OCR in Healthcare: Automating Medical Records | Case Study

OCR Automation for Medical Sector

Schematic Diagram of Solution

casestudy 2
The Problem:

The client services US Hospitals, performing Medical Records Review [MRR], processing their patients’ medical records, converting manually from unorganised documents into organised ones. The manual process by humans takes time and limits the number of requests that can be serviced per day. The results of this manual process are still a collection of diverse and unconnected documents, posing a difficult task for medical staff to obtain an overall picture of the patients’ medical history. Also, as the reports may have 100s of pages, human error may produce an inaccurate final report.

The above limitations restrict the client from scaling the business without adding staff, resulting in increased ongoing costs without any increase in accuracy or speed of delivery.

The Solution:

vInnovate implemented an OCR based solution, powered by vDigiDocr, our AI OCR Platform. A web application with a custom User Interface was provided to upload medical records along with a summary and index.Uploaded documents were processed by the OCR engine, making them searchable and sortable. A Summary and Index with hyper-links were added automatically, making it a fully-fledged organised document which can be readily used by doctors for each review of the patient’s health history and status. The software allows duplicates to be removed as well as blanks.


Business Value:

Each record, sometimes with several hundred pages, used to take hours for the staff to process. Now it is done in seconds or minutes, thus bringing a high level of productivity, with the capacity to onboard more clients with fewer staff on the ground.

Thus, the client has achieved high productivity with reduced cost.
Client engagement is achieved by providing access to the system so that whenever a report is generated, it is readily viewable and downloadable by the client in near real-time.


Technology :
  • MERN Stack for UI & API Implementation.
  • AI & OCR technology for core OCR Engine implementation.
  • Tesseract used as open source OCR library for English text extraction.
  • Usage of pdf and other libraries for pre and post processing of pdf and images.
  • Deployment on intranet server with powerful CPU of 8 core with 32GB memory.
Social Icons