Environmental sensor-based cognitive assessment

A cognitive facilitator assessment computing system comprising;

• a memory;
• a computing processor;
• a machine learning engine;
• and
• clinical assessment activity behavior tool stored in the memory and further comprising:
• sensor input logic configured, when executed, to provide one or more streams of continuous live smart-home based sensor data corresponding to activities of a resident in a residence, with the residence having a plurality of smart-home based sensors installed therein;
• an activity recognizer configured, when executed, to receive the continuous sensor data real time and to label the sensor data with activity labels to transform the data into measurements of activities of daily living;
• an activity performance features extractor configured, when executed, to automatically extract activity performance features from the activity labeled sensor data, the activities corresponding to activities of daily living and/or mobility, wherein each extracted activity performance feature includes a time-based feature and a sensor-based feature;
• a statistical feature extractor configured, when executed, to automatically extract statistical features from the extracted activity performance features, the extracted statistical features further measuring attributes over time of each extracted activity performance feature between at least two clinical cognitive testing point times;
• and
• a cognitive assessment predictor configured, when executed, to receive the extracted performance features and to submit them to the machine learning engine to automatically predict a cognitive assessment score using the sensor event data and the extracted statistical features, the machine learning engine modeling relationships between clinical cognitive assessment scores and activity performance features to correlate the predicted cognitive assessment score to a clinical cognitive assessment.

A non-transitory computer-readable memory medium containing contents for instructing a computer process to facilitate a cognitive assessment of a resident of a residence using a plurality of smart sensors installed in the residence, by performing a method comprising:

• receiving from the plurality of smart sensors, a continuous stream of raw output data of sensor events, the sensor events corresponding to a plurality of activities engaged in by the individual in the residence, the raw output data of each sensor event including at least a date, time, sensor identifier, and sensor value;
• automatically labeling the raw output data of each sensor event in the continuous stream with an activity label that corresponds to a detected activity of the resident, the automatic labeling transforming the raw output data into measurements of activities of daily living;
• automatically extracting activity performance features from the activity labeled sensor data, wherein each extracted activity performance feature includes a time-based feature and a sensor-based feature;
• automatically extracting statistical features from the extracted activity performance features, the extracted statistical features further measuring attributes over time of each extracted activity performance feature between at least two clinical cognitive testing point times;
• and
• submitting indications of the extracted statistical features to a machine learning engine to predict a cognitive assessment score using the continuous sensor event data and the extracted statistical features, the machine learning engine modeling relationships between clinical cognitive assessment scores and activity performance features to correlate the predicted cognitive assessment score to a clinical cognitive assessment.