DESCRIPTION

Technical Field

[001] One or more embodiments of the disclosure generally relate to ventilators. More particularly, certain embodiments of the disclosure relate to capturing images from a display of a ventilator connected to a patient, analyzing the captured images, converting the data from the captured images to numerical data, and providing suggestive or corrective measures for managing ventilator settings.

Background

[002] The following background information may present examples of specific aspects of the prior art (e.g., without limitation, approaches, facts, or common wisdom) that, while expected to be helpful to further educate the reader as to additional aspects of the prior art, is not to be construed as limiting the invention, or any embodiments thereof, to anything stated or implied therein or inferred thereupon.

[003] Currently various makes of ventilators and associated systems are being used in the industry to monitor health parameters of patients. Most of these ventilators include a display screen that displays the health parameters of the patient. When a health care professional (for example, a doctor or a radiologist) is in physical proximity to the ventilator, the health care professional may interpret the parameters displayed on the display screen or may review a patient report, printed as an output by the ventilator. Accordingly, the health care professional may suggest corrective measures that need to be made to the ventilator settings. Thus, for continuous observation and subsequent management of the ventilator settings, in-person presence of a health care professional becomes essential. This creates a problem for patients located in remote locations (for example, villages in deep interiors of a country), where such ventilators might be available, however, availability of expert health care professionals (for example, intensive care physicians) may be scarce.

[004] There is, therefore, requirement of a method and system to remotely monitor health parameters displayed on the display screen of the ventilator and to constantly provide inputs to a nurse (or a technician) available at the location of the ventilator to manage ventilator settings.

SUMMARY

[005] In one embodiment, a method of analysing data displayed by ventilators is disclosed. The method includes receiving one or more images of a display screen associated with a ventilator. The display screen is configured to display data representation corresponding to one or more health parameters obtained by the ventilator for a patient. The method further includes processing the one or more image, using at least one machine learning algorithm, to extract the data representation. The method includes generating numerical data corresponding to the one or more health parameters, based on the data representation. The method further includes retrieving the one or more health parameters associated with the patient using the numerical data.

[006] In another embodiment, a system for analysing data displayed by ventilators is disclosed. The system includes a processor and a memory communicatively coupled to the processor, wherein the memory stores processor instructions, which when executed by the processor, cause the processor to receive one or more images of a display screen associated with a ventilator. The display screen is configured to display data representation corresponding to one or more health parameters obtained by the ventilator for a patient. The processor instructions further cause the processor to process the one or more images, using at least one machine learning algorithm, to extract the data representation. The processor instructions cause the processor to generate numerical data corresponding to the one or more health parameters, based on the data representation. The processor instructions further cause the processor to retrieve the one or more health parameters associated with the patient using the numerical data.

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 illustrates an exemplary system for analyzing data displayed on ventilators, in accordance with an embodiment.

[009] FIG. 2 illustrates a functional block diagram of an exemplary system for analyzing data displayed on ventilators, in accordance with an embodiment.

[010] FIG. 3A is a flowchart of an exemplary method for analyzing data displayed on ventilators, in accordance with an embodiment.

[01 1] FIG. 3B illustrates process of denoising an image captured from a ventilator, in accordance with an exemplary embodiment.

[012] FIG. 4 is a flowchart of an exemplary method for processing images of a ventilator screen captured by an image sensor to extract data representation of a patient’s health parameters, in accordance with an embodiment. [013] FIG. 5 is a flowchart of an exemplary method for receiving an input for manipulating at least one parameter of a ventilator, in accordance with an embodiment.

DETAILED DESCRIPTION

[014] 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.

[015] Referring to FIG. 1 , an exemplary system 100 for analyzing data displayed on ventilators in accordance with an embodiment is illustrated. The system 100 includes a ventilator 102 configured to measure one or more health parameters of a patient 104, an image capturing device 106 that may include an image sensor, an audio capturing device 108, a computing system 116, a database 122, and a miscellaneous data input interface (not shown in figure).

[016] It may be appreciated by a person skilled in the art that the ventilator 102 is a machine that provides mechanical ventilation by moving breathable air into and out of the lungs of a patient, to deliver breaths to the patient 104 who is unable to breathe voluntarily or is breathing insufficiently. The display screen 110 may display patient monitoring parameters and machine settings. The patient monitoring parameters may include, but are not limited to, a respiratory rate set by the patient, a tidal volume per breath, a delivery flow rate for the breath, a waveform indicating air pressure, oxygen concentration, and trigger rate. It may further be appreciated by a person skilled in the art that though description of the current embodiment is limited to the ventilator 102, the invention is applicable for other health monitoring and maintaining devices that use a display screen to render health parameters of a patient.

[017] The image capturing device 106 may be installed or placed in the same location as the ventilator 102 (for example, a patient ward). Additionally, the image capturing device 106 may be placed, such that, the ventilator 102 is in direct line of sight of the image capturing device 106. Moreover, the image capturing device 106 may be configured to auto-focus on the display screen 1 10. In some embodiments, the image capturing device 106 may be remotely controlled by an administrator, for example, via the computing system 1 16. The image capturing device 106 may be communicatively coupled to the computing system 116 via a communication network, which may be any wired or wireless communication network and the examples may include, but are not limited to the Internet, Wireless Local Area Network (WLAN), Wireless Fidelity (Wi-Fi), Long Term Evolution (LTE), Worldwide Interoperability for Microwave Access (WiMAX), and General Packet Radio Service (GPRS). Thus, the administrator may be able to remotely control and manage positioning, focus, and viewing angle of the image capturing device 106, based on specific requirements.

[018] During a typical operation of the system 100, the image capturing device 106 may capture one or more images 1 12 of the display screen 1 10. The image capturing device 106 may further be configured to capture the one or more images 112 either iteratively after predefined time intervals or in response to predefined activities (for example, change in wave pattern) on the display screen 1 10. After capturing the one or more images 1 12, the image capturing device 106 may transmit the one or more images 1 12 to the computing device 116 via the communication network. In an embodiment, the image capturing device 106 may also be configured to record an uninterrupted video of the display screen 1 10, till the time the ventilator 102 is active. The video may then be split into the one or more images 1 12 at the computing device 1 16.

[019] In some embodiments, the system 100 may also include the audio capturing device 108 which may capture sound emanating from the ventilator 102 (for example, alarm sounds) or from the patient 104 as a sound 1 14. The audio capturing device 108 may be located in the same location as the ventilator 102. The audio capturing device 108 may be integrated with the image capturing device 106. Alternatively, the audio capturing device 108 may be a standalone device and may directly transmit the captured sound 1 14 to the computing device 1 16 via the communication network.

[020] As explained before, the information displayed on the display screen 1 10 of the ventilator 102 may be in the form of waveforms of pressure-time, flow-time and volume-time scalar graphs. Therefore, in one embodiment, the one or more images 1 12 captured by the image capturing device 106 may include the health parameters data represented in the form of graphs on the display screen 110 of the ventilator 102. It will be apparent to a person skilled in the art that the one or more images 1 12 may also include waveforms, numerical data, and colors as displayed on the display screen 1 10. Additionally, data of interest within the one or more images 1 12 may include numbers on the axis of the graphs, titles of the graphs, numbers displayed on the graphs or otherwise, name of the ventilator settings, or any error messages. [021 ] The computing device 116 may then identify a data representation 124 from the one or more images 1 12 and/or the captured voice 114. The miscellaneous data input interface, for example, a keyboard, touch screen, pulse sensor, temperature sensors, blood pressure sensor, may interact with the patient (via a health care professional) to gather further information on the patient. The computing device 1 16 may generate the one or more health parameters associated with the patient in the form of numerical data 126, based on the identified data representation 124.

[022] The one or more images 1 12, the captured voice 1 14, the numerical data 126, and the data collected by the miscellaneous patient interface may be stored in the database 122. The computing device 116 may then analyze the numerical data 126 to provide an analysis to determine a health condition of the patient. The analysis may be used to provide suggestive measures 128 that may be applied by making changes in settings of the ventilator 102. By way of an example, tidal volumes (i.e., size of a patient’s breath) may usually be limited to less than 6 ml/kg ideal bodyweight of the patient for safety of the patient. Thus, it is important to make sure that the tidal volume does not cross this threshold. The system 100 may promptly pick up these abnormal and unsafe values on the ventilator 102 and may promptly alert a health care professional. By way of another example, breathing patterns, such as, double triggering and auto peep, if not timely detected, may cause a lot of patients discomfort and harm. The system 100 may automatically analyze and identify such breathing patterns and may subsequently escalate to the health care professional for actionable input.

[023] The database 122 may include, but may not limited to, a plurality of data servers, and a memory card. It may be appreciated by a person skilled in the art, the database 122 (that includes patients health information in the form of images, voice, and numerical data) may contain virtually any patient data to enable the computing device 1 16 to analyze the data, generate the numerical data 126, and provide the suggestive measures 128 as described hereinabove. In another embodiment, the above discussed information (stored in the database 122) may be stored in a memory card within the computing system 1 16.

[024] In certain embodiments, the database 122 may be included within a storage device which may be part of a portable storage device or an internet based storage. Non-limiting examples of the portable storage device may include computer readable devices, such as, Universal Serial Bus (USB), memory card, DVD, etc. Nonlimiting examples of the internet based storage may include a cloud drive or a data downloadable web link, etc. The captured, analyzed, and generated information on the health parameters of the patient 104 for suggesting changes to the parameter settings of the ventilator may be stored in a local computing platform and/or network. In an alternative embodiment, the captured, analyzed, and generated information on the health parameters of the patient 104 may be located on a local computer network.

[025] The computing system 116 may further connect to any number of devices through the communication network. In one embodiment, the computing system 1 16 may connect to other devices for gathering additional information associated with the patient 104. In one embodiment, the additional information of the patient 104 may include, but is not limited to medical history of the patient 104 and genetic history data of the patient 104. In another embodiment, the computing system 1 16 may connect to other devices for transmitting the numerical data 126 and the suggestive measures 128 in order to make changes to the setting of the ventilator 102. To this end, the computing device 1 16 may include a memory 1 18 to store one or more instructions, wherein the one or more instructions, may be executed by the processor 120. The memory 118 may be a non-volatile memory (e.g., flash memory, Read Only Memory (ROM), Programmable ROM (PROM), Erasable PROM (EPROM), Electrically EPROM (EEPROM) memory, etc.) or a volatile memory (e.g., Dynamic Random Access Memory (DRAM), Static Random-Access memory (SRAM), etc.)

[026] It may be appreciated by a person skilled in the art, that the numerical data 126 generated from the analysis of the captured one or more images 112 and/or the audio 1 14 is stored in the database 122 and may further be employed to train a machine learning model (not shown in FIG. 1 ) within the computing device 1 16 to analyze a variety of images captured from the ventilator 102 and other models of ventilators.

[027] It may be appreciated by a person skilled in the art that the computing system 1 16 may be capable of processing images/voice data and subsequently generating numerical data representing health parameters of patients. Non-limiting examples of the computing system 116 may include a computer, laptop, a phablet, a tablet, an application server, a smart phone, and a cloud computing device. The computing system 116 may be a single device or may be a combination of communicatively coupled multiple devices. In embodiments where the computing device 1 16 is a single device, all the functions of receiving the video, receiving the audio, receiving other information of the patient 104, processing, and formatting gathered information, generating the numerical data 126, analyzing the numerical data 126, and generating the suggestive measures 128 to change the parameter settings of the ventilator 102, may be executed by the single computer system 116. In embodiments where the computing system 116 includes multiple devices these functions may be distributed between the multiple devices.

[028] The miscellaneous data input interface (not shown in FIG. 1 ) may be any data input interface capable of capturing additional information associated with the patient 104. The computing system 1 16 may include tools, for example, text notes and questionnaires which the patient 104 may have answered using the miscellaneous data input interface. The tools may be capable of gathering information on historical health data of the patient 104, genetic health information of the patient 104, preferences of the patient 104, and the like. This information may be employed by a data analyzing application software in the computing system 1 16 while analyzing and generating the numerical data 126 associated with the health parameters of the patient 104.

[029] FIG. 2 illustrates a functional block diagram of an exemplary system 200 for analyzing data displayed on ventilators, in accordance with an embodiment. The system 200 may include a computing system 212 that receives an input 214 from an information source 210 and provides an output 244 that includes recommendations/suggestive measures for modifying settings for the ventilator 102, if deemed necessary. The computing system 212 may include an image capture module 216 (may also be referred to as an image sensor module), an audio capture module 218, a patient information collection module 220, a database module 222, an information processing and formatting module 224, a numerical data generation module 226, a numerical data to graphical data re-generation module 228, an information communication module 230, an information verification module 232, an alert generating module 234, a reset module 236, a display module 238, a security module 240, and a heuristic module 242. [030] The image capture module 216, may have means of capturing an image (and videos), such as, without limitation, the camera 106 that may be used to capture the one or more images 1 12 from the display screen 110 of the ventilator 102. The captured one or more images 1 12 may act as the information source 210. Examples of the image capturing module 106 may include devices, such as, but not limited to, digital cameras, web cameras, video cameras, a Closed Circuit Television (CCTV), or inbuilt camera of a smart device, for example, a personal computer, a smartphone, a tablet, or a laptop.

[031] The audio capture module 218, may have means of capturing sound, such as, without limitation, the audio capturing device 108 that could be used to capture the sound 1 14. In such case, the captured sound 114 may act as the information source 210 for the computing system 212. The audio capturing module 218 may be any device capable of capturing sound of the patient 104 or a health care provider or any sound output from the ventilator 102. The audio capturing module 218 may include devices such as, but is not limited to digital cameras, web cameras, video cameras, voice recorders, etc. In another embodiment, the audio capturing module 218 may be an integrated sound recorder on a personal computer, a smartphone, a tablet, or a laptop.

[032] The patient information collection module 220 may capture patient information. The patient information, for example, without limitation, may be gathered by way of answers provided by the patient 104 or a health care provider using miscellaneous data capturing devices (for example, the audio capturing device 108). Alternatively, the patient information may be gathered tacitly by the audio capturing device 108 placed in close proximity to the patient 104. [033] The database module 222 may store information that includes a patient’s health information in the form of images, audio, and numerical data. Information in the database module 222 may enable one or more modules within the computing system 212 to analyze the information in order to generate numerical data and provide suggestions as described hereinabove. In one embodiment, the database module 222 may be located outside the computing system 212.

[034] The information processing and formatting module 224 may process an image, sound samples, and the gathered information. The information processing and formatting module 224, for example, may include, but is not limited to image recognition algorithms to process images and natural language processing algorithm to process audio samples. The image recognition algorithms may include one or more of, but are not limited to Bayesian networks, fuzzy logic, neural networks, template matching, Hidden Markov models, machine learning, data mining, feature extraction and data analysis/statistics, Natural Language Processing (NLP), Recurring Neural Network (RNN), Decision Trees, Numerical methods for preprocessing (for example, Fourier space analysis (used for curve smoothing), or optical character recognition. This is further explained in detail in conjunction with FIG. 4. The numerical data generation module 226 may process the output of the information processing and formatting module 224 to generate numerical data on the health parameters of the patient 104.

[035] In some optional embodiments, the numerical data to graphical data re-generation module 228 may re-generate graphical data from the numerical data generated on the health parameters of the patient 104, as received from the numerical data generation module 226. This may enable generation of a graph based report for the ventilator 102 at a location that is geographically separated from the ventilator

102.

[036] The information communication module 230 may facilitate exchange and sharing of information amongst various modules of the computing system 212. This information, for example, may include information gathered using the image capture module 216, the audio capture module 218, the patient information capture module 220, the numerical data generated using the numerical data generation module 226, the graphical data regenerated by the numerical data to graphical data re-generation module 228, or alerts generated by the alert generating module 234. The information communication module 230 may also enable exchange of information between the computing system 212 and the ventilator 102. Examples of such information may include reset information provided by the reset module 236 to the ventilator 102. The reset information is further explained in detail in subsequent paragraphs.

[037] The information verification module 232 may verify information that is generated by one or more modules within the computing system 212. The alert generating module 234 may generate alerts for a concerned person, such as, a health care provider, when a change in the health parameters of the patient 104 is detected. The alert may be a visual alert or a voice alert activated via the ventilator 102 or by any other system that is located in close proximity to the ventilator 102 and is communicatively coupled to the computing system 212. In some embodiments, the alert may be directly sent to a device that is accessible by a health care provider. The health care provider may either be present in close proximity to the ventilator 102 or may currently be in a remote location. Examples of the device may include, but are not limited to a smartphone, a tablet, a laptop, a mobile phone, a personal computer, and the like.

[038] In a situation where the ventilator 102 is required to be reset, the reset module 236 may generate a reset command to be sent to the ventilator 102. In some embodiments, the reset command may be used to change or modify settings of the ventilator 102. The display module 238 may display health parameters or alerts to a health care professional. The display module 238, for example, may be a display screen. The security module 240 may encrypt patient information that is captured at location of the ventilator 102 or existing patient information that is being retrieved from the database module 222, in order to secure patient data while in transmission. This ensures that patient data is not misused. The heuristic module 242 may process the data or information provided by the image capture module 216 and the audio capture module 218. Based on the processing, the heuristic module 242 may provide the output 244. The heuristic module 242 may include a machine learning model, which, for example, may be a neural network.

[039] In one embodiment, one or more modules of the computing system 212 may be embodied in a single device. In an alternative embodiment, all modules except the alert generating module 234 and the reset module 236 may be embodied in a personal computer or a laptop. The personal computer or laptop device may be capable of: receiving information on the health parameters of a patient in the form of image/voice information, processing, formatting, and organizing the information, designing and generating numerical data on the health parameters of a patient, and enabling the numerical data to be transmitted to a database (for example, the database module 222) accessible by a health care professional who may then provide suggestions for further treatment of the patient 104. In certain embodiments, the personal computer or laptop device may enable the health care professional to provide suggestions to change the settings of the ventilator 102.

[040] FIG. 3A is a flowchart illustrating an exemplary method 300 for analyzing data displayed on ventilators, in accordance with an embodiment. At step 302, one or more images (for example, the one or more captured images 1 12) of a display screen (for example, the display screen 1 10) associated with a ventilator (for example, the ventilator 102) is captured using an image sensor (for example, the image capturing device 106). The display screen is configured to display data representation corresponding to one or more health parameters obtained by the ventilator for a patient. Examples of the image sensor may include, but are not limited to a CCTV camera, a web camera, or an inbuilt camera of a smart device (for example, a smartphone).

[041] At step 304, the one or more images of the display screen associated with the ventilator may be received. The one or more images may be received by the computing system 1 16, for example. At step 306, the one or more images may be processed using at least one machine learning algorithm to extract the data representation from the one or more images of the display screen. This is further explained in detail in conjunction with FIG. 4. Based on the data representation extracted at step 306, numerical data (for example, the numerical data 126) corresponding to the one or more health parameters is generated at step 308. The numerical data, for example, may be in a digitized form. In an embodiment, if the numerical data is not accurately generated or if there appears to be an error in the determination of the numerical data, the computing system 1 16 may instruct the image capturing device 106 to re-capture images of the display screen of the ventilator. This may be iteratively repeated till the captured and processed images are sufficient to generate an appropriate numerical data.

[042] At step 310, the numerical data may be used to retrieve one or more health parameters associated with the patient. In an embodiment, once the one or more health parameters are retrieved, these may be simultaneously displayed to a health care professional sitting at a distant location (geographically separated from the location of the ventilator), along with the one or more captured images. In order to retrieve the one or more health parameters, the numerical data retrieved at step 310 is de-noised and a continuous curve is generated at step 310a. In other words, clean data is generated from the numerical data after noise is removed.

[043] An exemplary embodiment depicting the process of denoising a captured image is illustrated in FIG. 3B. The captured image may include a ventilator graph and may be noise-contaminated. Therefore, denoising the captured image enables estimation of the original image by suppressing the noise. Image noise may be a result of intrinsic conditions (for example, sensor based issues) and/or extrinsic conditions (for example, ambient environment based issues) that are unavoidable in practical situations. In the current context, the captured images (noise-contaminated) are processed to extract relevant graphs from captured images. The denoising technique of the exemplary embodiment is depicted by 316, 318, and 320.

[044] At 316, a captured image (which is noise contaminated) is straightened to generate a straightened image. Thereafter, the straightened image is converted into grayscale and edges are detected to generate edge images as depicted in 318. At 318, perspective transformation is performed on the edge images and noise is eliminated thereafter. Since most of the noise is eliminated after 318, pixels for the ventilator graph may be easily detected, as depicted in 320. [045] In an embodiment, the one or more health parameters associated with the patient may be based on one or more attributes of the respiratory system. The one or more attributes may include, but are not limited to an airway pressure, an elastance of the respiratory system, a resistance of the respiratory system, and a tidal volume. The one or more health parameters associated with the patient may include a mode of breath and an abnormality associated with breathing pattern. Mode of breadth may be determined based on an analysis of the numerical data. In an exemplary embodiment, abnormalities in breathing pattern may be determined as depicted below:

Step 1 : Noise is removed from digitized data (numerical data) using cubic spline interpolation and interpolation steps are performed in order to get a continuous curve (or cleaned data).

Step 2: Apply equation 1 (ventilator equation) given below to the cleaned data: P = EV + RF ... (1 ) where, P = Airway pressure

E = Elastance of respiratory system

R = Resistance of respiratory system

V = Tidal volume

F = dV/dt, i.e., change in the Tidal volume with respect to time Step 3: Differentiate the equation 1 to obtain equation 2 as given below: P’ = EF + RF’ ... (2)

[046] It may be noted that if volume graph is present there may not be a requirement to differentiate. However, when volume graph is not present differentiation may be used as depicted in equation 2. The equation 2 is used to estimate values of ‘E’ and ‘R’ and a ratio of R/E is determined by fitting parameters using gradient descent. Further, units for pressure and volume are restored (or retrieved) by performing OCR on PEEP, Tidal Volume, and/or Respiratory rate.

[047] At step 312, a plurality of recommendations (for example, the suggestive measures 128) may be generated based on the determined one or more health parameters. In an embodiment, a database (for example, the database 122) may include mapping of data interpretation to suggestive measures (recommendations). The data interpretation may be the determined one or more health parameters. Alternatively, data interpretation may be the numerical data after being de-noised. In an embodiment, if a specific data interpretation is not mapped to a suggestive measure in the database, a remotely located health care professional may be prompted to provide a recommendation or a suggestive measure. This recommendation or suggestive measure may then be mapped to the specific data interpretation in the database. In another embodiment, the incremental mapping in the database may be used to train a machine learning algorithm to predict one or more recommendations (or suggestive measures) based on the data interpretation. At step 314, the plurality of recommendations may be transmitted to a health care provider (nurse or a technician) to enable him/her to change settings of the ventilator and follow any other suggested mode of treatment required for the patient.

[048] FIG. 4 is a flowchart of an exemplary method 400 for processing images of a ventilator screen captured by an image sensor to extract data representation of a patient’s health parameters, in accordance with an embodiment. After one or more images (for example, the one or more images 112) of the display of the ventilator are captured, the one or more images are processed to extract data representation of a patient’s health parameter from the one or more images. To this end, at step 402, instance segmentation is performed on the one or more captured images. For instance segmentation, Mask Region based Convolutional Neural Network (RCNN) may be used on the one or more captured images. In one embodiment, the step 404 may also include identification and subsequent cropping of relevant objects in the one or more captured images using an object detection technique. RetinaNet from ImageAl may be used to facilitate object detection from the one or more captured images. In another embodiment, an object detection model may be created using transfer learning on RetinaNet.

[049] After instance segmentation, at step 404, image straightening of the one or more captured images may be performed. Image straightening may be performed using at least one of a canny edge detection, a Hough line transform, and a perspective warping on the one or more captured images. The steps 402 and 404 corresponds to pre-processing of the one or more captured images.

[050] After pre-processing, the one or more captured images may be digitized. To this end, at step 406, object detection may be performed on the one or more captured images to detect at least one of, but not limited to a pressure graph, a flow graph, and a volume graph. Object detection may lead to digitization of data contained in the one or more captured images. As discussed above, the one or more captured images may include different types of graphs, for example, a pressure graph, a flow graph, and a volume graph. The technique for object detection or digitization of each of these graphs may be unique.

[051] The digitization process may use dependencies such as OpenCV C++ and GNU scientific libraries. To this end, in an exemplary embodiment, a bounding box algorithm may be implemented to identify parts of an image that include one or more of a pressure graph, a volume graph, or a flow graphs. Bounding boxes of different graphs may be identified by pixel color filtering in Red Green Blue (RGB) color space and based on contour algorithms. Once a bounding box is identified, coordinates of the bounding box may be set. For the Y axis, at base of the bounding box (which is a square), Y may be set to zero and vertical size of the bounding box may be set as one unit length. In a similar manner, the X axis of the bounding box may also be set. Additionally or alternatively, in another exemplary embodiment, machine learning (that may use transfer learning) based bounding box identification process may be used.

[052] In an embodiment, machine learning algorithms may be used to identify specific parts of an input image and may provide one or more of bounding boxes and masks as an output. In an embodiment, these identified sub-parts of the input image (original image) may be identified as specific objects in a subsequent step. In this subsequent step, parameters may be identified from the identified sub-parts. These parameters may include, but are not limited to at least one of bounding box coordinates, crop image height, crop image width, or crop image pixel ratio are input to separate models. These models, for example, may include, but are not limited to Logistic regression models or Decision tree models. It may be noted that while bounding boxes may include pixels surrounding an area of interest, masks may include all the pixels in the area of interest. Bounding boxes of graphs may be determined from four rectangular coordinates of a mask.After setting coordinates of the bounding box, the points on the graph curves enclosed within the bounding box may be determined based on predefined criterion for different types of graphs. In an embodiment, at step 406a, a classification model may be trained to perform object detection, using transfer learning on RetinaNet. In other words, the classification model may be used to perform object detection (or digitization) for different types of graphs, i.e., pressure graphs, flow graphs, volume graphs, and the like.

[053] FIG. 5 is a flowchart of an exemplary method 500 for manipulating at least one parameter associated with a ventilator, in accordance with an embodiment. At step 502, one or more health parameters associated with a patient are retrieved using numerical data. This has already been explained in detail in conjunction with FIG. 3 and FIG. 4. At step 504, at least one of the one or more images and the one or more health parameters may be simultaneously displayed to a health care professional who is located at a place that is geographical separated from location of the ventilator. The simultaneous display may be via the computing system 116 depicted in the FIG. 1 . For example, the health care professional may be an intensive care specialist located in New York, while the ventilator may be supporting a patient located in a third world country that does not has expert doctors.

[054] Based on the at least one of the one or more images and the one or more health parameters, at step 506, an input for manipulating at least one parameter associated with the ventilator may be received by a staff who is managing or operating the ventilator. The input may include a text note or an audio recording created by the health care professional who is remotely located. The input, for example, may be transmitted from the computing system 1 16 to a staff or any other health care professional who is near the ventilator (for example, the ventilator 102). The input may be with respect to change in settings of the ventilator or any other procedure that is required to be performed in order to ensure that the patient’s health is not impacted.

[055] Having fully described at least one embodiment of the present invention, other equivalent or alternative methods of implementing the design for analyzing data displayed on ventilators using designated computer programs according to the present invention will be apparent to those skilled in the art. Various aspects of the invention have been described above by way of illustration, and the specific embodiments disclosed are not intended to limit the invention to the particular forms disclosed. The particular implementation for analyzing data displayed on ventilators using designated computer programs may vary depending upon the particular context or application. By way of example, and not limitation, the system for analyzing data displayed on ventilators using designated computer programs described in the foregoing were principally directed to capturing image from a display screen of a ventilator device connected to a patient and generating numerical data for health parameters of a patient from the captured image using a designated computer program, and using the numerical data to make suggestions on changes in settings of the ventilator machine. However, similar techniques may instead be applied to any health monitoring device, or financial monitoring device (stock exchange data), or sports data monitoring device, which implementations of the present invention are contemplated as within the scope of the present invention. The invention is thus to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the following claims. It is to be further understood that not all of the disclosed embodiments in the foregoing specification will necessarily satisfy or achieve each of the objects, advantages, or improvements described in the foregoing specification.

[056] 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 receiving one or more images (or snapshots) of a display screen associated with a ventilator, such that the display screen displays data representation corresponding to one or more health parameters obtained by the ventilator for a patient. The ventilator may be located in a location that is in a remote area and the patient may not have access to an intensive care physician. The one or more captured images are then transmitted over a communication network to a location where intensive care physicians are available.

[057] The technique further includes processing the one or more captured images at the different location, using at least one machine learning algorithm, to extract the data representation. The data representation is then used to generate numerical data corresponding to the one or more health parameters. Thereafter, the one or more health parameters associated with the patient are retrieved using the numerical data and a plurality of recommendations are generated based on the determined one or more health parameters. The plurality of recommendations are then transmitted back to the location of the ventilator via the communication network. This enables providing constant inputs to a nurse (or a technician) available at the location of the ventilator to manage ventilator settings. Thus, the problem of absence of an intensive care physician at the location of the ventilator is resolved.

[058] 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.

[059] The specification has described method and system for analysing data displayed on ventilators. 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.

[060] 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.

[061] It is intended that the disclosure and examples be considered as exemplary only, with a true scope of disclosed embodiments being indicated by the following claims.