Description:Various embodiments of the disclosure are discussed in detail below. While
20 specific implementations are discussed, it should be understood that this is done
for illustration purposes only. A person skilled in the relevant art will recognize
that other components and configurations may be used without parting from the
spirit and scope of the disclosure. Thus, the following description and drawings
are illustrative and are not to be construed as limiting. Numerous specific details
25 are described to provide a thorough understanding of the disclosure. However, in
certain instances, known details are not described in order to avoid obscuring the
description.
9
References to one or an embodiment in the present disclosure can be references to
the same embodiment or any embodiment; and such references mean at least one
of the embodiments.
Reference to "one embodiment", "an embodiment", “one aspect”, “some aspects”,
“an aspect” means that a particular feature, structure, or characteristic 5 described in
connection with the embodiment is included in at least one embodiment of the
disclosure. The appearances of the phrase "in one embodiment" in various places
in the specification are not necessarily all referring to the same embodiment, nor
are separate or alternative embodiments mutually exclusive of other embodiments.
10 Moreover, various features are described which may be exhibited by some
embodiments and not by others.
The terms used in this specification generally have their ordinary meanings in the
art, within the context of the disclosure, and in the specific context where each
term is used. Alternative language and synonyms may be used for any one or
15 more of the terms discussed herein, and no special significance should be placed
upon whether or not a term is elaborated or discussed herein. In some cases,
synonyms for certain terms are provided.
A recital of one or more synonyms does not exclude the use of other synonyms.
The use of examples anywhere in this specification including examples of any
20 terms discussed herein is illustrative only and is not intended to further limit the
scope and meaning of the disclosure or of any example term. Likewise, the
disclosure is not limited to various embodiments given in this specification.
Without intent to limit the scope of the disclosure, examples of instruments,
apparatus, methods and their related results according to the embodiments of the
25 present disclosure are given below. Note that titles or subtitles may be used in the
examples for convenience of a reader, which in no way should limit the scope of
the disclosure. Unless otherwise defined, technical and scientific terms used
herein have the meaning as commonly understood by one of ordinary skill in the
10
art to which this disclosure pertains. In the case of conflict, the present document,
including definitions will control.
Additional features and advantages of the disclosure will be set forth in the
description which follows, and in part will be obvious from the description, or can
be learned by practice of the herein disclosed principles. 5 The features and
advantages of the disclosure can be realized and obtained by means of the
instruments and combinations particularly pointed out in the appended claims.
These and other features of the disclosure will become more fully apparent from
the following description and appended claims or can be learned by the practice of
10 the principles set forth herein.
As mentioned above, there is a significant and unmet need for a system that can
ingest multi-dimensional patient data, compute a disease severity index using a
robust scoring algorithm, classify individuals into actionable risk tiers, and
generate care recommendations to support proactive care delivery.
15 Referring to Figure 1, according to an aspect of the present disclosure, a system
(100) for determining a disease severity index (DSI) score and predicting
healthcare resource utilization is illustrated. The system (100) is designed to
address the gap in personalized healthcare prediction by transforming raw,
multidimensional patient data into a structured score that quantifies disease
20 burden and future resource demands. The system (100) comprises a plurality of
interconnected modules that together enable the dynamic computation of the DSI
score and generation of actionable care insights.
In accordance with aspects of the present disclosure, the system (100) includes a
data acquisition module (10) that may be configured to receive one or more
25 patient-specific input data from internal and/ or external data sources. The data
acquisition module (10) may include electronic medical record (EMR) systems,
wearable health devices, insurance claims databases, mobile health platforms, and
manual data-entry portals. The input data may encompass demographic attributes
such as age, gender, and geographic location; clinical indicators including
11
comorbid conditions, diagnosis codes, vitals, and medication history; laboratory
results such as blood glucose, creatinine, lipid profiles, and inflammatory
markers; lifestyle-related variables like physical activity, diet, smoking status, and
alcohol consumption; as well as social determinants of health such as income,
employment, housing status, and 5 access to care.
In some aspects of the present disclosure, the data acquisition module (10) may
support real-time data streaming from connected devices or batch uploads from
enterprise systems using standard healthcare data exchange formats such as HL7,
FHIR, or CCD.
10 In accordance with aspects of the present disclosure, the data acquisition module
(10) may be operatively coupled to a data pre-processing module (20), which is
configured to clean, normalize, and validate the incoming data for further
computation. This data acquisition module (20) may include one or more rulebased
data validation engines, missing data imputation algorithms, value range
15 checks, unit conversions, and standardization of categorical variables using
clinically accepted taxonomies (e.g., ICD-10, LOINC). The pre-processing may
ensure that all data entering the scoring pipeline is harmonized and usable for
downstream modeling.
In some aspects of the present disclosure, the module (20) may also support data
20 de-duplication, time series alignment, and feature extraction based on domainspecific
logic.
In accordance with aspects of the present disclosure, the refined data may be
transmitted to a scoring engine (30) comprising at least one processor (31). The
scoring engine (30) may apply a proprietary algorithm to compute a Disease
25 Severity Index (DSI) score for the patient. The algorithm may incorporate a
weighted sum of condition-specific severity indicators, lab result thresholds,
behavioral adherence metrics, temporal health trends, and contextual risk signals
derived from geographic or population-level data. The resulting score is a
12
numerical value (e.g., 0–100) that quantifies the projected healthcare resource
consumption of the patient over a defined future period.
In some aspects of the present disclosure, the algorithm may be static (rule-based),
dynamically trained using supervised machine learning, or hybrid (combining
rules with 5 model inference).
In some aspects of the present disclosure, the scoring engine (30) may be updated
periodically through feedback loops from actual patient outcomes using the model
refinement engine (80), which will be described later in the description.
In accordance with aspects of the present disclosure, the DSI score generated by
10 the scoring engine (30) may be fed into a categorization module (40), which maps
the score to a predefined risk tier. The tiers may include, for example: no risk (0),
low risk (1), moderate risk (2), rising risk (3), and high risk (4). These thresholds
are configurable and may be customized based on provider-specific clinical
guidelines or insurer-defined stratification rules. In some embodiments, the
15 categorization module (40) may also incorporate clinical override rules, such as
reclassifying a patient with a recent hospitalization into a higher tier regardless of
algorithmic output. The assigned risk category becomes the basis for targeted
intervention planning and care coordination.
In accordance with aspects of the present disclosure, based on the risk category, a
20 recommendation engine (50) may be triggered to generate one or more care or
resource allocation suggestions. These may include, but not limited to, scheduling
follow-up visits, assigning a case manager, initiating home-based monitoring, or
triggering automated nudging sequences through patient communication
platforms. The recommendations can be tailored to the risk tier, with high-risk
25 patients receiving intensive outreach and no-risk patients receiving low-frequency
wellness reminders. The engine (50) may also integrate with external clinical
decision support systems to fetch protocol-based action sets.

In some aspects of the present disclosure, the engine (50) may prioritize
recommendations based on intervention cost-effectiveness or provider resource
constraints.
In accordance with aspects of the present disclosure, the outputs of the scoring
engine (30), categorization module (40), and recommendation 5 engine (50) may be
visually rendered via a user interface module (60). This module (60) may be
accessible to various stakeholders, including, but not limited to, primary care
physicians, specialist providers, care coordinators, health insurers, or population
health managers. The module (60) may display patient-level dashboards showing
10 the DSI score, assigned risk tier, underlying data contributors, and recommended
actions. The module (60) may also offer visualization features such as risk trend
graphs, forecasted healthcare costs, and comparison against cohort benchmarks.
For population-level users, the interface (60) may include aggregated DSI
heatmaps, resource planning charts, and provider-level performance dashboards.
15 In some aspects of the present disclosure, the module (60) may be implemented as
a web-based application, embedded widget, or secure API for integration into
existing EHR or payer systems.
In accordance with aspects of the present disclosure, the system (100) may include
a data storage module (70) configured to store historical patient data, calculated
20 DSI scores, intervention logs, and outcome records for longitudinal analysis. This
module (70) may enable retrospective model audits, patient trajectory tracking,
and care impact assessments. The stored data may also feed into the model
refinement engine (80), which uses one or more supervised machine learning
(ML) techniques to improve the performance of the scoring engine (30). The
25 engine (80) may retrain model weights based on actual utilization patterns, missed
escalations, or successful interventions, thus enabling continuous performance
optimization.
In operation, the system (100) for determining a disease severity index (DSI)
score and predicting healthcare resource utilization functions by first receiving
14
patient-specific input data through the data acquisition module (10). This data
may include demographic characteristics, clinical conditions, laboratory test
results, lifestyle factors, and social determinants of health collected from multiple
sources such as electronic medical records (EMRs), wearable devices, insurance
databases, and digital health platforms. The received data is 5 then passed to the
data pre-processing module (20), which cleanses, normalizes, and validates the
input to ensure consistency and accuracy. Once the data is pre-processed, it is
analyzed by the scoring engine (30), which applies a proprietary algorithm using
one or more processors (31) to compute a DSI score - a numerical representation
10 of the predicted healthcare resource burden associated with the patient. The
computed DSI score is subsequently sent to the categorization module (40), where
it is classified into one of several predefined risk tiers such as no risk, low risk,
moderate risk, rising risk, or high risk, based on configurable threshold values.
Based on the risk category, the recommendation engine (50) generates one or
15 more personalized care recommendations or resource planning suggestions,
including follow-up scheduling, escalation alerts, or behavioral nudging protocols.
The output, including the DSI score, assigned risk category, and suggested
actions, is then visually presented to relevant stakeholders - such as healthcare
providers, care coordinators, or insurers - through the user interface module (60).
20 Further, all relevant data and outputs are stored in a data storage module (70) for
trend analysis, auditing, and future retrieval, while a model refinement engine (80)
may periodically retrain or adjust the scoring algorithm using real-world patient
outcomes to enhance the accuracy and adaptability of the system over time. This
end-to-end operation enables intelligent, personalized, and proactive decision25
making in clinical and administrative healthcare environments.
In one exemplary embodiment, a 65-year-old female patient with a history of
chronic kidney disease and type 2 diabetes visits a tertiary care hospital for
routine follow-up. The hospital's electronic medical record (EMR) system is
integrated with the data acquisition module (10) of the DSI system (100),
30 allowing automatic retrieval of her demographic details, clinical conditions, lab
15
results such as creatinine and HbA1c levels, and social factors including
transportation limitations. The data pre-processing module (20) cleans and
standardizes these inputs, correcting unit inconsistencies and identifying missing
fields. The scoring engine (30) computes a DSI score of 82 using a proprietary
algorithm that assigns high weightage to renal impairment and 5 erratic glucose
control. The categorization module (40) classifies the score into a "high risk" tier.
The recommendation engine (50) responds by suggesting weekly remote
monitoring, bi-monthly nephrology consultations, and enrollment into a homebased
nutrition coaching program. These recommendations, along with the score
10 and patient summary, are displayed to her care coordinator through the user
interface module (60). The patient's longitudinal data is stored in the data storage
module (70), enabling trend analysis over successive visits, while the model
refinement engine (80) may eventually use the patient’s outcomes to adjust the
algorithm’s weightings.
15 In another exemplary embodiment, a health insurance company deploys the
system (100) across its network to proactively identify members at risk of
hospitalization. The data acquisition module (10) collects structured claims data,
pharmacy refill logs, and health survey responses from over 10,000 policyholders.
The data pre-processing module (20) aligns time-series lab results and removes
20 redundant data entries. The scoring engine (30) identifies a 45-year-old male
policyholder with borderline hypertension, obesity, and recent mental health
claims, and assigns him a DSI score of 58. The categorization module (40) places
him in the "rising risk" tier. Based on this, the recommendation engine (50)
advises assigning a health coach, scheduling a preventive care screening, and
25 pushing behavioral nudges via a connected mobile health app. The user interface
module (60) displays this profile to the insurer’s population health manager in a
dashboard view, while the patient's journey is continuously tracked and updated in
the data storage module (70).
In another exemplary embodiment, a public health agency uses the system (100)
30 to assess population risk during an influenza season. The data acquisition module
16
(10) interfaces with wearable device providers and regional EMR repositories to
collect anonymized patient vitals, symptom reports, and vaccination status. The
data pre-processing module (20) maps location-based risk variables and social
exposure indices. The scoring engine (30) identifies several individuals over the
age of 70 with declining activity levels and recent respiratory 5 complaints. A DSI
score of 73 is assigned to one such individual. The categorization module (40)
places him in the “moderate risk” tier. The recommendation engine (50) generates
a flu-testing recommendation, schedules a teleconsultation, and pushes locationbased
alerts regarding local outbreaks. Through the user interface module (60), a
10 public health coordinator reviews a geospatial risk heatmap and deploys resources
such as mobile clinics and vaccination drives. Meanwhile, aggregated data is
stored in the data storage module (70), and used to update model sensitivity
parameters in the model refinement engine (80).
In another exemplary embodiment, a telemedicine platform integrates the system
15 (100) to support virtual chronic disease management. A 52-year-old female
patient logs into her care app, and wearable sensor data is sent to the data
acquisition module (10), including step count, blood pressure, sleep patterns, and
dietary logs. These inputs are filtered through the data pre-processing module
(20), which fills missing values and flags inconsistent measurements. The scoring
20 engine (30) calculates a DSI score of 39, which the categorization module (40)
classifies as “moderate risk.” The recommendation engine (50) assigns her to a biweekly
coaching cycle with tailored lifestyle goals. This care plan and risk level
are visualized in the user interface module (60), which also shows improvements
in her score over time. The backend stores her evolving health records in the data
25 storage module (70), and insights from her adherence patterns help retrain the
scoring model through the model refinement engine (80).
These embodiments demonstrate the adaptability and scalability of the system
(100) across hospitals, insurers, public health networks, and digital care platforms.
By integrating structured and unstructured data, computing a dynamic disease
30 severity score, and generating targeted care recommendations, the system
17
provides real-time decision support and personalized interventions that improve
patient outcomes and optimize resource utilization.
Figure 2 illustrates a method (200) for determining a disease severity index (DSI)
score and predicting healthcare resource utilization using the system (100). The
method (200) includes the 5 following steps.
At step (201), one or more patient-specific input data is received through a data
acquisition module (10). The input data includes clinical, demographic, lifestyle,
and social health indicators sourced from EMR systems, insurer databases,
wearable devices, or digital health platforms.
10 At step (202), the received data is passed to a data pre-processing module (20),
where it is cleaned, normalized, and validated. This step ensures the accuracy and
compatibility of data by performing value standardization, range verification, and
missing data imputation.
At step (203), a scoring engine (30) applies a proprietary algorithm to the pre15
processed data and computes a numerical disease severity index (DSI) score,
indicating the projected level of healthcare resource utilization associated with the
patient.
At step (204), the computed DSI score is evaluated by a categorization module
(40), which assigns the patient to a predefined risk tier such as no risk, low risk,
20 moderate risk, rising risk, or high risk based on configurable thresholds.
At step (205), a recommendation engine (50) generates one or more personalized
care recommendations, which may include scheduling of follow-ups, physician
escalations, or targeted intervention workflows, based on the risk tier
classification.
25 At step (206), the DSI score, assigned risk category, and care recommendations
are presented to relevant stakeholders—including healthcare providers, care
managers, or insurers—through a user interface module (60), facilitating timely
and data-driven decision-making.
18
In some aspects of the present disclosure, the DSI score is periodically
recalculated based on updated lab results, adherence behavior, or new clinical
conditions.
In some aspects of the present disclosure, the method (200) further comprising the
step of storing the DSI scores and associated patient metadata 5 in a data storage
module (70) for historical trend analysis.
In some aspects of the present disclosure, the scoring engine (30) is updated using
feedback data and machine learning (ML) techniques from a model refinement
engine (80).
10 The implementation set forth in the foregoing description do not represent all
implementations consistent with the subject matter described herein. Instead, they
are merely some examples consistent with aspects related to the described subject
matter. Although a few variations have been described in detain above, other
modifications or additions are possible. In particular, further features and/or
15 variations can be provided in addition to those set forth herein. For example, the
implementation described can be directed to various combinations and sub
combinations of the disclosed features and/or combinations and sub combinations
of the several further features disclosed above.
In addition, the logic flows depicted in the accompany figures and/or described
20 herein do not necessarily require the particular order shown, or sequential order,
to achieve desirable results. Other implementations may be within the scope of the
following claims. , Claims:1. A system (100) for determining a disease severity index (DSI) score and
predicting healthcare resource utilization, said system (100) comprising:
a data acquisition module (10) configured to receive one or more
patient-specific input data, including demographic information, 5 comorbid
conditions, laboratory test values, lifestyle factors, and social determinants
of health;
a data pre-processing module (20) configured to clean, normalize,
and validate the received one or more patient-specific input data for further
10 processing;
a scoring engine (30) operatively coupled to the data preprocessing
module (20), and comprising at least one processor (31)
configured to apply a proprietary algorithm on the pre-processed data to
compute a disease severity index (DSI) score, wherein the score being on a
15 numerical scale indicating the predicted healthcare resource consumption
of a patient;
a categorization module (40) operatively coupled to the scoring
engine (30), and configured to classify the computed DSI score into one of
a plurality of risk tiers, including no risk, low risk, moderate risk, rising
20 risk, or high risk;
a recommendation engine (50) configured to generate at least one
care or resource allocation recommendation based on the risk tier assigned
to the patient; and
a user interface module (60) operatively coupled to at least the
25 scoring engine (30), categorization module (40), and recommendation
engine (50), said user interface module (60) configured to display at least
one of the DSI score, risk category, and care recommendations to one or
more stakeholders including healthcare providers, insurers, or care
coordinators.
20
2. The system (100) as claimed in claim 1, wherein the one or more patientspecific
input data is received from a plurality of sources including EMR
systems, patient wearables, insurer databases, or digital health platforms.
3. The system (100) as claimed in claim 1, wherein the data acquisition
module (10) comprises an interface configured to integrate 5 with electronic
medical records (EMR), insurance claims databases, wearable devices, or
mobile health applications.
4. The system (100) as claimed in claim 1, wherein the scoring engine (30)
applies a weighted algorithm comprising condition-specific severity
10 parameters, lab score thresholds, behavioral adherence scores, and regional
disease trends.
5. The system (100) as claimed in claim 1, wherein the categorization
module (40) is configurable with customizable threshold values for each
risk tier based on insurer-specific or provider-specific policies.
15 6. The system (100) as claimed in claim 1, wherein the recommendation
engine (50) provides intervention scheduling, physician review flags, or
digital nudging sequences based on the patient’s risk profile.
7. The system (100) as claimed in claim 1, wherein the user interface module
(60) supports population-level DSI heatmaps, claim forecasting
20 dashboards, and patient-level drill-down for predictive insights.
8. The system (100) as claimed in claim 1, wherein the system (100)
comprising a data storage module (70) configured to store at least one of
historical patient data, DSI scores, and care interventions for longitudinal
analysis.
25 9. The system (100) as claimed in claim 1, wherein the system (100)
comprising a model refinement engine (80) configured to periodically
retrain the scoring engine (30) based on real-world patient outcomes using
machine learning (ML) techniques.
21
10. A method (200) for determining a disease severity index (DSI) score and
predicting healthcare resource utilization using the system (100) as
claimed in claim 1, said method (200) comprising steps of:
receiving one or more patient-specific input data through a data
acquisition 5 module (10);
pre-processing the received one or more patient-specific input data
using a data pre-processing module (20);
computing a disease severity index (DSI) score using a scoring
engine (30) based on a proprietary scoring algorithm;
10 classifying the computed DSI score into one of a plurality of risk
tiers, including no risk, low risk, moderate risk, rising risk, or high risk,
using a categorization module (40);
generating one or more care recommendations, include scheduling
of follow-ups, escalation alerts to physicians, or modification of the care
15 intensity based on risk stratification, using a recommendation engine (50);
and
displaying at least one of a DSI score, risk category, and
recommendations to stakeholders through a user interface module (60).
11. The method (200) as claimed in claim 10, wherein the input data is
20 received from a plurality of sources including EMR systems, patient
wearables, insurer databases, or digital health platforms.
12. The method (200) as claimed in claim 10, wherein the DSI score is
periodically recalculated based on updated lab results, adherence behavior,
or new clinical conditions.
25 13. The method (200) as claimed in claim 10, wherein the method (200)
comprising the step of storing the DSI scores and associated patient
metadata in a data storage module (70) for historical trend analysis.
22
14. The method (200) as claimed in claim 10, wherein the scoring engine (30)
is updated using feedback data and machine learning (ML) techniques
from a model refinement engine (80)