AITeky Learn
Mini-project Healthcare & Hospitals

Lab report extractor

Turn lab reports into structured values with units and reference ranges, and flag abnormal results.

The brief

Clinics receive lab reports as PDFs and photos in many layouts. Build an extractor that returns each test, value, unit and reference range as validated JSON, and flags results outside the range for the doctor. Good looks like: accuracy per field on labelled reports, unit and range checks, and a clear abnormal-results list.

What you will build

  1. Create sample reports from synthetic patient data
  2. Extract values with a vision-capable LLM and structured outputs
  3. Validate units and reference ranges
  4. Measure accuracy on 30 labelled reports
  5. Flag abnormal results in a simple view

Free, open data to build it with

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.

We use essential cookies to run the site. With your permission we also set one anonymous ID to count unique visitors, and Google Analytics cookies to see which pages help learners. Cookie notice

Join AiTeky Learn. It's free.

Create a free account to track your progress, save items for later and get study suggestions.

Sign up free Log in