US 11,923,075 B2
System and method associated with determining physician attribution related to in-patient care using prediction-based analysis
Todd R. Griffin, St. James, NY (US); Erin J. Healy, Bethpage, NY (US); I. V. Ramakrishnan, Setauket, NY (US); and Vikas Ganjigunte Ashok, Port Jefferson, NY (US)
Assigned to The Research Foundation for The State University of New York, Albany, NY (US)
Filed by THE RESEARCH FOUNDATION FOR THE STATE UNIVERSITY OF NEW YORK, Albany, NY (US)
Filed on Aug. 11, 2021, as Appl. No. 17/399,511.
Application 17/399,511 is a continuation of application No. 16/244,314, filed on Jan. 10, 2019, granted, now 11,114,197.
Application 16/244,314 is a continuation of application No. 14/723,619, filed on May 28, 2015, abandoned.
Claims priority of provisional application 62/004,507, filed on May 29, 2014.
Prior Publication US 2021/0375443 A1, Dec. 2, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G16H 40/20 (2018.01); G16H 10/60 (2018.01)
CPC G16H 40/20 (2018.01) [G16H 10/60 (2018.01)] 18 Claims
OG exemplary drawing
 
1. A system of determining physician attribution among a plurality of physicians related to in-patient care of a patient, the system comprising:
a processing device;
a non-transitory memory storing instructions that, when executed by the processing device, cause the processing device to perform operations comprising:
receiving a plurality of sets of physician treatment notes related to the in-patient care of the patient, each set of the plurality of sets of physician treatment notes associated with a respective physician of the plurality of physicians;
assigning a feature vector to each physician of the plurality of physicians, the feature vector including assignment of a plurality of feature values, each feature value of the plurality of feature values being associated with a different type of physician treatment note and indicating a number of issued physician treatment notes of the different type, indicative of clinical progress of the patient at various stages of the in-patient care;
automatically determining a weight vector for each feature vector, the weight vector including determination of a plurality of predicted feature weights associated with respective plurality of feature values using machine learning analysis, each predicted feature weight of the plurality of predicted feature weights indicating a level of priority associated with the type of physician treatment note of a respective feature value, the level of priority being different for each of the plurality of predicted feature weights, wherein the machine learning analysis that determines the predicted feature weights is based on a training data set of patients, physicians, and associated physician treatment notes related to providing in-patient care to the patients;
computing an attribution score for each physician of the plurality of physicians based on the feature values of the feature vector and the predicted feature weights of the weight vector associated with the physician; and
generating a clinical representation on a user interface, the clinical representation displaying a plurality of attribution scores of the plurality of physicians to the in-patient care of the patient, and identifying a physician with a highest attribution score from the plurality of attribution scores as an attributed physician predicted to be most responsible for the in-patient care of the patient.