System and method for characterizing and monitoring health of an animal based on gait and postural movements

Abstract

One variation of a method includes: accessing a video feed recorded at a training apparatus configured to dispense primary reinforcer units; dispensing a first primary reinforcer unit toward a first location; in response to detecting the animal at the first location, dispensing a second primary reinforcer unit toward a second location intersecting a pathway extending from the first location; in response to detecting movement of the animal along the pathway, collecting a timeseries of position data representing changes in position of a set of anatomical features of the animal; based on the timeseries of position data, deriving a movement profile representing movement of the animal along the pathway; based on a difference between the movement profile and a baseline movement profile defined for the animal, interpreting an abnormality and predicting a causal pathway for the abnormality; and selecting a mitigation protocol based on the causal pathway.

I claim: 
     
         1 . A method for monitoring health of an animal during autonomous training with a training apparatus, comprising:
 during a first training session:
 accessing a video feed recorded by an optical sensor defining a field of view intersecting a working field and integrated into the training apparatus configured to dispense units of a primary reinforcer toward a range of locations within the working field via a dispenser integrated into the training apparatus; and 
 in the video feed, detecting the animal in the working field; 
   in response to detecting the animal at a first location within the working field:
 calculating a target location for dispensation of a first unit of a primary reinforcer based on the first location and a target pathway defining a target distance and a target orientation of the animal relative the training apparatus; and 
 dispensing the first unit of the primary reinforcer toward the target location; and 
   during a first test period, in response to detecting movement of the animal along the target pathway toward the target location in the video feed:
 collecting a first timeseries of position data representing changes in position of a set of anatomical features of the animal during the first test period; 
 deriving a first movement profile for the animal based on the first timeseries of position data, the first movement profile representing movement of the animal along the target pathway during the first test period; 
 accessing a baseline movement profile defined for the animal; and 
 characterizing health of the animal based on a difference between the baseline movement profile and the first movement profile. 
   
     
     
         2 . The method of  claim 1 :
 further comprising, at an initial time during the first training session, dispensing an initial unit of the primary reinforcer toward a first target location, within the working field, via the dispenser occupying a first dispenser position aligned to the first target location;   wherein calculating the target location based on the first location and the target pathway in response to detecting the animal at the first location comprises, in response to detecting the animal at the first location, falling within a threshold distance of the first target location, calculating a second target location for dispensation of the first unit of the primary reinforcer based on the first target location and the target pathway; and   wherein dispensing the first unit of the primary reinforcer toward the target location comprises dispensing the first unit of the primary reinforcer toward the second target location via the dispenser occupying a second dispenser position aligned to the second target location.   
     
     
         3 . The method of  claim 1 :
 wherein characterizing health of the animal based on the difference between the baseline movement profile and the first movement profile comprises:
 characterizing the difference between the baseline movement profile and the first movement profile, the baseline movement profile derived from timeseries of position data collected for the set of anatomical features of the animal during a preceding time period; 
 in response to the difference exceeding a threshold difference, identifying a first abnormality in the first movement profile for the animal; 
 predicting a first causal pathway, in a set of causal pathways, for the first abnormality based on the difference and characteristics of the first movement profile; and 
 characterizing a first health score for the animal based on detection of the first abnormality in the first movement profile and prediction of the first causal pathway; and 
   further comprising selecting a first treatment protocol for the animal based on the first health score.   
     
     
         4 . The method of  claim 1 , wherein characterizing health of the animal based on the difference between the baseline movement profile and the first movement profile comprises:
 accessing a diagnostic model linking abnormalities in movements of animals to a set of causal pathways;   interpreting a first confidence score for a first causal pathway, in the set of causal pathways, for the first abnormality based on the difference and the diagnostic model; and   characterizing health of the animal based on the first confidence score and the first causal pathway.   
     
     
         5 . The method of  claim 1 :
 wherein characterizing health of the animal based on the difference between the baseline movement profile and the first movement profile comprises characterizing a first health score for the animal based on the difference between the baseline movement profile and the first movement profile; and   in response to the first health score falling below a threshold health score:
 terminating the first training session; 
 generating a notification indicative of the first health score and comprising a suggestion to further investigate health of the animal; and 
 transmitting the notification to a user associated with the animal. 
   
     
     
         6 . The method of  claim 1 , further comprising:
 dispensing a second unit of the primary reinforcer toward a second target location within the working field;   in response to detecting the animal at the second target location within the working field, dispensing a third unit of the primary reinforcer toward a third target location within the working field, the third target location defining:
 a second target distance between the training apparatus and the third target location; and 
 intersecting a second target pathway extending between the second target location and the third target location and defining a second target orientation of the animal relative the training apparatus; 
   in response to detecting the animal at the third target location and in the second target orientation:
 outputting a command associated with a target pose in a set of target poses; 
 during a first transition period, in response to detecting transition of the animal into the target pose, collecting a second timeseries of position data representing changes in position of the set of anatomical features of the animal during the first transition period; and 
 at a first time, in response to detecting the animal in a first pose falling within a threshold deviation of the target pose, collecting a set of position data representing position of the set of anatomical features of the animal in the target pose at the first time; and 
   where deriving the first movement profile for the animal based on the first timeseries of position data comprises:
 deriving a gait profile for the animal based on the first timeseries of position data, the gait profile representing movement of the animal along the target pathway during the first test period; 
 deriving a transition profile for the animal based on the second timeseries of position data, the transition profile representing movement of the animal transitioning into the target pose during the first transition period; 
 deriving a pose profile for the animal based on the set of position data, the pose profile representing posture of the animal in the target pose at the first time; and 
 deriving the first movement profile for the animal during the first training session based on the gait profile, the transition profile, and the pose profile.