital Focal Cortical Dysplasia -Type 1 3 12.1
P49 20 Normal Neocortical resection Temporal Features of Focal Cortical Dysplasia Taylor-Type IIA. 3 18.3
P50 11 Bilateral subependimal nodules temporo occipital disconnection Parietal Reactive gliosis 3 19.9
P51 34 Left post-central sulucs on the left - possible sulcal widening & G-W junction blurring. Possible Type 1, HV : normal MRI: normal Radio frequency ablation occipital Left occipital FCD -Focal Cortical Dysplasia -Type 1 is a Possibility 3 14.0
P52 13 Normal Neocortical resection Parietal - 3 34.1
P53 20 Reportedly normal, Very unusual bisecting sulcus over the omega sign of the left motor cortex. Neocortical resection Fronto-Parietal Suspicious of dysmorphic neurons 3 21.1
P54 21 Bilateral symmetric parietal gliosis Neocortical resection Parietal Reactive Gliosis 3 39.6
P55 9 FCD in the right premotor region (Brodman 44-8-6), but also has
sulcation anomaly of the hand-knob Neocortical resection Frontal FCD-Type IIA 3 32.7
P56 26 Multiple lesions: large FCD in the right anterior ventral insula, FCD of the right superior frontal sulcus with tail sign, right anterior inter parietal sulcus FC , left inferior frontal gyrus (smaller) DOSD Neocortical resection Frontal Focal Cortical Dysplasia Type 1 Cannot be Ruled out. The fragmented nature of the Biopsy interferes with the assessment of Cortical Lamination. 3 14.1
P57 23 Right posterior cingulate / precuneus band heterotopia, extensive and involving the calcarine sulcus & cuneus.
HV - normal Radio frequency ablation Parietal - 3 12.8
P58 30 Left Frontal Premotor FCD- residual lesion Neocortical resection Frontal Focal Cortical Dysplasia Taylor Type IIB 3 12.9
P59 23 Bilateral parieto-occipital ulegyria, possible right > left
hippocampus is almost symmetric Neocortical resection Parietal Reactive Gliosis 4 12.7
P60 22 The gliosis appears more or less same as compared to previous MRI. Right hippocampus and amygdala are not visualized consistent with post operative status. Right mammillary body and right fornix show thinning. The Left hippcocampus, left mammillary body and left fornix appear normal. Radio frequency ablation Frontal - 4 15.7
P61 40 Normal Neocortical resection Parietal No Evidence to Suggest Focal Cortical Dysplasia 4 12.9
P62 26 Normal: post injury gliotic changes Neocortical resection Parietal No evidence of any Cortical dysplasia/Malignancy seen. 4 56.6
P63 34 Normal Neocortical resection Frontal Evidence of Cortical Dyslamination. 4 38.4

Table 4. Estimating region-specific normative HFO rates.
ctx-lh = left cortex and ctx-rh = right cortex.

Brain Structures Ripples Fast Ripples
Median 95th percentile Median 95th percentile Number of contacts per brain structure
ctx-caudalanteriorcingulate 0.62 1.1 0.04 0.04 37
ctx-caudalmiddlefrontal 0.26 0.29 0.08 0.12 4
ctx-cuneus 0.38 0.38 0 0 27
ctx-entorhinal 0.81 0.81 0 0 51
ctx-fusiform 0.32 0.45 0 0 45
ctx-inferiorparietal 0.97 1.67 0.24 0.24 33
ctx-inferiortemporal 0.35 1.15 0.09 0.09 40
ctx-isthmuscingulate 0.28 0.55 0.82 1.6 43
ctx-lateraloccipital 0.05 0.05 0 0 24
ctx-lateralorbitofrontal 0.65 1.26 0 0 29
ctx-lingual 1.39 4.15 0.32 0.43 41
ctx-medialorbitofrontal 0.1 0.19 0 0 42
ctx-middletemporal 0.59 0.96 0.07 0.07 21
ctx-parahippocampal 0.18 0.27 0.21 0.35 4
ctx-paracentral 3.44 4.5 0 0 18
ctx-parsopercularis 0.16 0.57 0.16 0.16 11
ctx-parsorbitalis 0.17 0.24 0.1 0.11 11
ctx-parstriangularis 0.13 0.27 0.12 0.35 28
ctx-pericalcarine 0.88 0.88 0 0 14
ctx-postcentral 0.46 1.39 0.16 0.2 21
ctx-posteriorcingulate 0.22 0.36 0.1 0.16 23
ctx-precentral 0.22 0.64 0.45 0.86 13
ctx-precuneus 0.57 1.54 0.1 0.1 23
ctx-rostralanteriorcingulate 0.1 0.1 0.05 0.05 20
ctx-rostralmiddlefrontal 0.11 0.31 0.06 0.06 22
ctx-superiorfrontal 0.14 0.35 0.06 0.08 33
ctx-superiorparietal 1.3 2.5 0.12 0.13 17
ctx-superiortemporal 0.3 0.38 0.15 0.19 14
ctx-supramarginal 0.61 1.69 0.03 0.03 2
ctx-frontalpole 0.18 0.18 0 0 11
ctx-temporalpole 0.46 0.29 0 0 10
ctx-transversetemporal 0.34 0.41 0 0 26
ctx-insula 0.3 0.82 0.04 0.04 36
Hippocampus 0.51 0.68 0.21 0.23 112
Amygdala 0.72 1.49 0.07 0.07 6
wm-caudalanteriorcingulate 0.05 0.06 0.04 0.04 60
wm-caudalmiddlefrontal 0.4 0.67 0.05 0.05 6
wm-cuneus 2.81 2.81 0 0 30
wm-entorhinal 0.08 0.08 0 0 51
wm-fusiform 0.24 0.53 0.05 0.05 68
wm-inferiorparietal 0.63 0.81 0.35 0.36 39
wm-inferiortemporal 0.46 1.14 0.14 0.14 35
wm-isthmuscingulate 0.2 0.22 0 0 37
wm-lateraloccipital 0.07 0.08 0.1 0.1 33
wm-lateralorbitofrontal 0.12 0.18 0.21 0.35 30
wm-lingual 0.29 0.46 0.03 0.03 37
wm-medialorbitofrontal 1.42 2.65 0.6 1.02 11
wm-middletemporal 1.85 3.67 0.11 0.11 30
wm-parahippocampal 1.41 4.46 0.48 1.24 12
wm-paracentral 0.09 0.14 0.04 0.04 21
wm-parsopercularis 0.17 0.32 0.06 0.08 21
wm-parsorbitalis 0.24 0.75 0.15 0.26 17
wm-parstriangularis 0.4 0.61 0.04 0.04 13
wm-pericalcarine 0.11 0.11 0 0 20
wm-postcentral 0.51 1.01 0.17 0.19 25
wm-posteriorcingulate 0.11 0.18 0.12 0.12 23
wm-precentral 1.21 2.96 0.29 0.65 26
wm-precuneus 0.54 0.67 0 0 34
wm-rostralanteriorcingulate 0 0 0 0 13
wm-rostralmiddlefrontal 0.2 0.76 0.05 0.09 21
wm-superiorfrontal 0.21 0.78 0.11 0.11 33
wm-superiorparietal 0.1 0.11 0 0 19
wm-superiortemporal 0.63 1 0.16 0.18 33
wm-supramarginal 0.1 0.12 0.11 0.11 2
wm-frontalpole 0 0 0 0 15
wm-temporalpole 0.05 0.05 0 0 8
wm-transversetemporal 0.06 0.09 0.04 0.04 25
wm-insula 0.06 0.14 0.09 0.12 16

Table 5. Localizing EZ in seizure free patients by comparing with region-specific normal HFO rate database. Optimum HFO rate ratio: ie., patient 1 - HFO rate ratio < 9 would lead to more cortical structures beyond these 12.
Patient No. DK Atlas-based structures resected

Total number of structures resected (TSR)
Ripples
Fast Ripples
Concordance to resection cavity
(Yes / No)
Calculated EZ based on HFO

Percentage of Resection based on calculated EZ (PR) Optimum HFO rate ratio for 100 % match Calculated EZ based on HFO

Percentage of Resection based on calculated EZ (PR) Optimum HFO rate ratio for 100 % match
P1 rh-insula
rh-lateral orbitofrontal
rh-medial orbitofrontal
rh-middle temporal
rh-pars opercularis
rh-pars orbitalis
rh-pars triangularis
rh-rostral anterior cingulate
rh-rostral middle frontal
rh-superior temporal
Right-Amygdala,
Right-Hippocampus

TSR = 12 rh-insula
rh-lateral orbitofrontal
rh-medial orbitofrontal
rh-pars opercularis
rh-pars triangularis
rh-rostral anterior cingulate
Right-Amygdala

PR = 1 + (5/ TSR)
= 142 % 9 rh-insula
rh-medial orbitofrontal
rh-pars triangularis
rh-rostral anterior cingulate
rh-rostral middle frontal
Right-Amygdala
Right-Hippocampus

PR = 1 + (5/ TSR)
= 142 % 3 Yes
P2 rh-supramarginal

TSR = 1
rh-supramarginal

PR = 1 + (0/ TSR)
= 100 % 5.9 rh-supramarginal

PR = 1 + (0/ TSR)
= 100 % 5.2 Yes
P3 rh-lateral orbitofrontal
rh-pars orbitalis
rh-pars triangularis
rh-rostral middle frontal

TSR = 4
rh-lateral orbitofrontal
rh-pars orbitalis
rh-pars triangularis
rh-rostral middle frontal

PR = 1 + (0/ TSR)
= 100 % 1 rh-pars orbitalis
rh-pars triangularis
rh-rostral anterior cingulate

PR = 1 + (0/ TSR)
= 100 % 1 Yes
P4 Left-Amygdala
Left-Hippocampus
lh-middle temporal
lh-superior temporal

TSR = 4

Left-Amygdala
Left-Hippocampus

PR = 1 + (2/TSR)
= 150 % 3 Left-Amygdala
Left-Hippocampus

PR = 1 + (2/TSR)
= 150 % 2.1 Yes
P5 rh-inferior temporal
rh-middle temporal
Right-Hippocampus

TSR = 3
Left-Amygdala
Left-Hippocampus
Right-Hippocampus

PR = 1 + (2/TSR)
= 166 % 3.5 Left-Hippocampus
lh-superior temporal
Right-Hippocampus

PR = 1 + (1/TSR)
= 166 % 2 Yes
P6 lh-insula
lh-lateral orbitofrontal
lh-pars orbitalis

TSR = 3 lh-lateral orbitofrontal
lh-pars orbitalis

PR = 1 + (1/TSR)
= 133 % 1.3 - - Yes
P7 lh-cuneus
lh-precuneus

TSR = 2
lh-cuneus

PR = 1 + (1/TSR)
= 150 % 10 - - Yes
P8 rh-rostral middle frontal
rh-superior frontal

TSR = 2
rh-rostral middle frontal
rh-superior frontal

PR = 1 + (0/TSR)
= 100 % 5.6 rh-rostral middle frontal
rh-superior frontal

PR = 1 + (0/TSR)
= 100 % 1 Yes
P9 rh-insula
rh-pars triangularis
rh-precentral
rh-superior temporal

TSR = 4 rh-insula

PR = 1 + (3/TSR)
= 175 % 1
rh-insula

PR = 1 + (5/TSR)
= 175 %
1.5 Yes
P10 Left-Hippocampus
Left-Amygdala

TSR = 2
Left-Hippocampus
Left-Amygdala

PR = 1 + (0/TSR)
= 100 % 2.4 Left-Hippocampus
Left-Amygdala

PR = 1 + (0/TSR)
= 100 % 4.9 Yes
P11 rh-caudal middle frontal
rh-superior frontal

TSR = 2
rh-caudal middle frontal
rh-superior frontal

PR = 1 + (0/TSR)
= 100 % 2 rh-caudal middle frontal
rh-superior frontal

PR = 1 + (0/TSR)
= 100 % 1 Yes
P12 rh-fusiform
rh-inferior temporal
rh-middle temporal
Right-Amygdala
Right-Hippocampus

TSR = 5

Right-Amygdala
Right-Hippocampus

PR = 1 + (3/TSR)
= 160 %
5 Right-Hippocampus

PR = 1 + (4/TSR)
= 180 % 1 Yes
P13 Left-Amygdala
Left-Hippocampus
lh-inferior temporal
lh-lingual
lh-parahippocampal
lh-superior temporal

TSR = 6
lh-lingual

PR = 1 + (5/TSR)
= 183 %
4.8
lh-lingual

PR = 1 + (5/TSR)
= 183 % 1.7 Yes
P14 lh-posterior cingulate
lh-superior frontal

TSR = 2
- - - - No
P15 lh-fusiform
lh-lateral occipital
lh-lingual
lh-parahippocampal
lh-pericalcarine
lh-precuneus

TSR = 6
lh-lateral occipital
lh-parahippocampal

PR = 1 + (4/TSR)
= 167 % 2 - - Yes
P16 Left-Hippocampus
lh-insula
lh-parahippocampal

TSR = 3
lh-insula
lh-parahippocampal
lh-superior temporal

PR = 1 + (1/TSR)
= 133 %
1.1 lh-parahippocampal
lh-superior temporal

PR = 1 + (5/TSR)
= 133 % 1.1 Yes
P17 Left-Amygdala
Left-Hippocampus
lh-entorhinal
lh-fusiform
lh-inferior temporal
lh-superior temporal

TSR = 6
Left-Amygdala
Left-Hippocampus
lh-superior temporal

PR = 1 + (3/TSR)
= 150 %

8.3 Left-Hippocampus
Left-Amygdala

PR = 1 + (4/TSR)
= 166 % 2.8 Yes
P18 Left-Amygdala
Left-Hippocampus
lh-fusiform
lh-inferior temporal
lh-middle temporal
lh-superior temporal
lh-temporal pole

TSR = 7
Left-Amygdala
Left-Hippocampus
lh-fusiform
lh-inferior temporal

PR = 1 + (3/TSR)
= 150 %
4 Left-Hippocampus
lh-fusiform

PR = 1 + (5/TSR)
= 133 %
1.5 Yes
P19 rh-entorhinal
rh-fusiform
rh-inferior temporal
rh-middle temporal
rh-temporal pole
Right-Hippocampus

TSR = 6
rh-fusiform
rh-inferior temporal
rh-temporal pole
Right-Hippocampus

PR = 1 + (2/TSR)
= 133 % 8 Right-Hippocampus

PR = 1 + (5/TSR)
= 183 % 2 Yes
P20 rh-caudal middle frontal
rh-superior frontal

TSR = 2
rh-caudal middle frontal
rh-superior frontal

PR = 1 + (0/TSR)
= 100 % 1 rh-caudal middle frontal

PR = 1 + (1/TSR)
= 150 %
1 Yes
P21 rh-fusiform
rh-inferior temporal
rh-lateral orbitofrontal
rh-lingual
rh-medial orbitofrontal
rh-parahippocampal
rh-rostral middle frontal
Right-Hippocampus

TSR = 8 Right-Hippocampus

PR = 1 + (7/TSR)
= 188 % 3.1 Right-Hippocampus

PR = 1 + (7/TSR)
= 188 % 1 Yes
P22 Left-Amygdala
Left-Hippocampus
lh-entorhinal
lh-fusiform
lh-inferior temporal
lh-lingual
lh-middle temporal

TSR = 7 Left-Hippocampus

PR = 1 + (6/TSR)
= 186 % 2 Left-Hippocampus

PR = 1 + (6/TSR)
= 186 % 1 Yes
P23 lh-caudal middle frontal
lh-precentral

TSR = 2 lh-caudal middle frontal
lh-precentral
lh-superior frontal
lh-superior temporal

PR = 1 + (0/TSR)
= 100 % 2 - - Yes
P24 rh-insula
rh-lateral orbitofrontal
rh-medial orbitofrontal
rh-pars opercularis
rh-superior temporal
Right-Amygdala
Right-Hippocampus

TSR = 7 - - rh-pars opercularis
Right-Hippocampus
Right-Amygdala

PR = 1 + (4/TSR)
= 157 % 1 Yes
P25 rh-inferior temporal
rh-middle temporal
rh-parahippocampal
rh-superior temporal
rh-temporal pole
Right-Amygdala
Right-Hippocampus

TSR = 7
- - Right-Hippocampus
rh-parahippocampal

PR = 1 + (5/TSR)
= 171 % 10 Yes
P26 lh-supramarginal
lh-precentral
lh-transverse temporal

TSR = 3
diffuse - - - No
P27 rh-superior frontal

TSR = 1 rh-superior frontal

PR = 1 + (0/TSR)
= 100 % 3.9 - - Yes
P28 rh-insula
rh-supramarginal

TSR = 2 rh-supramarginal

PR = 1 + (1/TSR)
= 150 % 4 no HFO - Yes
P29 rh-precuneus
rh-superior parietal

TSR = 2
rh-precuneus
rh-superior parietal

PR = 1 + (0/TSR)
= 100 % 1.5 rh-precuneus

PR = 1 + (1/TSR)
= 150 % 2.5 Yes
P30 lh-rostral middle frontal

TSR = 1 lh-rostral middle frontal

PR = 1 + (0/TSR)
= 100 % 11.9 lh-rostral middle frontal

PR = 1 + (0/TSR)
= 100 % 5.3 Yes
P31 Left-Hippocampus
lh-insula
lh-superior temporal

TSR = 3
lh-superior temporal

PR = 1 + (2/TSR)
= 142 % 2 lh-superior temporal
lh-insula

PR = 1 + (1/TSR)
= 133 % 2 Yes
P32 lh-insula
lh-pars triangularis

TSR = 2 lh-pars triangularis

PR = 1 + (1/TSR)
= 150 % 1 no HFOs - Yes
P33 Right-Amygdala
rh-parahippocampal
rh-inferior temporal

TSR = 3 rh-parahippocampal

PR = 1 + (2/TSR)
= 167 % 2 rh-parahippocampal

PR = 1 + (2/TSR)
= 167 % 2.5 Yes
P34 lh-insula
lh-superior temporal

TSR = 2 lh-insula

PR = 1 + (1/TSR)
= 150 % 1 lh-insula

PR = 1 + (1/TSR)
= 150 % 1 Yes
P35 lh-posterior cingulate

TSR = 1 no HFOs no HFOs No
P36 Left-Hippocampus

TSR = 1 Left-Hippocampus

PR = 1 + (0/TSR)
= 100 % 1.8 - - Yes
P37 rh-precuneus

TSR = 1
rh-precuneus

PR = 1 + (0/TSR)
= 100 %
4 rh-precuneus
rh-supramarginal

PR = 1 + (0/TSR)
= 100 % 16 Yes

Table 6. Localizing EZ in non-seizure free patients by comparison with region- specific normal HFO rate database.
lh = left and rh=right
Patient No. DK Atlas-based structures resected

Total number of structures resected (TSR)
Ripples Fast Ripples Concordance to resection cavity
(Yes / No)
Calculated EZ based on HFO

Percentage of Resection (PR) = Calc EZ / Calc EZ included in resection
Optimum HFO rate ratio Calculated EZ based on HFO

Percentage of Resection (PR) = Calc EZ / Calc EZ included in resection Optimum HFO rate ratio
P38 lh-insula
lh-lateral orbitofrontal
lh-medial orbitofrontal
lh-pars triangularis
lh-rostral middle frontal
lh-superior frontal

TSR = 6 lh-rostral middle frontal
lh-pars triangularis

PR= Calc EZ / Calc EZ included in resection

PR = 100 % 5.8 - 2.7 Concordant
P39 rh-postcentral

TSR = 1 rh-postcentral

PR = 100 % 5.8 rh-postcentral

PR = 100 % 2.7 Concordant
P40 Right-Hippocampus
Right-Amygdala

TSR = 2 - 5.8 Left-Hippocampus

PR = 0 % 2.7 Discordant
P41 rh-inferior parietal
rh-inferior temporal
rh-isthmus cingulate
rh-lateral occipital
rh-lingual
rh-middle temporal
rh-precuneus
rh-superior parietal

TSR = 8 almost all channels

5.8 almost all channels
2.7 almost all channels
P42 Left-Hippocampus
lh-entorhinal
lh-inferior temporal
lh-middle temporal
lh-superior temporal
lh-temporal pole

TSR = 6 - 5.8 lh-middle temporal

PR = 100 % 2.7 Concordant
P43 lh-superior frontal
lh-insula

TSR = 2 lh-insula
lh-pars orbitalis
lh-rostral middle frontal
lh-superior frontal
lh-superior temporal
lh-medial orbitofrontal
lh-rostral middle frontal

PR= 100%

5.8 lh-rostral middle frontal
lh-superior frontal

PR = 50 %
2.7 Concordant + Other area
P44 rh-medial orbitofrontal
rh-rostral anterior cingulate
rh-rostral middle frontal
rh-superior frontal

TSR = 4 lh-superior frontal
lh-caudal middle frontal

PR= 0%
5.8 - 2.7 Discordant
P45 lh-posterior cingulate
lh-caudal anterior cingulate rh-posterior cingulate

PR= 1 + (0/1)
= 100 % 5.8 - 2.7 Concordant
P46 lh-lateral orbitofrontal
lh-medial orbitofrontal
lh-pars orbitalis
lh-rostral anterior cingulate

TSR = 6 almost all channels 5.8 almost all channels 2.7 almost all channels
P47 lh-lateral occipital

TSR = 1 lh-lateral occipital
lh-pericalcarine
lh-superior parietal
rh-lateral occipital
rh-middle temporal
rh-pericalcarine

PR= 16 % 5.8 - 2.7 Concordant + Other area
P48 lh-inferior parietal
lh-lateral occipital

TSR = 2 lh-lateral occipital
lh-inferior parietal
lh-isthmus cingulate
lh-fusiform
lh-inferior temporal

PR= 40 % 5.8 lh-lingual
lh-inferior temporal
lh-inferior parietal
lh-superior frontal
lh-caudal middle frontal
lh-insula

PR= 17 % 2.7 Concordant + Other area
P49 lh-insula
lh-superior temporal
lh-inferior temporal

TSR = 3 - 5.8 Left-Amygdala

PR= 0 % 2.7 Discordant
P50 Left-Hippocampus
lh-inferior temporal
lh-lingual
lh-fusiform
lh-lateral occipital

TSR = 5
lh-lingual
lh-lateral occipital
lh-fusiform
rh-supramarginal
rh-lateral occipital
rh-fusiform

PR= 50 % 5.8 lh-lingual
Left-Hippocampus
rh-lateral occipital
PR= 33 % 2.7 Concordant + Other area
P51 lh-precuneus
lh-isthmus cingulate
TSR = 2
almost all channels 5.8 almost all channels 2.7 almost all channels
P52 rh-inferior parietal
rh-supramarginal
rh-superior temporal
rh-inferior parietal
rh-postcentral

TSR = 5 lh-middle temporal
rh-insula
rh-isthmus cingulate
rh-middle temporal
rh-posterior cingulate
rh-precuneus

PR= 0 % 5.8 rh-isthmus cingulate
rh-posterior cingulate

PR= 0%
2.7 Discordant
P53 lh-precuneus
lh-cuneus

TSR = 2 - 5.8 - 2.7 No detections
P54 lh-superior parietal
lh-inferior parietal
lh-pericalcarine
lh-lateral occipital
lh-precuneus

TSR = 5 lh-inferior parietal
lh-superior parietal
rh-lateral occipital
rh-parahippocampal
rh-superior parietal
rh-supramarginal

PR= 33 % 5.8 - 2.7 Discordant
P55 lh-superior frontal

TSR = 1 - 5.8 lh-superior frontal

PR= 100 % 2.7 Concordant
P56 lh-insula
lh-pars opercularis
lh-postcentral
lh-precentral - 5.8 - 2.7 No detections
P57 rh-posterior cingulate
rh-supramarginal

TSR = 6 rh-lateral occipital

PR = 0 % 5.8 - 2.7 Discordant
P58 lh-caudal middle frontal
lh-precentral
lh-rostral middle frontal
lh-superior frontal

TSR = 4 lh-supramarginal
lh-caudalanteriorcingulate

PR= 0 % 5.8 - 2.7 Discordant
P59 rh-inferior parietal
rh-superior parietal
rh-postcentral

TSR = 3 lh-inferior parietal
rh-inferior parietal
rh-lateral occipital
rh-paracentral
rh-postcentral
rh-superior parietal

PR= 50% 5.8 rh-postcentral
rh-supramarginal
lh-inferior parietal

PR= 100 % 2.7 Concordant + Other area
P60 rh-insula
rh-lateral orbitofrontal
rh-medial orbitofrontal
rh-pars opercularis
rh-pars triangularis
rh-rostral anterior cingulate
rh-rostral middle frontal

TSR = 7 lh-insula
lh-temporal pole
rh-insula
rh-lateral orbitofrontal
rh-medial orbitofrontal
rh-pars triangularis
rh-rostral anterior cingulate
rh-temporal pole

PR= 62% 5.8 rh-caudal middle frontal
rh-insula
rh-rostral middle frontal

PR= 33 % 2.7 Concordant + Other area
P61 lh-supramarginal
lh-superior temporal
rh-supramarginal
rh-inferior parietal
rh-posterior cingulate

PR= 0% 5.8 - 2.7 Discordant
P62 lh-inferior parietal
lh-precuneus

TSR = 4 lh-superior parietal
lh- precuneus

PR= 50% 5.8 - 2.7 Concordant + Other area
P63 lh-insula
lh-pars opercularis

TSR = 2 lh-insula
lh-lateral orbitofrontal
lh-medial orbitofrontal
lh-pars opercularis
lh-pars orbitalis
lh-precentral
lh-superior temporal
lh-supramarginal

PR= 25% 5.8 almost all channels 2.7 Concordant + Other area

 
, Claims:We claim:
1. A method for localizing epileptogenic zone using HFO analysis in SEEG wherein said method comprises of steps of:
- building region specific normative HFO database comprising of steps of:
• selecting patients (Pa) who had undergone epilepsy surgery to include seizure-free and non-seizure free patients (Pa) based on the availability of postoperative volumetric MRI scans and interictal HFO recordings for acquiring of data,
• acquiring of data comprising of steps of:
o collecting medical history concerning epilepsy surgery of the selected patients (Pa),
o acquiring volumetric T1- weighted and FLAIR MRI, of selected patients (Pa), referred to as preimplantation MRI scans,
o employing the preimplantation MRI scans specifically the T1 weighted MRI scans to construct three-dimensional representations of individual patient’s (Pa) brain surfaces,
o acquiring SEEG recordings, also referred to as SEEG signal, of the selected patients (Pa) for analysis, by implanting SEEG electrodes intracranially, and selecting a 30-minute interictal SEEG recording during non-rapid eye movement sleep in each patient (Pa),
o reviewing of recordings independently by experienced epileptologist and exporting them in EDF format for subsequent analysis and
o acquiring post-implantation CT scans of individual patients (Pa),
• detecting HFOs and identifying ripples in the range of 80–250 Hz and fast ripples in the range of 250-500 Hz from the selected SEEG recordings of individual patients (Pa),
• mapping patient (Pa) SEEG signal to brain structures using computational methods comprising of steps of:
o conducting cortical parcellations by using the three-dimensional representations of individual patient’s (Pa) brain surfaces and leveraging a brain atlas that provides a comprehensive cortical and subcortical map of individual patient’s (Pa) brains, facilitated by an analysis and visualization tool,
o co-registering post-implantation CT scans, pre-implantation MRI scans, and the cortical parcellations of the individual patients’ (Pa) brains using a processing tool,
o segmenting the post-implantation CT scans by partitioning the images into multiple segments, each corresponding to a different anatomical structure of the brain and
o mapping patient (Pa) SEEG contacts from SEEG recordings to their corresponding cortical location as per the brain atlas,
• identifying SEEG contacts located in the EZ comprising of steps of:
o co-registering the post-implantation CT scans with the postoperative MRI using mapping tool,
o identifying the SEEG contacts located within the resection cavity which defines the EZ in seizure-free patients (Pa) by careful visual examination of the co-registered images,
o classifying subsequently the HFOs identified from seizure free patients (Pa) (Engel 1) on the basis of HFOs identified in the resection cavity as EZ-HFOs, and HFOs identified in non-EZ areas as non-EZ HFOs and
o estimating the Percentage of Resection (PR) by measuring the calculated EZ that was actually resected compared to the cortical resection performed,
• estimating region-specific physiological (non-epileptogenic) HFO rates and database, comprising of steps of:
o defining normal brain area encompassing any non-lesional, non-epileptogenic, and non-irritative part of the cerebral cortex,
o estimating physiological HFO rates in the non-EZ areas using SEEG signals recorded from the normal brain area,
o estimating region-specific normal HFO database by computing physiological HFO rates from respective anatomical electrode locations across all Engel 1 patients (Pa) and
o computing standard deviation, and percentile for each cortical region in the brain atlas to build a region-specific normal (physiological) HFO rate database,
• defining region-specific HFO rate thresholds for localization of EZ, comprising of steps of:
o determining the EZ for each patient (Pa) as the SEEG contacts that showed a statistically significant deviation from the normative data,
o defining a parameter called the HFO rate ratio that is calculated as the quotient of a patient’s (Pa) region-specific HFO rate divided by the normal physiological HFO rates at the 95th percentile value, to accurately localize the EZ,
o estimating an optimum ratio for each seizure-free patient (Pa) by testing ratios at each level, at increments of 0.1 starting from 1, to detect pathological HFO rates,
o selecting the optimum ratio by finding the ratio that best matches the resection volume which defines the EZ in the seizure-free patients (Pa) and
o using the optimum ratios to localize the EZ in seizure-free and non-seizure free patients (Pa),
• performing statistical analysis to estimate the optimum ratios for localizing the EZ to include plotting a receiver operating characteristic (ROC) curve and testing whether the pathological HFOs rates accurately localize in seizure free patients (Pa), testing median statistical difference between the HFO rates in the EZ and non-EZ channels etc.,
- implementing the region-specific HFO rate ratio database and detection tool in a graphical user interface (GUI) and stored as a customized EZ delineation tool (A2) for automation and easy usability,
- localization of the Epileptogenic Zone (EZ) in patients (Pa) using region specific HFO rate ratio and database, comprising of steps of:
• acquiring of intracranial SEEG recordings of the subject/patient (Pa) to be studied and selecting a 30-minute SEEG recording during non-rapid eye movement sleep for analysis,
• detecting HFOs and identifying ripples and fast ripples from the selected SEEG recording of the subject/patient (Pa),
• mapping patient (Pa) SEEG contacts from SEEG recordings to their corresponding cortical location as per the brain atlas,
• calculating the region specific HFO rates for the patient (Pa) and comparing these with region specific normative HFO database,
• calculating HFO rate ratio as patient’s (Pa) region specific HFO rate divided by normal physiological region specific HFO rate at 95th percentile and
• analysing HFO rate ratio by comparing it with values above specific threshold to localize the epileptogenic zone in the patient (Pa) brain.
2. The method as claimed in claim 1, wherein said medical history concerning epilepsy surgery of the selected patients (Pa) comprises of baseline characteristics, pre and post - surgical or operative data to include MRI scans, histopathology, surgical outcomes of the selected patients (Pa) etc. collected retrospectively from electronic medical records.
3. The method as claimed in claim 1, wherein said defining of normal brain area is based as cortical regions in seizure-free patients (Pa) that fulfill the criteria below :
- non-lesional cortex on visual inspection of preoperative MRI by an experienced neuroradiologist,
- the corresponding SEEG contact does not fall within the irritative zone with interictal discharges and
- ensuring it did not overlap with the resection cavity on postoperative MRI.
4. A system (S) for localizing epileptogenic zone using HFO analysis in SEEG wherein said system (S) comprises of:
- at least one data acquisition module (D) for acquiring data from a group of epilepsy patients (Pa) who have undergone epilepsy surgery, for preparation of region specific HFO rate database and subsequently for delineation of EZ in a patient (Pa) under study, said module (D) comprising of :
• at least one SEEG submodule (D1) for acquiring SEEG recordings of patient(s) (Pa), said submodule (D1) comprising of a SEEG system,
• at least one CT submodule (D2) for acquiring post implantation of SEEG electrodes- CT scans of patient(s) (Pa), said submodule comprising of a CT machine system,
• at least one MRI submodule (D3) for acquiring preimplantation of SEEG electrodes- volumetric T1- weighted and FLAIR MRI scans, said submodule comprising of an MRI system and
• at least one medical records submodule (D4) for acquiring medical history of patients (Pa), said submodule comprising of medical records in health care systems,
- at least one processing module (P) for processing of acquired data from data acquisition module (D), said processing (P) module comprising of:
• at least one HFO detector submodule (P1) for detecting HFOs from SEEG recordings, said submodule comprising of HFO detector system,
• at least one brain atlas tool (P2) that provides a comprehensive cortical and subcortical map of patient’s (Pa) brains and thus provides a spatial framework and standardized labels for brain regions, facilitating processing, research, analysis etc.,
• at least one tool for analysis and visualization (P3) that facilitates usage of brain atlas for cortical parcellations,
• at least one processing tool (P4) that facilitates co-registering of post-implantation CT scans, preimplantation MRI scans, and the cortical parcellations for enabling spatial alignment across different imaging modalities,
• at least one mapping tool (P5) that facilitates co registering of post-implantation CT scans with post operative MRI scans that facilitates identification of the SEEG contacts in the epileptogenic zone and
• at least one memory and storage unit (P6) for storage of acquired or processed data,
- at least one analysis and display module (A) for estimating region specific HFO rates database and subsequently facilitating delineation of EZ in patients (Pa) under study, said module comprising of:
• at least one computing unit (A1) for housing the various tools and units of the system,
• at least one customized EZ delineation tool (A2) that facilitates derivation of region specific HFO rates and database, HFO rate ratios, detection of pathological HFO rates, selection of optimum HFO rate ratios, delineation of EZ etc., and
• at least one display unit (A3) with GUI for visualization and communication of data and results
wherein
- the system facilities use of HFO rates to identify the epileptogenic zone (EZ) by distinguishing it from physiological HFO rates detected in interictal SEEG recordings,
- the system (S) provides a region-specific normative HFO database that is retrospectively constructed using interictal physiological HFO detected from non-resected brain areas of a group of patients (Pa) who remained seizure-free after epilepsy surgery,
- the system (S) provides and uses region-specific HFO rate ratio and specific thresholds in a method to aid in the detection of region-specific pathological HFOs to localize the EZ,
- acquiring of SEEG recordings is done by implanting SEEG electrodes intracranially and selecting a 30-minute SEEG recording during non-rapid eye movement sleep in each patient (Pa) and
- the system (S) facilitates delineation of EZ in a patient (Pa) under study by detecting HFOs from the SEEG recordings and comparing the patient’s (Pa) HFO rate with region specific normative database, deriving the HFO rate ratio and delineation of EZ based on the ratios exceeding specific threshold
that enables a precise, accurate, reliable and computationally efficient mapping of region-specific physiological HFO rates of the cerebral cortex of the human brain for improved epileptogenic zone delineation from stereo-EEG in patients (Pa) with drug-refractory epilepsy.
5. The system (S) claimed in claim 4, wherein said SEEG system comprises of SEEG electrodes intracranially implanted using preferably a SEEG robotic or frame-based technique and the SEEG recordings are undertaken using systems selected from Natus EEG Monitoring system, any other clinically validated EEG system, etc.
6. The system (S) claimed in claim 4, wherein said CT machine systems are selected from Siemens, GE etc.
7. The system (S) claimed in claim 4, wherein said MRI systems are selected from Siemens Biograph mMR system, GE Discovery MR 750W, any 3T MRI Machine etc., preferably the Siemens Biograph mMR system.
8. The system (S) as claimed in claim 4, wherein said brain atlas is selected from Desikan–Killiany (DK) atlas, VEP atlas etc., preferably the DK atlas.
9. The system (S) as claimed in claim 4, wherein said HFO detector system is selected from Montreal Neurological Institute (MNI) HFO detector, HFOApp, pyHFO, Spiking Neural Network (SNN) based system, any HFO detector systems etc., preferably the MNI HFO detector.
10. The system (S) as claimed in claim 4, wherein said analysis and visualization tool is selected from FreeSurfer 6.3 or 7 etc.
11. The system (S) as claimed in claim 4 wherein the processing tool is selected from any version of Gardel tool etc.
12. The system (S) as claimed in claim 4, wherein said mapping tool is selected from to Statistical Parametric Mapping (SPM) software version 8 or 12 etc.
13. The system (S) as claimed in claim 4, wherein said system (S) finds application in the domain of neurology, specifically within the field of epilepsy for diagnosis, treatment planning (including surgery), and understanding the mechanisms of epileptic seizures. etc.

Dated this the 14th day of August 2025

________________________
Daisy Sharma
IN/PA- 3879
of SKS Law Associates
Attorney for the Applicant

To
The Controller of Patents,
The Patent Office, Chennai