Diagnosis and monitoring treatment effectiveness for cleansing activities and purging behaviors in eating disorders

Abstract

A method and a system are provided for taking physiological markers measurements of patients who have an eating disorder (e.g., anorexia nervosa, bulimia nervosa, binge eating disorder, pica syndrome, rumination disorder or refrained food consumption). Mathematical analysis of the physiological markers measurements is used to create a unique personal prediction model and data set for an individual patient, including personal predictions for the entire stages of the disorder (binge, nausea and cleansing). The unique personal data set is used to diagnose and monitor a particular problem of the individual patient associated with the eating disorder, determine the disorder's severity, provide intraday alerts to assist in the ongoing disorder management or to recommend a treatment for a particular problem of the individual patient associated with these eating disorders, or to predict an outcome of a treatment for a particular problem of the individual patient associated with these eating disorders.

Description

DIAGNOSIS AND MONITORING TREATMENT EFFECTIVENESS FOR CLEANSINGACTIVITIES AND PURGING BEHAVIORS IN EATING DISORDERSFIELD OF THE INVENTION

[0001] The present invention generally relates to a method for analysis of eating disorders and in particular, but not limited to, measuring, predicting and alerting during the different stages of the disorder including binge, nausea and cleansing activities and purging behaviors in eating disorders and the effectiveness of treatments, including, but not limited to, behavioral, psychological treatments, drugs, appropriate dosage, and especially a combination of different types of treatments and / or drugs, used to neutralize eating disorders or other related clinical treatments.BACKGROUND OF THE INVENTION

[0002] Eating disorders is an umbrella term for a group of disorders that manifests itself in an excessive preoccupation with eating and weight, along with problematic eating patterns that lead to emotional difficulties, health problems and significant impairment of function. Eating disorders, unlike other mental health disorders, may have significant physical and life-threatening consequences. In various cases, the eating disorders are accompanied by cleansing / purging activities.

[0003] The present invention refers to eating disorders as defined in the Diagnostic and Statistical Manual of Mental Disorders (DSM), for example:

[0004] Anorexia Nervosa - an eating disorder characterized by a drastic decrease in body weight, which occurs through starvation and / or excessive sports activity. The disorder is accompanied by a huge fear of obesity and an excessive preoccupation with food.

[0005] Bulimia nervosa - an eating disorder characterized by binge eating accompanied by “compensatory behaviors” whose purpose is to prevent weight gain. These behaviors can be vomiting, fasting, using laxatives, etc.

[0006] Binge eating disorder (BED) - an eating disorder characterized by repeated eating bouts in which an excessive amount of food is consumed accompanied by a feeling of inability to control eating, guilt and self-disgust. Unlike bulimia, in this disorder there are no compensatory behaviors.

[0007] Pica Syndrome - an eating disorder characterized by eating non-nutritious substances, which are not food, over a long period of time.

[0008] Rumination Disorder - Returning swallowed food back into the oral cavity, for the purpose of chewing, swallowing or spitting, and not as a result of a medical condition or as part of another eating disorder.

[0009] Refrained food consumption - an eating or feeding disorder characterized by avoiding or limiting food intake, without disturbance in the way weight and body shape are experienced, to the point of drastic weight loss, significant nutritional deficiency, dependence on feeding or nutritional supplements to the point of dysfunction.

[0010] Eating disorders can lead to severe physiological and health problems and even death, as it is one of the most deadly mental disorders. Underweight conditions are lifethreatening and in most cases lead to physiological damage. Even when the weight itself is not dangerously low, symptoms such as binge eating, self-starvation and vomiting may harm various physical functions such as the menstrual cycle regularity, cardiac issue, esophageal function and more. In addition, extreme obesity may lead to diabetes, heart disease, high blood pressure, fertility difficulties and more.

[0011] In addition to the health problems, eating disorders often lead to functioning impairment. Eating disorders in children and adolescence often damages the parents functioning as well as the entire family. Moreover, eating disorders often lead to social and interpersonal problems, such as hiding the condition, particularly low body image and self-esteem. As such, eating disorders may lead to the neglect of significant areas of life that are not related to eating, and thereby to significant harm and distress to those who suffer from it.

[0012] Cleansing activities, often referred to as purging behaviors, are common in the different types of eating disorders, typically associated with attempts to compensate for overeating or to prevent weight gain. The cleansing activities are typically associated with the following activities:

[0013] Self-Induced Vomiting: This involves intentionally bringing up food that has been consumed by inducing vomiting. Individuals may use their fingers, objects, or other methods to trigger vomiting.

[0014] Laxative Abuse: Some people misuse laxatives, which are substances that help increase bowel movements. Abusing laxatives is an attempt to eliminate calories before they can be absorbed by the body.

[0015] Diuretic Abuse: Diuretics are medications that increase urine production and are often used to treat conditions like edema. However, in the context of eating disorders, individuals may misuse diuretics to try to reduce water weight.

[0016] Excessive Exercise: Intense and frequent exercise is another way individuals may attempt to compensate for overeating or prevent weight gain. This can involve compulsive or driven exercise, even in the absence of excessive food consumption.

[0017] Fasting or Extreme Caloric Restriction: Some individuals may engage in periods of fasting or extreme caloric restriction as a way to offset the effects of overeating or to achieve weight loss.

[0018] Enemas: Enemas involve introducing liquid into the rectum to induce bowel movements. Like laxatives, enemas are sometimes misused as a means of purging.

[0019] It is important to note that purging behaviors can have serious health consequences, including electrolyte imbalances, dehydration, gastrointestinal issues, and damage to the esophagus and teeth. They are also indicative of underlying emotional and psychological distress.

[0020] Current treatments

[0021] The physical and psychological problems described above, as well as mortality danger, emphasize the importance of finding adequate treatment. According to current approaches, treatment of eating disorders first includes a nutritional-medical plan focused on achieving and maintaining a normal weight, guided by a physician and a nutritionist or clinical dietitian. In addition, psychological and psychiatric interventions are needed.

[0022] There are a variety of common treatment methods for eating disorders that may help in dealing with the disorder, among others are dynamic treatments (individual and / or group) which help in identifying the role of the disorder in the patient’s life and regaining control over the patient’s body, cognitive behavioral treatments designed to deal with obsessive thinking and the behaviors that accompany it (e.g., excessive occupation in sports) and family interventions, which are focused on a more adaptive resolution of conflicts. Pharma treatments may assist in cases were the disorder results from or is accompanied by other emotional and mental difficulties such as depression or anxiety. However, the pharma treatment does not treat the eating disorder directly.

[0023] The treatment of an eating disorder may be carried out in a clinical setting, when the patient is able to continue the daily routine. Together with that, in cases where the patient’s life is at risk or the compliance to the treatment is low, treatment is provided at the clinic. These cases may include comprehensive medical supervision, strict supervision of proper nutrition, and a mental treatment framework prevent deterioration.

[0024] In addition to the lack of objective tools to assist the patient on a daily level, there is a severe problem of a lack of objective tools and measures to quantify the severity of the problem by the physicians, both in the initial screening phase for prioritizing patients according to the severity of the condition, and during the ongoing treatment.

[0025] Due to the lack of data, one of the major problems is the lack of objective information for diagnosing and determining severity (for example - the frequency of vomiting). Due to this, there are currently no objective tools to quantify the severity of the disorder in the patients. As a result, there is no possibility to prioritize the patients for processes such as hospitalization.

[0026] In addition, the treatment processes typically aim to increase the self-awareness, so that the patient can batter prepare and / or ask for assistance before the occurrence of such activity. As of now, there are no objective tools that can provide prediction to the occurrence of such behaviors.SUMMARY

[0027] The present invention seeks to provide objective analysis of physiology by developing a method and machine learning based engine for monitoring and personal condition evaluation and providing treatment related alerts. The system will analyze changes in various physiological markers in order to make assist the patient in identifying patterns and situations (for example, binge occurrence followed by nausea and then followed by cleansing activities), and will personalize the treatment analysis to determine personalized treatment per patient.

[0028] Without limitation, the system may include the following:

[0029] 1. Identification of the binge and the following nausea and cleansing conditions by a machine-learning based engine to predict changes in a variety of physiological indicators to identify the above conditions. The collection of the physiological indicatorswill be carried out using wearable products such as smart watches, smart stickers and more.

[0030] 2. Initial screening: quantifying the disorder severity as objectively documented by measuring the frequency disorder occurrence (e.g., vomiting, binge eating, etc.).

[0031] 3. Ongoing treatment: real-time alert before an appearance of binge, nausea and cleansing events.

[0032] As part of the behavioral therapy, the system will provide an alert which will allow the patient and to better implementation of the behavioral protocol prior to the disorder occurrence.

[0033] 4. Ongoing treatment: trends analysis

[0034] By providing objective and ongoing analysis, the invention will analyze long-term trends regarding the severity of the disease by objectively quantify cases (e.g., . the desire to vomit).

[0035] 5. Personalized treatment by assigning patients to clusters, based on their background data (such as gender, age and ethnicity) and physiological responses.

[0036] The present invention seeks to provide a method and a tracking and evaluation algorithmic system that analyzes changes in various physiological markers, for eating disorders diagnosis, continuous monitoring (including real-time alerts for stakeholders to enable prevention measures), analyzing the personalized severity of the disorder and providing real time continuous objective data to the different stakeholders (patients / parents, physicians) to enable better management of the disorder.

[0037] The method and system may include A) allocating the patient to one of predetermined groups of eating disorders clusters (using both severity and specific patient attributes); B) personalized analysis of their disorder severity; C) provide real-time data to alert stakeholders for prevention measures; and D) collection and presentation of objective data to allocate long term patterns.

[0038] Unlike the current trial and error mitigation processes that use subjective inputs and self-reporting, in the present invention these processes are done using objective mathematical analysis of changes in an ensemble of different physiological markers. The mathematical analysis is done using trained machine learning based algorithms,designated specifically for eating disorders analysis and then personally calibrated per group of patients (e.g., gender, aged, comorbidities).

[0039] As opposed to the prior art, the physiological markers measurements are used to objectively define the disorder severity, probability of the treatment success in lowering the eating disorder’s symptoms and the ongoing treatment effectiveness and patterns, all of these by analyzing an ensemble of physiological markers.

[0040] The process will include A) defining pre-determined personal groups of patients; and B) create a personal profile, baseline and prediction pattern per patient to analyze the patient’s specific attributes and current disorder severity. The patients in each cluster have similar attributes (e.g., gender, age, comorbidities) and physiological markers measurements that are similar, so that within each cluster marker variably will be significantly smaller than the between clusters variability.

[0041] The innovative process will include:

[0042] A. Allocation to a patient cluster: providing patient cluster based analysis of each specific treatment success with respect to combination of treatment regimens;

[0043] B. Objectively analyzing the current personal disorder severity (e.g., number of vomiting cases);

[0044] C. Real time alerts (e.g., prior to cleansing activity) using changes in physiological markers to predict such episode; and

[0045] D. Ongoing treatment effectiveness analysis: by calibrating the generic model assigned to each cluster to create unique treatment effectiveness personal model. This will be done by using machine and deep learning techniques.

[0046] There is thus provided in accordance with a non-limiting embodiment of the invention a method for dealing with an eating disorder comprising taking physiological markers measurements of patients who each have an eating disorder; using the physiological markers measurements to create a cluster of eating disorder patients, wherein the patients in the cluster have similar attributes and physiological markers measurements, so that within each cluster physiological markers variably will be significantly smaller than the between clusters variability; and processing differences between the physiological marker measurements of the patients, both between clusters and specifically for each patient while in normal stage or during a disorder related eventto create a unique personal data set for an individual patient; and using the unique personal prediction machine learning based models and data set to A) set a personal disorder scale; B) set personal severity disorder scoring; C) provide intraday event alerts; and D) provide ongoing analysis to detect treatment effectiveness patterns.BRIEF DESCRIPTION OF DRAWINGS

[0047] The present invention will be understood and appreciated more fully from the following detailed description, taken in conjunction with the drawings in which:

[0048] Fig. 1 is an illustration of the data flow concept: retrieving physiological data (smart wearable agnostic), analyzing the data with our machine learning based engine and assigning various analysis reports to the different stakeholders.

[0049] Fig. 2 is an illustration of a phone app that will present peaks in physiological activity that will indicate a prediction of disorder event (e.g., vomit).

[0050] Fig. 3 is an illustration of the significance of physiological change under different stages of eating disorders and different scenarios (e.g., binge eating disorder, followed by nausea and self-induced vomiting episodes).

[0051] Fig. 4 is a sample table of overall disorder severity.DETAILED DESCRIPTION

[0052] Reference is now made to Fig. 1. The illustrates, in accordance with a nonlimiting embodiment of the present invention, retrieving physiological data (smart wearable agnostic), analyzing the data with our machine learning based engine and assigning various analysis reports to the different stakeholders.

[0053] Personally mapping events and aggregating the overall changes in the ensemble of physiological markers are illustrated in Fig. 2 and Fig. 3.

[0054] Once the machine learning personal baseline is set, the system aggregates data to set the personal severity of the disorder scoring, as shown in Fig. 4.

[0055] The invention provides continuous analysis to track intraday events and overall changes patterns in the disorder severity.

[0056] As mentioned above, in contrast to the prior art, the present invention uses mathematical analysis of physiological measurements to define clusters of patients and, for each patient, to create a personal objective severity baseline, intraday real time alerts and ongoing patterns of the treatment effectiveness. The unique personal models arecalibrated by machine and deep learning techniques which are specifically assigned per patient, for the provision of output that assess eating disorder severity and patterns.

[0057] The personal method may include the followings steps:

[0058] A: Cluster allocation:

[0059] Using patients’ measured physiological marker data to build an ensemble of predictive models for clinical eating disorder diagnosis tools used by physicians. This step uses physiological marker data measured during the clinical diagnosis process, and would use a host of pattern recognition, machine learning and Al algorithms.

[0060] Using pattern recognition, machine learning and Al algorithms to cluster patients based on external features (such as gender, age, comorbidities, ethnical origin) and internal features (physiological data), and to modify the predictive model to each cluster.

[0061] B: Personal baseline calibration and continues monitoring of medical treatment effectiveness:

[0062] Using pattern recognition, machine learning and Al algorithms to calibrate model features (e.g., physiological markers weighting and / or selecting the importance of a particular physiological feature and / or modifying the personalized detection algorithms) and tailor a predictive model to each individual, based on its cluster association and on its unique combination of external and internal features.

[0063] Using the trained personalized predictive model to continuously and passively monitor patients to detect patterns and objectively assign severity score and continuously detect treatment ongoing treatment patterns.

[0064] C: Real time alerts:

[0065] Once the models will detect an anomaly (e.g., binge, nausea and cleansing episodes) they will alert the relevant stakeholders to start real time prevention protocols

[0066] The method also assesses the objective impact of other treatment options for eating disorders, such as various cognitive and behavioral treatments.

[0067] The output of the method is achieved by demonstrating personal machine learning based analysis and assessment models (for both all 5 eating disorder subtypes) and personal baseline of the individual with and without effective treatment. This output supports physicians’ evaluation of treatment success and provides initials objectivescreening tool and additional, ongoing analysis which recommends possible advantageous treatment modifications.

[0068] The method uses pattern recognition, machine learning, Al algorithms, and other techniques. It is based on physiological marker measurements (described below) and external information (such as gender, age, etc.). A predictive model is designed upon collection of all samples, then modified within each cluster of patients, and then further modified to match the specific personal pattern of every patient, with separated personal models for initial objective baseline creation and ongoing analysis. Eventually, every individual patient is characterized by a unique, personalized, predictive model.

[0069] The following is a description of one non-limiting embodiment of the method and system of the invention.

[0070] First, a combination of the following of biomarkers’ measurements may be automatically collected and recorded by a device, such as a wearable device

[0071]

[0072] Afterwards, patients may be assigned to a pre-defined cluster using the physiological markers and attributes. For example, initial ongoing physiological markers measurement, together with patient’s attributes (e.g., female, 15 years old, Afro- American, which was born with a low birth weight) may be used to assign a patient into a designated cluster. The method may start with a certain number of clusters, and in parallel there may be ongoing cluster calibration and addition of new clusters.

[0073] The methods for clustering and for building predictive models may be based on an ensemble of models, including linear models (e.g., Fisher Discriminant Analysis and Linear Discriminant analysis) and nonlinear ones (e.g., neural network classifiers, random forests). The ensemble of models may be adjusted and constantly re-trained whenever more patient data become available.

[0074] The method may also include measuring physiological markers in parallel to physician diagnostics. For example, the measurement may be carried out together with the clinical evaluation to associate physiological markers with eating disorder severity scoring as determined by the physician.

[0075] The method may also include calculation of a personal calibration model. The method may assign a personal calculation model to setup baseline and ongoing prediction models.

[0076] The method may also include setting up personalized physiological markers weightings and significance level. For example, these may include personal physiological markers weighting and / or selecting the importance of a particular feature while assessing non-linear models.

[0077] The method may also include both monitoring ongoing treatment and / or preliminary eating disorder diagnosis and baseline settings. This may include utilizingpersonal prediction patterns and performing ongoing analysis and real time prediction and alerts, using machine learning based process.

[0078] The method may also include ongoing cluster calibration and shifts between clusters. For example, there may be ongoing evaluation of the patient to decide whether he / she should shift between clusters (e.g., age group changed from 14-10 to 18-14), or there may be ongoing creation of new clusters or sub-clusters.

[0079] Personal process - samples demonstrating the significance of personal modeling:

[0080] The importance of personal modeling is now explained with reference to Fig. 3. The figure relates to measurements of various eating disorder events. During the initial calibration process, the system will record specific event, such as different cleansing activities

[0081] The system will label the personal overall changes in the raw of physiological markers (e.g., changes in HRV), and will further use mathematical models to further analyze the overall physiological change (e.g., changes in all or the markers, their standard deviation, their correlations and further advanced measures that can be extracted using machine learning based methods).

[0082] Once the machine learning based personal calibration will be set, the system will initially gather events over a period of time to create a personal disorder severity score, for the initial screening by the professional stakeholders.

[0083] After that, the system will continuously both monitor the patient and adjust the personal model to verify major changes (e.g., patient shift between different patient’s clusters).

[0084] Technical Process:

[0085] All of the physiological markers’ measurements are automatically collected and recorded by a wearable device. Data may be transformed to a centralized hub. Data may be stored within the wearable device memory for later downloading. Thereafter, the data may be transmitted or uploaded to a digital data repository (cloud or other) and analyzed by implementing machine and deep learning techniques.

[0086] System output is a translation of the processed analysis into medical indicators and status reports and is provided to physicians through a medical web application, and tothe individuals (patient, child, parents etc.) through a mobile application to be used for both real time alerts and aggregated data.

[0087] Analysis is seamlessly transferred to the individual’s smartphone (or to other communication device), or via a transmitting, add-on dedicated unit for real-time data transmission.

[0088] The web application provides the physicians the results of the analysis which include continuous diagnosis and monitoring of the disorder, which may include indication of severity scoring, treatment effective analysis long term patterns and trends and other clinical aspects (e.g., effectiveness of other clinical treatments), predictions related to treatment changes, adherence, false reporting and overall treatment effectiveness snapshots. Thus, the method provides medical professionals big data insights into real-life patient clinical and behavioral patterns.

[0089] In addition, the method may use patient clustering analysis to provide deep learning-based recommendations (e.g., preliminary type of recommended treatment to be allocated to each cluster for the initial treatment process).

[0090] System’s elements / components:

[0091] The system implementing the method is an loT platform which may include a user-friendly wearable device. The wearable device may include, without limitation, a sensor hub that include sensors for all biomarkers mentioned in the above table, a user interface (e.g., including graphics, sound and vibration methods of interface), and a software application. For example, without limitation, the application may operate the related services and processes to read the sensors, perform the analysis and send the real time feedback through the user interface to the user. This software application may also communicate with the cloud-based software on a real time or a periodic basis, to update the data measured from the sensors in the device.

[0092] The system may further include a cloud-based software tool or other data repository, which aggregates and analyzes the sensors data, updates related algorithms and generates output in dash boards and reports to stakeholders.

[0093] The system may further include a web application interface for medical professionals - providing output on diagnosis of eating disorder and treatmenteffectiveness, predictions of recommended treatment changes, level of disorder severity, adherence, false reporting and overall treatment effectiveness snapshots.

[0094] The system may further include a mobile application interface for users (individuals who have eating disorders and stakeholders (e.g. parents) providing a real time event alerts throughout the day, including specific event (e.g. vomit).

[0095] Method outputs and example of the importance of personal analysis:

[0096] Tests of the methods of the invention have demonstrated their efficacy, resulting in outputs objectively quantifying the disorder severity, while avoiding false alerts, in accordance to clinically grade indications by physicians.

Claims

CLAIMSWhat is claimed is:

1. A method for dealing with an eating disorder comprising: taking physiological markers measurements of patients who each have an eating disorder; using said physiological markers measurements to create a cluster of eating disorder patients, wherein the patients in said cluster have similar attributes and physiological markers measurements, so that within each cluster physiological markers variably will be significantly smaller than the between clusters variability; and processing differences between said physiological marker measurements of the patients, both between clusters and specifically for each patient while in normal stage or during a disorder related event to create a unique personal data set for an individual patient; and using the unique personal prediction machine learning based models and data set to A) set a personal disorder scale; B) set personal severity disorder scoring; C) provide intraday event alerts; and D) provide ongoing analysis to detect treatment effectiveness patterns.

2. The method according to claim 1, wherein said disorder related event comprises selfinduced vomiting.

3. The method according to claim 1, wherein said eating disorder comprises anorexia nervosa, bulimia nervosa, binge eating disorder, pica syndrome, rumination disorder or refrained food consumption.

4. The method according to claim 1, wherein the step of processing differences is done by pattern recognition, machine learning or Al algorithms.

5. The method according to claim 1, further comprising: assigning other patients to one of various patients’ cluster depending on physiological markers measurements of said other patients; using pattern recognition of differences between the physiological markers measurements of said other patients to create a unique personal data set for an individual patient of said other patients; and using additional personal pattern recognition models of differences between the physiological markers measurements to analyze disorder severity, recurrence and ongoing patterns.

6. The method according to claim 5, wherein the unique personal data set further comprises calibration and calculation of a personal baseline.

7. The method according to claim 1, further comprising performing ongoing treatment, including delivering personal analysis of disorder severity, treatment effect and intraday predictions using a personal pattern based on the unique personal data set.

8. The method according to claim 1, further comprising performing ongoing cluster calibration using automation machine learning and shifts between clusters.

Images