Acquisition of device information for medical devices using optical character recognition (OCR)
A new method for acquiring device information using optical character recognition (OCR) for remote monitoring of the extracorporeal membrane oxygenation (ECMO) device was developed in this study.
Abstract
A new method for acquiring device information using optical character recognition (OCR) for remote monitoring of the extracorporeal membrane oxygenation (ECMO) device was developed in this study. The ECMO device used was the SP-200 (Terumo Corp.) because initially its information could not be directly acquired for remote monitoring. In addition, the data management system (DMS) is a software that runs on a tablet PC and can collectively manage and display device information. Therefore, we used the screen output of a tablet PC running DMS to acquire the device information of SP-200 with OCR.A Raspberry Pi 4 Model B was used to capture the laptop PC display output that displayed a DMS image and read the device information with OCR. The captured image was then processed with Raspberry Pi into an image suitable for OCR. By reading with OCR, it was possible to acquire device information from a device from which it was initially not possible to directly do so.