Monitoring surface cleaning of medical surfaces using video streaming

Inventors

Ferrante, Joseph • GROSS, Kelsey • Swart, Elliot

Assignees

Digital Diagnostics Inc

Abstract

A cleaning wizard monitors and provides feedback for cleaning of medical equipment to ensure that cleaning is performed based on best practices. The cleaning wizard receives a video stream comprising an item of medical equipment and inputs a first set of video frames from the video stream into a first machine learning model. The first machine learning model is trained to output whether the first set of video frames corresponds to activity that initiates a cleaning protocol for the item of medical equipment. Responsive to the cleaning protocol being initiated, the cleaning wizard inputs a second set of video frames into a second machine learning model trained to output whether the second set of frames meets criteria of the cleaning protocol. Responsive to all criteria of the cleaning protocol being met, the cleaning wizard transmits a notification to an operator that the cleaning protocol is complete.

Core Innovation

A cleaning wizard monitors medical-surface cleaning by receiving a video stream and analyzing frames to determine whether cleaning actions initiate a cleaning protocol for an item of medical equipment. The cleaning initiation module inputs a first set of frames into a first machine learning model trained to output whether the first set of frames corresponds to activity that initiates the cleaning protocol.

After detecting initiation, a second machine learning model inputs a second set of frames and outputs a pixel map representing one or more aspects of the cleaning protocol based on the second set of frames. The pixel map is then input into one or more classifiers, which output whether pixels of the pixel map conform to required values of the cleaning protocol, so that the second set of frames meet criteria of the cleaning protocol.

Responsive to the second set of frames meeting the criteria, the system transmits a notification to an operator that the cleaning protocol is complete. The monitoring can incorporate criteria expressed in required values and deficiency values for aspects of the cleaning protocol, with operator notifications used when protocol criteria are not met and additional monitoring used to restart the cleaning protocol when criteria continue to fail.

Claims Coverage

The document provides three independent claim sets (method, non-transitory computer-readable storage medium, and system). Across these, the claims include at least five core inventive features: receiving a video stream, two-stage machine-learning-based detection and evaluation, pixel-map generation, pixel conformity classification against required values, and notification to an operator upon protocol completion.

The independent claims consistently require a video-stream driven workflow that uses a first machine learning model to detect initiation of a cleaning protocol and a second machine learning model to produce a pixel map of cleaning protocol aspects. The pixel map is classified for conformity to required values, and only when conformity is met does the system determine the frames meet cleaning protocol criteria and transmit a cleaning-protocol-complete notification to an operator.

Stated Advantages

Documented Applications