Optical Character Recognition of Medical Records Based on Deep Learning

Retrieving the medical records of the insured is a crucial step in the insurance company’s claim investigation and helps to prevent insurance fraud. Due to the huge workload and sometimes even additional expenses incurred by entrusting a third party, it’s necessary to develop a medical record processor, which can automatically recognize medical records. Optical character recognition (OCR) has been a hot research topic in recent years owing to its outstanding ability to transmute scanned images and other visuals into text. A deep-learning-based OCR software was developed for the detection and recognition of medical records recognition and automatic calibration of medical records OCR in both Chinese and English are studied and realized, including classifying the physical drawings and scanned copies and removing the seal on the medical record sheet of scanned ones. Based on the observations from the analysis we give several suggestions related to the design and implementation of medical record processing approaches.