Methods and apparatus for injury prediction based on machine learning techniques

Abstract

Systems and methods of the present disclosure enable injury prediction using one or more processors for receiving a time-varying signal of sensor measurements from a sensor device associated with a user. The processor(s) generate time windows of the time-varying signal, including a series of the sensor measurements across a predetermined time period, and generate motion features based at least in part on the series of the sensor measurements of the time windows. The processor(s) utilize an injury risk classification machine learning model to predict an injury risk during each time window based at least in part on the motion features. An injury alert message is generated based at least in part on the injury risk being predicted; and transmitting the injury alert message to at least one user computing device.

What is claims is: 
     
         1 . A method comprising:
 receiving, by at least one processor, from at least one device, time-varying movement data of a user;
 wherein the at least one device comprises at least one of:
 an imaging device, 
 an accelerometer, or 
 a gyroscope; 
 
 wherein the time-varying movement data comprises at least one movement data vector that represents at least one movement performed by the user; 
   extracting, by the at least one processor, at least one motion feature vector based at least in part on the time-varying data;
 wherein the at least one motion feature vector comprises an encoding of at least one of:
 at least one time-domain feature, or 
 at least one frequency domain feature; 
 
 wherein the at least one motion feature vector represents at least one characteristic of the at least one movement performed by the user; 
   utilizing, by the at least one processor, at least one injury prediction model to predict an injury risk score for the at least one movement as indicating increased risk of injury based at least in part on a difference between the at least one motion feature vector and historical motion feature vectors;
 wherein the historical motion feature vectors are associated with historical motion data representative of historical time-varying data associated with safe movements performed by at least one user; and 
   generating, by the at least one processor, an injury alert message based at least in part on the injury risk score exceeding a threshold value.   
     
     
         2 . The method as recited in  claim 1 , wherein the imaging device comprises a camera. 
     
     
         3 . The method as recited in  claim 1 , wherein the at least one injury prediction model comprises a rules-based algorithm. 
     
     
         4 . The method as recited in  claim 1 , wherein the at least one injury prediction model comprises at least one of:
 a machine learning classifier model,   a novelty detection model,   density based machine learning models,   tree-based machine learning models, or   variance based machine learning models.   
     
     
         5 . The method as recited in  claim 1 , wherein the time-varying movement data comprises at least one of:
 raw data captured by the at least one device, or   derived data produced by processing of the raw data.   
     
     
         6 . The method as recited in  claim 1 , wherein the at least one injury prediction model is configured to:
 determine a movement peak metric based at least in part on the at least one motion feature vector, wherein the movement peak metric is representative of a greatest value associated with the at least one motion feature,   determine that the at least one movement indicates the increased risk of injury based at least in part on the movement peak metric.   
     
     
         7 . The method as recited in  claim 1 , further comprising:
 receiving, by the at least one processor, at least one first historical motion feature vector;   receiving, by the at least one processor, at least one non-injury event label associated with the at least one first historical motion feature vectors;   receiving, by the at least one processor, at least one second historical motion feature vector;   receiving, by the at least one processor, at least one injury event label associated with the at least one second historical motion feature vectors; and   training, by the at least one processor, the at least one injury prediction model with the at least one first historical motion feature vectors, the at least one non-injury event label, the at least one second historical motion feature vectors, and the at least one injury event label in order to predict a non-injury event, an injury event, or both in the time-varying signal.   
     
     
         8 . The method as recited in  claim 1 , wherein the at least one motion feature vector comprises at least one of:
 at least one yaw measurement,   at least one pitch measurement,   at least one roll measurement,   at least one x-axis acceleration measurement,   at least one y-axis acceleration measurement,   at least one z-axis acceleration measurement,   at least one x-axis gyroscopic measurement,   at least one y-axis gyroscopic measurement,   at least one z-axis gyroscopic measurement,   sagittal angle,   twist angle,   lateral angle,   sagittal velocity,   twist velocity, and   lateral velocity.   
     
     
         9 . The method of  claim 1 , further comprising:
 accessing, by the at least one processor, a user profile associated with the user;
 wherein the user profile records a recorded injury risk score representative of a risk of injury to the user; and 
   modifying, by the at least one processor, the recorded injury risk score in the user profile based at least in part on the injury risk score.   
     
     
         10 . A system comprising:
 at least one non-transitory computer readable medium storing processor instructions;   at least one processor in communication with the at least one non-transitory computer readable medium, wherein the at least one processor is configured, upon execution of the processor instructions, to:
 receive, from at least one device, time-varying movement data of a user;
 wherein the at least one device comprises at least one of:
 an imaging device, 
 an accelerometer, or 
 a gyroscope; 
 
 wherein the time-varying movement data comprises at least one movement data vector that represents at least one movement performed by the user; 
 
 extract at least one motion feature vector based at least in part on the time-varying data;
 wherein the at least one motion feature vector comprises an encoding of at least one of:
 at least one time-domain feature, or 
 at least one frequency domain feature; 
 
 wherein the at least one motion feature vector represents at least one characteristic of the at least one movement performed by the user; 
 
 utilize at least one injury prediction model to predict an injury risk score for the at least one movement as indicating increased risk of injury based at least in part on a difference between the at least one motion feature vector and historical motion feature vectors;
 wherein the historical motion feature vectors are associated with historical motion data representative of historical time-varying data associated with safe movements performed by at least one user; and 
 
 generate an injury alert message based at least in part on the injury risk score exceeding a threshold value. 
   
     
     
         11 . The system as recited in  claim 10 , wherein the imaging device comprises a camera. 
     
     
         12 . The system as recited in  claim 10 , wherein the at least one injury prediction model comprises a rules-based algorithm. 
     
     
         13 . The system as recited in  claim 10 , wherein the at least one injury prediction model comprises at least one of:
 a machine learning classifier model, or   a novelty detection model,   density based machine learning models,   tree-based machine learning models, or   variance based machine learning models.   
     
     
         14 . The system as recited in  claim 10 , wherein the time-varying movement data comprises at least one of:
 raw data captured by the at least one device, or   derived data produced by processing of the raw data.   
     
     
         15 . The system as recited in  claim 10 , wherein the at least one injury prediction model is configured to:
 determine a movement peak metric based at least in part on the at least one motion feature vector, wherein the movement peak metric is representative of a greatest value associated with the at least one motion feature,   determine that the at least one movement indicates the increased risk of injury based at least in part on the movement peak metric.   
     
     
         16 . The system as recited in  claim 10 , wherein the at least one processor is further configured to:
 receive at least one first historical motion feature vector;   receive at least one non-injury event label associated with the at least one first historical motion feature vectors;   receive at least one second historical motion feature vector;   receive at least one injury event label associated with the at least one second historical motion feature vectors; and   training the at least one injury prediction model with the at least one first historical motion feature vectors, the at least one non-injury event label, the at least one second historical motion feature vectors, and the at least one injury event label in order to predict a non-injury event, an injury event, or both in the time-varying signal.   
     
     
         17 . The system as recited in  claim 10 , wherein the at least one motion feature vector comprises at least one of:
 at least one yaw measurement,   at least one pitch measurement,   at least one roll measurement,   at least one x-axis acceleration measurement,   at least one y-axis acceleration measurement,   at least one z-axis acceleration measurement,   at least one x-axis gyroscopic measurement,   at least one y-axis gyroscopic measurement,   at least one z-axis gyroscopic measurement,   sagittal angle,   twist angle,   lateral angle,   sagittal velocity,   twist velocity, and   lateral velocity.   
     
     
         18 . The system as recited in  claim 10 , wherein the at least one processor is further configured to:
 access a user profile associated with the user;
 wherein the user profile records a recorded injury risk score representative of a risk of injury to the user; and 
   modify the recorded injury risk score in the user profile based at least in part on the injury risk score.