US 12,002,580 B2
System and method for customized patient resources and behavior phenotyping
Anjali Kataria, Chevy Chase, MD (US); and Vinay Bhargava, Chevy Chase, MD (US)
Assigned to MYTONOMY INC., Chevy Chase, MD (US)
Filed by Mytonomy Inc., Bethesda, MD (US)
Filed on Jul. 17, 2018, as Appl. No. 16/037,994.
Claims priority of provisional application 62/533,939, filed on Jul. 18, 2017.
Prior Publication US 2019/0027248 A1, Jan. 24, 2019
Int. Cl. G16H 50/20 (2018.01); G06N 3/08 (2023.01); G06N 20/00 (2019.01); G10L 25/63 (2013.01); G16H 10/20 (2018.01); G16H 20/70 (2018.01); G16H 50/70 (2018.01); H04N 7/18 (2006.01)
CPC G16H 50/20 (2018.01) [G06N 3/08 (2013.01); G06N 20/00 (2019.01); G10L 25/63 (2013.01); G16H 10/20 (2018.01); G16H 20/70 (2018.01); G16H 50/70 (2018.01); H04N 7/183 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A method comprising:
receiving, at a computer system, training data sets describing relationships between: (1) patients diagnosed with a physical illness, (2) patient health risk levels, and (3) patient trajectories without intervention;
training a neural network using the training data sets, resulting in a machine learning algorithm;
receiving, at a processor from a mobile device, answers to a behavioral questionnaire, wherein the answers are received in real-time as a patient self-reports answers to the behavioral questionnaire on the mobile device;
recording, via a camera, a video of the patient as the answers are completed;
identifying, via the processor from the video of the patient, at least one facial expression of the patient made during completion of the behavioral questionnaire;
identifying, via the processor from the video of the patient, at least one repeated body posture of the patient made during completion of the behavioral questionnaire;
calculating a time period between instances of the at least one repeated body posture;
calculating, via the processor, a numerical score of the patient based upon the answers to the behavioral questionnaire, the time period, the at least one repeated body posture, and the at least one facial expression of the patient made during completion of the behavioral questionnaire;
receiving, at the processor, a physical medical condition of the patient;
generating, via the processor executing a playlist algorithm using the physical medical condition of the patient and the numerical score, a playlist of videos associated with the physical medical condition;
receiving, at the computer system, a record of which videos within the playlist of videos were watched by the patient;
executing, via the computer system, the machine learning algorithm, with inputs to the machine learning algorithm comprising the record and outputs of the machine learning algorithm comprising a modification to the playlist algorithm; and
modifying, via the processor, the playlist algorithm according to the modification.