I want to collect data from a longitudinal medical study. This means several tens of thousands of physical documents, that can contain printed text, handwritten annotations, checkboxes, anatomical schemes and images, with many different layouts due to various sources and temporal distribution.
Is there a software option or SDK that specializes in the medical field and would aid me with this task? I know about the common solutions from ABBYY, IRIS, LEADTOOLS and so on and they partly advertise document processing in the medical environment, but are either specialized in highly structured forms or say nothing about their application at all. I also know abaout several approaches to document classification using AI in the non-medical sector, maybe somebody has dared the adaptation allready.
Requirements:
- Input: rasterized images
- Languages: western european (mostly german and italian)
- Output: preferably in JSON format
- Platform: Linux / Windows
- On-Premises
(I am a bit surprised that the associated tags icr, iwr, hwr and omr do not exist yet, but I can't create them either.)