Predictive model for substance monitoring and impact prediction

Patent metadata

Field	Value
Publication number	US-11615876-B2
Application number	US-201816161659-A
Country	US
Kind code	B2
Filing date	Oct 16, 2018
Priority date	Oct 16, 2018
Publication date	Mar 28, 2023
Grant date	Mar 28, 2023

Abstract

A method, computer program product, and a system where a processor(s) obtains data related to physical activities performed by an individual from a sensor(s) proximate to the individual. The processor(s) cognitively analyzes the data to identify baseline behavioral patterns of the individual, when the individual is engaged in each of the physical activities. The processor(s) obtains data indicating consumption of a substance by the individual at a first time. The processor(s) determines impacts of the consumption on the baseline behavioral patterns of the individual and generates a data structure (a predictive model) that includes expected deviations from the baseline behavioral patterns of the individual, when the individual has consumed the substance.

First claim

What is claimed is:

1. A computer-implemented method, comprising:
continuously obtaining, by one or more processors, training data for a machine learning algorithm, wherein the training data comprises data related to physical activities performed by an individual from one or more sensors proximate to the individual, wherein a portion of the data related to the physical activities performed by the individual is selected from the group consisting of: physiological data, heart rate, blood pressure, blood oxygen saturation, respiration, movement data indicating a restful state, movement data indicating an active state, temperature, ambient light readings, eye focus, and noise readings, wherein the data related to the physical activities is collected by the one or more sensors contemporaneously with engagement of the individual in the physical activities;
obtaining, by the one or more processors, data indicating consumption of a substance by the individual at a first time;
generating, by the one or more processors, a predictive model utilizing baseline behavioral patterns of the individual when the individual is engaged in each of the physical activities and expected deviations from the baseline behavioral patterns of the individual when the individual has consumed the substance, wherein the predictive model is utilized to determine one or more probabilities that the individual will exhibit one or more behaviors comprising the expected deviations, and to determine an interval subsequent to consuming the substance in which the individual will exhibit the one or more behaviors, the generating comprising:
training, by the one or more processors, the machine leaning algorithm, utilizing a first set of the training data and a second set of training data, wherein the first set of the training data was obtained prior to consumption of the substance by the individual, and wherein the second set of the training data was obtained subsequent to the first time, to identify impacts of the consumption of the substance at the first time on the identified baseline behavioral patterns of the individual, wherein the first set of training data trains the machine leaning algorithm to identify baseline behavioral patterns of the individual when the individual is engaged in each of the physical activities, and wherein the second set of training data trains the machine learning algorithm to identify of the consumption of the substance at the first time on the identified baseline behavioral patterns of the individual; and
generating, by the one or more processors, based on the training of the machine leaning algorithm, the predictive model;

obtaining, by the one or more processors, data indicating consumption of the substance by the individual at a second time;
determining, by the one or more processors, based on applying the predictive model, a probability that the individual will exhibit one or more behaviors comprising the expected deviations, and an interval subsequent to the second time in which the individual will exhibit the one or more behaviors;
determining, by the one or more processors, that the individual is engaged in a physical activity of the physical activities, wherein the one or more behaviors comprising the expected deviations impact the physical activity;
based on applying the predictive model determining that probability will exceed a pre-defined threshold during the interval subsequent to the second time and impact the physical activity during the interval subsequent to the second time, and the determination that the individual is engaged in the physical activity, transmitting, by the one or more processors, the probability and the interval subsequent to the second time, to the individual, via a computing device comprising a portion of the one or more sensors;
obtaining, by the one or more processors, data indicating consumption of the substance by the individual at a third time;
obtaining, during the continuously obtaining, training data at a given time;
cognitively analyzing, by the one or more processors, the training data obtained at the given time, wherein the cognitively analyzing comprises determining that the training data obtained at the given time comprises values which are outliers to the baseline behavioral patterns of the individual when the individual is engaged in each of the physical activities and to the expected deviations from the baseline behavioral patterns of the individual when the individual has consumed a substance;
determining, by the one or more processors, a number of times values obtained during the continuously obtaining are consistent with the outlier values obtained at the given time;
based on determining that the number of times the values obtained during the continuously obtaining are consistent with the outlier values obtained at the given time exceeds a threshold number, determining, by the one or more processors, the probability that the individual will exhibit one or more behaviors comprising the expected deviations, the determining comprising:
updating, by the one or more processors, the predictive model, wherein the updating comprises re-training, by the one or more processors, the machine learning algorithm with the training data obtained at a given time; and
applying, by the one or more processors, the predictive model, to determine the probability that the individual will exhibit one or more behaviors comprising the expected deviations, and an interval subsequent to a third time in which the individual will exhibit the one or more behaviors; and

based on determining that the number of times the values obtained during the continuously obtaining are consistent with the outlier values obtained at the given time does not exceed a threshold number, determining, by the one or more processors, the probability based on the training data obtained at the given time.

2. The computer-implemented method of claim 1 , wherein the one or more sensors monitor biometrics, behaviors, and motion of the individual when the individual is engaged in the physical activities.

3. The computer-implemented method of claim 1 , wherein the computing device comprising the portion of the one or more sensors is an Internet of Things device.

4. The computer-implemented method of claim 1 , wherein the data indicating consumption of the substance by the individual at the first time comprises contextual data describing the consumption.

5. The computer-implemented method of claim 4 , wherein the contextual data comprises a quantity of the substance consumed by the individual at the first time.

6. The computer-implemented method of claim 1 , wherein obtaining the data indicating consumption of the substance by the individual at the first time comprises obtaining, by the one or more processors, a schedule of planned consumption times for the substance, wherein the first time comprises a planned consumption time, wherein the schedule is accessible via a communication connection to at least one computing resource, and wherein the at least one computing resource is communicatively coupled to the one or more processors.

7. The computer-implemented method of claim 1 , wherein obtaining the data indicating consumption of the substance by the individual at the first time comprises obtaining, by the one or more processors, the data from a device selected from the group consisting of: at least one sensor of the one or more sensors and an image capture device proximate to the individual.

8. The computer-implemented method of claim 1 , wherein obtaining the data indicating consumption of the substance by the individual at the first time comprises capturing, by the one or more processors, the data from a personal computing device utilized by the individual, wherein the

Assignees

• IBM

Inventors

• Bender Michael
• Childress Rhonda L
• Childress Edward Tyrone
• Boss Gregory J

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G16H50/50

  for simulation or modelling of medical disorders · CPC title

• G16H50/70

  for mining of medical data, e.g. analysing previous cases of other patients · CPC title

• G06N3/063

  using electronic means · CPC title