Sensor-based machine learning in a health prediction environment
A machine learning prediction system can analyze a dataset of users with self-reported symptoms and associated data from a wearable device to impact measure the impact of an acute health condition (such as the flu) at the population level. The machine learning prediction system can train a machine learning model to recognize individual acute health condition patterns based on differences in user activity with respect to the characteristics of determined baseline periods. For example, per-individual normalized change with respect to baseline aggregated at the population level can be used to determine individual acute health condition patterns and predict the onset of certain acute health conditions using a trained machine learning model. In response to predictions, the machine learning prediction system can take interventions to manage the impact of a predicted acute health condition on an individual.
1 . A method comprising:
(a) obtaining first time series data in a non-standardized format, remotely over a computer network, at least in part by a first wearable health sensor of a target user;
(b) obtaining second time series data in the non-standardized format, remotely over the computer network, at least in part by a second wearable health sensor of the target user, wherein the second wearable health sensor is different than the first wearable health sensor;
(c) converting the first time series data and the second time series data to a standardized format to generate physical statistics data for the target user, wherein the converting comprises normalizing the first time series data and the second time series data based at least in part on one or both of: (1) prior first time series data of the target user obtained prior to obtaining said first time series data in (a), or (2) prior second time series data of the target user obtained prior to obtaining said second time series data in (b);
(d) determining a probability of onset of an acute health condition for the target user by applying a trained machine learning model to the physical statistics data for the target user;
(e) in response to the probability of onset of the acute health condition exceeding a threshold, automatically generating a notification comprising a warning to the target user of the acute health condition; and
(f) transmitting the notification, over the computer network, to a plurality of user devices to cause modification of each interface displayed by the plurality of user devices to display the notification, thereby providing a plurality of users with access to up-to-date health condition information.
2 . The method of claim 1 , further comprising causing modification of an interface displayed by a user device of the target user to display a notification configured to change a behavior of the target user.
3 . The method of claim 1 , further comprising, in response to the probability of onset of the acute health condition exceeding the threshold, causing a test kit corresponding to the acute health condition to be sent to the target user.
4 . The method of claim 1 , further comprising, in response to the probability of onset of the acute health condition exceeding the threshold, causing a doctor's appointment to be scheduled for the target user.
5 . The method of claim 1 , wherein one or more of a pedometer, a sleep tracker, a smart watch, a smartphone, or a mobile device comprises one or both of the first wearable health sensor of the target user or the second wearable health sensor of the target user.
6 . The method of claim 1 , wherein the trained machine learning model was trained using a set of training data that comprises acute health condition symptom data for a plurality of users.
7 . The method of claim 1 , wherein the acute health condition is an influenza-like illness.
8 . The method of claim 1 , wherein the acute health condition is COVID-19.
9 . The method of claim 1 , wherein the physical statistics data comprises a measurement of one or more of: resting heart rate, activity level, daily step count, or sleep time.
10 . The method of claim 1 , wherein the physical statistics data comprises a measurement of one or more of: respiration rate, heart rate variability, or galvanic skin response.
11 . A non-transitory computer-readable storage medium comprising instructions which, when executed by a processor, cause the processor to perform operations comprising:
(a) obtaining first time series data in a non-standardized format, remotely over a computer network, at least in part by a first wearable health sensor of a target user;
(b) obtaining second time series data in the non-standardized format, remotely over the computer network, at least in part by a second wearable health sensor of the target user, wherein the second wearable health sensor is different than the first wearable health sensor;
(c) converting the first time series data and the second time series data to a standardized format to generate physical statistics data for the target user, wherein the converting comprises normalizing the first time series data and the second time series data based at least in part on one or both of: (1) prior first time series data of the target user obtained prior to obtaining said first time series data in (a), or (2) prior second time series data of the target user obtained prior to obtaining said second time series data in (b);
(d) determining a probability of onset of an acute health condition for the target user by applying a trained machine learning model to the physical statistics data for the target user;
(e) in response to the probability of onset of the acute health condition exceeding a threshold, automatically generating a notification comprising a warning to the target user of the acute health condition; and
(f) transmitting the notification, over the computer network, to a plurality of user devices to cause modification of each interface displayed by the plurality of user devices to display the notification, thereby providing a plurality of users with access to up-to-date health condition information.
12 . The non-transitory computer-readable storage medium of claim 11 , wherein the operations further comprise causing modification of an interface displayed by a user device of the target user to display a notification configured to change a behavior of the target user.
13 . The non-transitory computer-readable storage medium of claim 11 , wherein the operations further comprise, in response to the probability of onset of the acute health condition exceeding the threshold, causing a test kit corresponding to the acute health condition to be sent to the target user.
14 . The non-transitory computer-readable storage medium of claim 11 , wherein the operations further comprise, in response to the probability of onset of the acute health condition exceeding the threshold, causing a doctor's appointment to be scheduled for the target user.
15 . The non-transitory computer-readable storage medium of claim 11 , wherein one or more of a pedometer, a sleep tracker, a smart watch, a smartphone, or a mobile device comprises one or both of the first wearable health sensor of the target user or the second wearable health sensor of the target user.
16 . The non-transitory computer-readable storage medium of claim 11 , wherein the acute health condition is an influenza-like illness.
17 . The non-transitory computer-readable storage medium of claim 11 , wherein the physical statistics data comprises one or more of: resting heart rate, activity level, daily step count, or sleep time.
18 . The non-transitory computer-readable storage medium of claim 11 , wherein the physical statistics data comprises one or more of: respiration rate, heart rate variability, or galvanic skin response.
19 . The method of claim 1 , wherein one or both of: (i) the first time series data uses a different data format than the second time series data or (ii) the first time series data uses a different measurement frequency than the second time series data.
20 . The method of claim 1 , wherein converting the first time series data and the second time series data at (c) comprises transforming one or both of the first time series data or the second time series data.
21 . The non-transitory computer-readable media of claim 11 , wherein one or both of: (i) the first time series data uses a different data format than the second time series data or (ii) the first time series data uses a different measurement frequency than the second time series data.
22 . The non-transitory computer-readable media of claim 11 , wherein converting the first time series data and the second time series data at (c) comprises transforming one or both of the first time series data or the second time series data.