Providing optical character recognition in healthcare contexts

About The Customer

The customer is a not-for-profit healthcare network with a range of services. It is New York state’s largest healthcare provider with 23 hospitals and more than 700 outpatient facilities.

The Project

As they maintain a medical storehouse, the customer receives massive quantities of medical goods and equipment. They decided to turn to Exadel to create an AI-based module for asset management that would read labels and recognize serial numbers, expiration dates, and other critical pieces of information.

Solution

The Exadel team developed an AI-based asset management module that processes photos of manufacturer labels on medical goods and extracts target data: serial numbers, brands, models, humidity, temperature, etc. It detects symbols, understands their relationships with surrounding text, and makes appropriate decisions. The solution doesn’t depend on a predefined format and uses AI to work with any image type.

The Optical Recognition module that uses three different methods. The barcode scanner involves a convolutional deep learning neural network that finds standardized barcodes and retrieves required information. Then the program finds text on the page that corresponds with predefined patterns. All these methods are used in parallel, and then the module combines results for better accuracy.

The new AI asset management solution is built with a modern technology stack:

  • Back end: Java, PostgreSQL
  • Mobile: Ionic
  • AI module: Python, Tesseract, and Tensorflow

Results

The embedded AI-assisted tool for asset management helps automate manual work, minimize human errors, and expedite business processes. Due to the solution’s ability to easily integrate with existing systems and solutions, the client automated repetitive work and reduced costs.

How can we help you?
Contact Us