US 12,261,871 B2
Real-time precision care plan support during remote care management
Anthony Dohrmann, El Paso, TX (US); Bryan John Chasko, Las Cruces, NM (US); Juseung Park, Las Cruces, NM (US); and Isaac Davalos, Las Cruces, NM (US)
Assigned to Electronic Caregiver, Inc., Las Cruces, NM (US)
Filed by Electronic Caregiver, Inc., Las Cruces, NM (US)
Filed on Jul. 5, 2024, as Appl. No. 18/765,121.
Application 18/765,121 is a continuation of application No. 17/187,568, filed on Feb. 26, 2021, granted, now 12,034,748.
Claims priority of provisional application 62/983,455, filed on Feb. 28, 2020.
Prior Publication US 2024/0364726 A1, Oct. 31, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. H04L 9/40 (2022.01); G06N 5/043 (2023.01); G16H 40/67 (2018.01); G06N 20/10 (2019.01)
CPC H04L 63/1425 (2013.01) [G06N 5/043 (2013.01); G16H 40/67 (2018.01); G06N 20/10 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A secure intelligent networked architecture for real-time precision care plan remote support comprising:
a secure intelligent data receiving agent having a specialized hardware processor and a memory, the secure intelligent data receiving agent configured to automatically receive a digital data element over a network from a wireless transmission-equipped peripheral device, the digital data element representing an output in response to a predetermined plan;
the secure intelligent data receiving agent configured to perform a risk stratification for a patient's capacity for self-managed care; and
the secure intelligent data receiving agent configured to process the digital data element and configured with logic for anomaly detection, the logic for the anomaly detection being executed by a machine learning model comprising:
at least one support vector machine classifier comprising a plurality of support vectors, the at least one support vector machine classifier configured to detect anomalies by classifying outlier data;
a plurality of input training vectors determined from one or more outcomes from the anomaly detection;
a plurality of input measurements over a number of days;
support vector classification finding a hyperplane, the plurality of support vectors representing data points closest to the hyperplane and defining a decision boundary;
a kernel transforming the plurality of input measurements that support the finding of the hyperplane;
parameter c, a regularization parameter in the support vector classification that controls a trade-off between maximizing a distance between the hyperplane and a nearest data point of a class; and
the plurality of input measurements and corresponding predicted outcomes tested for a predictive performance of the logic for the anomaly detection.