Grant: Idea to Innovation

Many medical devices lack the ability to export information in a digital format or have this option protected behind proprietary options, thus treatment and physiologic information must be collected manually. This factor hinders the ability to capture high frequency data and burdens medical staff with the need to manually collect the data, introducing risk of errors and burnout. The development of optical character recognition (OCR) techniques allows for the conversion of text within images to a digital format. The goal for our team is to develop a complete system that continuously converts information from various medical device screens into a digital format using computer vision, providing a universal hands-free digital data transcription which reduces staff burnout and errors. Through our past work, we have previously created a prototype system (completed technology readiness level (TRL) 1 through 3) that uses a consumer grade web camera with OCR techniques to digitize medical data (in a continuous fashion) that typically required manual recording by nurses (patient monitors and drug pumps). This is done through taking a picture with the camera, using image manipulation techniques to make the image optimal for OCR, then using OCR to extract automatically all of the text within the image. We have subsequently expanded this to extract data from multiple devices and device types simultaneously. NSERC I2I supported a formal Canadian market assessment (MA; I2IPJ 586104-23) demonstrating wide administrative, clinical, and academic interest in the technology with no existing competitors. Our intellectual property (IP) now has provisional US patent application filed by the University of Manitoba (UM; May 27th, 2024). However, before this system can be widely implemented, aspects of the OCR and image manipulation implementation require more refinement to work under less ideal conditions, and an expanded international MA is required to attract future industry partners. Through this NSERC I2I Phase I grant, we propose a multi-faceted approach to bridge these existing gaps, optimally positioning us for upcoming industry partner discussions around IP acquisition and bringing the device to market.