Claims:CLAIMS
WHAT IS CLAIMED IS:
1. A method (300) for determining physiological parameters of a patient through a Machine Learning (ML) model (204), the method (300) comprising:
receiving (302), by a patient monitoring device (102), visual data (216) corresponding to a set of medical devices associated with the patient through at least one image capturing device in real-time, wherein the visual data (216) comprises a unique Quick Response (QR) code associated with each of the set of medical devices;
transforming (304), by the patient monitoring device (102), the visual data (216) into numerical data, wherein the numerical data corresponds to a plurality of pixels in the visual data (216);
determining (306), by the patient monitoring device (102), through the ML model (204), a set of physiological parameters based on the numerical data; and
identifying (308), by the patient monitoring device (102), the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and a Hospital Information System (HIS).

2. The method of claim 1, further comprising:
associating (310) the set of physiological parameters with the patient in an Electronic Medical Record (EMR) based on the identifying;
generating (312) a diagnosis for the patient through the ML model (204) based on one or more of the numerical data, the set of physiological parameters corresponding to the patient, and historical diagnostic data; and
generating (314) a treatment course for the patient through the ML model (204) based on one or more of the numerical data, the set of physiological parameters corresponding to the patient, the diagnosis associated with the patient, and historical treatment data.

3. The method of claim 1, further comprising training the ML model (204) through a transfer learning technique, wherein the transfer learning technique comprises training one or more layers of the ML model (204) trained on a large dataset, with an annotated image dataset.

4. The method of claim 1, wherein the visual data (216) comprises at least one of real-time video data and real-time image data, and wherein the image capturing device is a high-definition (HD) camera.

5. The method of claim 1, wherein the set of medical devices comprises a telemetry monitor.

6. The method of claim 6, wherein the set of physiological parameters comprises at least one of atrial depolarization, ventricular depolarization, ventricular repolarization, rate of atrial cycle, rate of ventricular cycle, oxygen saturation, body temperature, pulse rate, and cardiac output.

7. A system (200) for determining physiological parameters of a patient through a Machine Learning (ML) model (204), the system (200) comprising:
a processor (104); and
a memory (106) communicatively coupled to the processor (104), wherein the memory stores processor instructions, which when executed by the processor (104), cause the processor (104) to:
receive (302) visual data (216) corresponding to a set of medical devices associated with the patient through at least one image capturing device in real-time, wherein the visual data (216) comprises a unique Quick Response (QR) code associated with each of the set of medical devices;
transform (304) the visual data (216) into numerical data, wherein the numerical data corresponds to a plurality of pixels in the visual data (216);
determine (306), through the ML model (204), a set of physiological parameters based on the numerical data; and
identify (308) the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and a Hospital Information System (HIS).

8. The system of claim 7, wherein the processor instructions, on execution, further cause the processor (104) to:
associate (310) the set of physiological parameters with the patient in an Electronic Medical Record (EMR) based on the identifying;
generate (312) a diagnosis for the patient through the ML model (204) based on one or more of the numerical data, the set of physiological parameters corresponding to the patient, and historical diagnostic data; and
generate (314) a treatment course for the patient through the ML model (204) based on one or more of the numerical data, the set of physiological parameters corresponding to the patient, the diagnosis associated with the patient, and historical treatment data.

9. The system of claim 7, wherein the processor instructions, on execution, further cause the processor (104) to train the ML model (204) through a transfer learning technique, wherein the transfer learning technique comprises training one or more layers of the ML model (204) trained on a large dataset, with an annotated image dataset.

10. The system of claim 7, wherein the visual data (216) comprises at least one of real-time video data and real-time image data, and wherein the image capturing device is a high-definition (HD) camera.

, Description:METHOD AND SYSTEM FOR DETERMINING PHYSIOLOGICAL PARAMETERS OF A PATIENT THROUGH MACHINE LEARNING MODEL
DESCRIPTION
Technical Field
[001] This disclosure generally relates to Machine Learning (ML) models and more particularly to method and system for determining physiological parameters of a patient through an ML model.
Background
[001] Globally, healthcare industry is increasingly shifting towards automation to provide efficient management of patients, early detection of abnormalities, and improved decision-making. In present state of art, patient management is either manual or semi-automated. For example, for an Electrocardiogram (ECG) associated with a patient, manual management requires staff members to frequently check on patient health. Such methods are prone to observational errors and inaccurate monitoring. Further, in many cases, such methods fail to provide early detection of abnormalities. In case of semi-automated methods, certain key health parameters such as pulse rate, respiratory rate, and the like, are determined and monitored with predefined threshold values. However, frequent checks by staff members are still required to determine a diagnosis.
[002] Other conventional techniques for patient management include determination of patient health based on analysis of screenshots of videos of monitors of health devices associated with the patient. However, such techniques fail to accurately identify physiological parameters in real-time.
[003] In short, existing techniques fall short in providing a mechanism for monitoring physiological parameters of patients in real-time. Further, existing techniques fail to provide intelligent diagnosis and treatment course based on the physiological parameters of the patient.
SUMMARY
[004] In one embodiment, a method for determining physiological parameters of a patient through a Machine Learning (ML) model is disclosed. In one example, the method includes receiving visual data corresponding to a set of medical devices associated with the patient through at least one image capturing device in real-time. The visual data includes a unique Quick Response (QR) code associated with each of the set of medical devices. The method further includes transforming the visual data into numerical data. The numerical data corresponds to a plurality of pixels in the visual data. The method further includes determining, through the ML model, a set of physiological parameters based on the numerical data. The method further includes identifying the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and a Hospital Information System (HIS).
[005] In one embodiment, a system for determining physiological parameters of a patient through an ML model is disclosed. In one example, the system includes a processor and a computer-readable medium communicatively coupled to the processor. The computer-readable medium stores processor-executable instructions, which, on execution, cause the processor to receive visual data corresponding to a set of medical devices associated with the patient through at least one image capturing device in real-time. The visual data includes a unique QR code associated with each of the set of medical devices. The processor-executable instructions, on execution, further cause the processor to transform the visual data into numerical data. The numerical data corresponds to a plurality of pixels in the visual data. The processor-executable instructions, on execution, further cause the processor to determine, through the ML model, a set of physiological parameters based on the numerical data. The processor-executable instructions, on execution, further cause the processor to identify the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and an HIS.
[006] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.
BRIEF DESCRIPTION OF THE DRAWINGS
[007] The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and, together with the description, serve to explain the disclosed principles.
[008] FIG. 1 is a block diagram of an exemplary system for determining physiological parameters of a patient through a Machine Learning (ML) model, in accordance with some embodiments.
[009] FIG. 2 is a functional block diagram of a patient monitoring device implemented by the exemplary system of FIG. 1, in accordance with some embodiments.
[010] FIG. 3 is a flow diagram of an exemplary process for determining physiological parameters of a patient through an ML model, in accordance with some embodiments.
[011] FIGS. 4A and 4B illustrate an exemplary architecture of an ML model for determining physiological parameters of a patient, in accordance with some embodiments.
DETAILED DESCRIPTION
[012] Exemplary embodiments are described with reference to the accompanying drawings. Wherever convenient, the same reference numbers are used throughout the drawings to refer to the same or like parts. While examples and features of disclosed principles are described herein, modifications, adaptations, and other implementations are possible without departing from the spirit and scope of the disclosed embodiments. It is intended that the following detailed description be considered as exemplary only, with the true scope and spirit being indicated by the following claims.
[013] Referring now to FIG. 1, an exemplary system 100 for determining physiological parameters of a patient through a Machine Learning (ML) model is illustrated, in accordance with some embodiments. The system 100 may include a patient monitoring device 102 (for example, server, desktop, laptop, notebook, netbook, tablet, smartphone, mobile phone, or any other computing device), in accordance with some embodiments. The patient monitoring device 102 may determine physiological parameters of a patient through the ML model from numerical data corresponding to a plurality of pixels in visual data of a patient. It should be noted that, in some embodiments, the patient monitoring device 102 may identify the patient associated with the set of medical devices and the set of physiological parameters based on a unique Quick Response (QR) code associated with the patient and a Hospital Information System (HIS) to generate a diagnosis and a treatment course for the patient.
[014] As will be described in greater detail in conjunction with FIGS. 2 – 4, the patient monitoring device may receive visual data corresponding to a set of medical devices associated with the patient through at least one image capturing device in real-time. The visual data may include a unique QR code associated with each of the set of medical devices. The patient monitoring device may further transform the visual data into numerical data. The numerical data corresponds to a plurality of pixels in the visual data. The patient monitoring device may further determine through the ML model, a set of physiological parameters based on the numerical data. The patient monitoring device may further identify the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and an HIS.
[015] In some embodiments, the patient monitoring device 102 may include one or more processors 104 and a computer-readable medium 106 (for example, a memory). The computer-readable storage medium 106 may store instructions that, when executed by the one or more processors 104, cause the one or more processors 104 to determine physiological parameters of a patient through the ML model, in accordance with aspects of the present disclosure. The computer-readable storage medium 106 may also store various data (for example, visual data (for example, video data or image data), numerical data based on the visual data, set of physiological parameters corresponding to the patient, training data, historical medical data, set of parameters for the ML model, and the like) that may be captured, processed, and/or required by the system 100.
[016] The system 100 may further include a display 108. The system 100 may interact with a user via a user interface 110 accessible via the display 108. The system 100 may also include one or more external devices 112. In some embodiments, the patient monitoring device 102 may interact with the one or more external devices 112 over a communication network 114 for sending or receiving various data. The external devices 112 may include, but may not be limited to, a remote server, a digital device, or another computing system.
[017] Referring now to FIG. 2, a functional block diagram of a patient monitoring device 200 is illustrated, in accordance with some embodiments. In an embodiment, the patient monitoring device 200 is analogous to the patient monitoring device 102. In particular, the patient monitoring device 200 may include a data transformation module 202, an ML model 204, a patient identification module 206, a diagnosis generation module 208, a treatment course generation module 210, a training module 212, and a database 214. In some embodiments, each of the modules 202-214 may be implemented within a memory (such as the computer-readable medium 106). The patient monitoring 200 may receive visual data 216 corresponding to a set of medical devices associated with the patient through at least one image capturing device in real-time. The visual data 216 includes a unique QR code associated with each of the set of medical devices. It may be noted that the visual data 216 includes at least one of real-time video data and real-time image data. In an embodiment, the image capturing device is a high-definition (HD) camera. By way of an example, the set of medical devices may include, but may not be limited to, a telemetry monitor.
[018] Further, the data transformation module 202 receives the visual data 216. The data transformation module 202 transforms a plurality of pixels of the visual data 216 into numerical data. Further, the data transformation module 202 sends the numerical data to the ML model 204. Through the ML model 204, a set of physiological parameters is determined based on the numerical data. By way of an example, the set of physiological parameters includes, but is not limited to, at least one of atrial depolarization, ventricular depolarization, ventricular repolarization, rate of atrial cycle, rate of ventricular cycle, oxygen saturation, body temperature, pulse rate, cardiac output, and the like. The ML model 204 may include one or more layers. In some embodiments, the ML model 204 is based on at least one of a Mask Region-based Convolutional Neural Network (MRCNN) and a logistic regression technique. In some embodiments, the visual data 216 may be analyzed in real-time on edge and on cloud. In such embodiments, skin tone, hands, eye, head, torso, and leg movements associated with the patient may be analyzed to determine physiological state of the patient. Further, the ML model 204 may provide insights on the physiological state of the patient. By way of an example, the skin tone may be used to calculate physiological parameters such as, oxygen saturation, temperature, pulse rate, cardiac output, and the like.
[019] Further, the training module 212 trains the ML model 204 using training data. In an embodiment, the training module 212 receives the training data from the database 214. It may be noted that the training module 212 trains the ML model 204 through a transfer learning technique. The transfer learning technique includes training one or more layers of the ML model trained on a large dataset (for example, a COCO dataset), with an annotated image dataset. The set of physiological parameters may be stored in the database 214. Further, the ML model 204 sends the set of physiological parameters to the patient identification module 206. The patient identification module 206 identifies the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and an HIS. Further, the patient identification module 206 associates the set of physiological parameters with the patient in an Electronic Medical Record (EMR) based on the identification.
[020] Further, the patient identification module 206 sends patient information to the diagnosis generation module 208. The diagnosis generation module 208 generates a diagnosis for the patient through the ML model 204 based on one or more of the numerical data, the set of physiological parameters corresponding to the patient, and historical diagnostic data. The diagnosis may be stored in the database 214. It may be noted that the diagnosis of the patient may be generated as an output 218 of the patient monitoring device 200. Further, the diagnosis generation module 208 sends the diagnosis for the patient to the treatment course generation module 210. The treatment course generation module 210 generates a treatment course for the patient through the ML model based on one or more of the numerical data, the set of physiological parameters corresponding to the patient, the diagnosis associated with the patient, and historical treatment data. The treatment course may be stored in the database 214. It may be noted that the treatment course for the patient may be generated as an output 218 of the patient monitoring device 200. In an embodiment, the output 218 is generated in the form of a report. The report includes the patient information, the set of physiological parameters, the diagnosis, and the treatment course associated with the patient. By way of an example, visual data 216 in the form of images of Electrocardiogram (ECG) tracings received from a cardiac monitor may be analyzed using an OpenCV package. Further, shape and regularity of the ECG tracings may be used for early identification of abnormalities such as atrial fibrillation, supraventricular tachycardia, hyperkalemia, myocardial infarction, and the like.
[021] It should be noted that all such aforementioned modules 202 – 214 may be represented as a single module or a combination of different modules. Further, as will be appreciated by those skilled in the art, each of the modules 202 – 214 may reside, in whole or in parts, on one device or multiple devices in communication with each other. In some embodiments, each of the modules 202 – 214 may be implemented as dedicated hardware circuit comprising custom application-specific integrated circuit (ASIC) or gate arrays, off-the-shelf semiconductors such as logic chips, transistors, or other discrete components. Each of the modules 202 – 214 may also be implemented in a programmable hardware device such as a field programmable gate array (FPGA), programmable array logic, programmable logic device, and so forth. Alternatively, each of the modules 202 – 214 may be implemented in software for execution by various types of processors (e.g., processor 104). An identified module of executable code may, for instance, include one or more physical or logical blocks of computer instructions, which may, for instance, be organized as an object, procedure, function, or other construct. Nevertheless, the executables of an identified module or component need not be physically located together, but may include disparate instructions stored in different locations which, when joined logically together, include the module and achieve the stated purpose of the module. Indeed, a module of executable code could be a single instruction, or many instructions, and may even be distributed over several different code segments, among different applications, and across several memory devices.
[022] As will be appreciated by one skilled in the art, a variety of processes may be employed for determining physiological parameters of a patient through an ML model. For example, the exemplary system 100 and the associated patient monitoring device 102, 200 may determine physiological parameters of the patient through the ML model by the processes discussed herein. In particular, as will be appreciated by those of ordinary skill in the art, control logic and/or automated routines for performing the techniques and steps described herein may be implemented by the system 100 and the associated patient monitoring device 102, 200 either by hardware, software, or combinations of hardware and software. For example, suitable code may be accessed and executed by the one or more processors on the system 100 to perform some or all of the techniques described herein. Similarly, application specific integrated circuits (ASICs) configured to perform some or all of the processes described herein may be included in the one or more processors on the system 100.
[023] Referring to FIG. 3, an exemplary process 300 for determining physiological parameters of a patient through an ML model (for example, the ML model 204) is illustrated via a flow chart, in accordance with some embodiments. The process 300 may be implemented by the patient monitoring device 102 of the system 100. The process 300 includes receiving visual data corresponding to a set of medical devices associated with the patient through at least one image capturing device in real-time, at step 302. The visual data includes a unique QR code associated with each of the set of medical devices. It may be noted that the visual data includes at least one of real-time video data and real-time image data. It may also be noted that the image capturing device is a high-definition (HD) camera. In some embodiments, the set of medical devices includes a telemetry monitor. By way of an example, the data transformation module 202 receives the visual data 216 corresponding to a set of medical devices associated with the patient from an image capturing device.
[024] Further, the process 300 includes transforming the visual data into numerical data, at step 304. The numerical data corresponds to a plurality of pixels in the visual data. Further, the process 300 includes determining through the ML model, a set of physiological parameters based on the numerical data, at step 306. It may be noted that the ML model may be trained through a transfer learning technique. The transfer learning technique includes training one or more layers of the ML model trained on a large dataset (for example, the COCO dataset), with an annotated image dataset. The set of physiological parameters includes at least one of atrial depolarization, ventricular depolarization, ventricular repolarization, rate of atrial cycle, rate of ventricular cycle, oxygen saturation, body temperature, pulse rate, and cardiac output. In continuation of the example above, the data transformation module 202 transforms the visual data 216 into numerical data. The ML model 204 determines a set of physiological parameters based on the numerical data.
[025] Further, the process 300 includes identifying the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and an HIS, at step 308. Further, the process 300 includes associating the set of physiological parameters with the patient in an EMR based on the identifying, at step 310. Further, the process 300 includes generating a diagnosis for the patient through the ML model based on one or more of the numerical data, the set of physiological parameters corresponding to the patient, and historical diagnostic data, at step 312. Further, the process 300 includes generating a treatment course for the patient through the ML model based on one or more of the numerical data, the set of physiological parameters corresponding to the patient, the diagnosis associated with the patient, and historical treatment data, at step 314. In continuation of the example above, the patient identification module 206 identifies the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and the HIS. Further, the patient identification module 206 associates the set of physiological parameters with the patient. Further, the diagnosis generation module 208 generates the diagnosis for the patient based on the set of physiological parameters and historical diagnostic data of the patient. The treatment course generation module 210 generates the treatment course for the patient, the diagnosis, the set of physiological parameters, and the historical diagnostic data of the patient.
[026] Referring to FIGS. 4A and 4B, an exemplary architecture 400 of an ML model for determining physiological parameters of a patient is illustrated, in accordance with some embodiments. The ML model (such as, the ML model 204) includes one or more layers. In the architecture 400, the one or more layers of the ML model include model A 402, model B 404, and model C 406. In an embodiment, the model A 402 is based on an MRCNN architecture. The model A 402 is trained based on labelled images (such as, a labelled image 408) of patient monitoring device 200. The model A 402 receives the labelled image 408 as an input and generates masks and bounding boxes of Regions of Interest (RoI) from the labelled image 408 as output. Further, image parameters 410 such as, but not limited to, width (w’), height (h’), distance of RoI center (x) from a corner (x,y), and the like, are calculated in relation with width (w) and height (h) of the labelled image 408.
[027] Further, the model B 404 receives the image parameters 410. In an embodiment, the model B 404 is based on linear regression. The model B 404 is trained using the image parameters 410. Further, the model B 404 determines a label for the RoI based on the image parameters 410. A set of labels 412 is determined for a set of RoIs. Further, the model C 406 receives the set of labels 412. In some embodiments, the model C 406 is based on one of an Optical Character Recognition (OCR) or a classifier model. It may be noted that crops of the set of RoIs may be rectified using a homography technique prior to receiving by the model C 406. Further, the model C 406 identifies content of each of the set of RoIs and generates a digitized output 414 including a content associated with each of the set of labels 412.
[028] The disclosed methods and systems may be implemented on a conventional or a general-purpose computer system, such as a personal computer (PC) or server computer.
[029] Thus, the disclosed method and system try to overcome the technical problem of determining physiological parameters of a patient through a Machine Learning (ML) model. The method and system provide a high accuracy (discrimination and calibration) solution to determine physiological parameters of a patient from visual data corresponding to a set of medical devices. Image and video data from the set of medical devices are transformed into numerical data for efficient and less resource-intensive computation. The method and system further provide techniques for identifying the patient associated with the set of medical devices. The method and system further generate a diagnosis and a treatment course for the patient based on the set of physiological parameters and historical medical data of the patient.
[030] As will be appreciated by those skilled in the art, the techniques described in the various embodiments discussed above are not routine, or conventional, or well understood in the art. The techniques discussed above provide for determining physiological parameters of a patient through an ML model. The techniques first receive visual data (for example, a unique QR code) corresponding to a set of medical devices associated with the patient through at least one image capturing device in real-time. The techniques may then transform the visual data into numerical data. The numerical data corresponds to a plurality of pixels in the visual data. The techniques may then determine, through the ML model, a set of physiological parameters based on the numerical data. The techniques may then identify the patient associated with the set of medical devices and the set of physiological parameters based on the unique QR code and an HIS.
[031] In light of the above mentioned advantages and the technical advancements provided by the disclosed method and system, the claimed steps as discussed above are not routine, conventional, or well understood in the art, as the claimed steps enable the following solutions to the existing problems in conventional technologies. Further, the claimed steps clearly bring an improvement in the functioning of the device itself as the claimed steps provide a technical solution to a technical problem.
[032] The specification has described method and system for determining physiological parameters of a patient through a Machine Learning (ML) model. The illustrated steps are set out to explain the exemplary embodiments shown, and it should be anticipated that ongoing technological development will change the manner in which particular functions are performed. These examples are presented herein for purposes of illustration, and not limitation. Further, the boundaries of the functional building blocks have been arbitrarily defined herein for the convenience of the description. Alternative boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed. Alternatives (including equivalents, extensions, variations, deviations, etc., of those described herein) will be apparent to persons skilled in the relevant art(s) based on the teachings contained herein. Such alternatives fall within the scope and spirit of the disclosed embodiments.
[033] Furthermore, one or more computer-readable storage media may be utilized in implementing embodiments consistent with the present disclosure. A computer-readable storage medium refers to any type of physical memory on which information or data readable by a processor may be stored. Thus, a computer-readable storage medium may store instructions for execution by one or more processors, including instructions for causing the processor(s) to perform steps or stages consistent with the embodiments described herein. The term “computer-readable medium” should be understood to include tangible items and exclude carrier waves and transient signals, i.e., be non-transitory. Examples include random access memory (RAM), read-only memory (ROM), volatile memory, nonvolatile memory, hard drives, CD ROMs, DVDs, flash drives, disks, and any other known physical storage media.
[034] It is intended that the disclosure and examples be considered as exemplary only, with a true scope and spirit of disclosed embodiments being indicated by the following claims.