Office of Technology Transfer – University of Michigan

Machine Learning for Hepatitis C

Technology #6348

Questions about this technology? Ask a Technology Manager

Download Printable PDF

Ulysses Balis
Managed By
Drew Bennett
Associate Director - Software Licensing 734-615-4004
Patent Protection
US Patent Pending

This technology provides a novel method for the prognosis of Hepatitis C progression. Hepatitis C is often chronic, and of those with chronic Hepatitis C, ~65% eventually develop chronic liver disease, and ~15% develop cirrhosis which is often a precursor for liver cancer. It is currently difficult to predict which individuals will develop complications from chronic Hepatitis C infection. Confident prediction of the disease course could save patients hundreds of thousands of dollars in unnecessary treatment options.

Giving Confidence to a Measured Response

This technology used a machine learning approach to classify patients based on their likely outcomes with very high accuracy rates. Research from developing laboratories has shown that this method outperforms other methods that generally have high accuracy. This has implications for assigning treatment regimens to patients, which can be very expansive. Using this technology, patients and doctors concerned about the financial impact of treatments can have a better idea of the treatment’s necessity.


  • Identification of patients with chronic HepC at high risk for fibrosis progression and adverse clinical outcomes.
  • Guidance for treatment options.


  • The only tool of its kind currently available.
  • Compared to differing models to select the most effective prognostic approach.