- using machine learning to aid diagnosis of rare diseases

Funding: DTU and the Lundbeck Foundation
Project management: Ole Winther and Dan Svenstrup 
Duration: 01.09.14-31.12.17

"When you hear hoofbeats, think of horses not zebras". Sometimes, however, physicians are confronted with rare diseases (the zebras). FindZebra is a tool for helping diagnosis of rare diseases. It uses freely available high quality curated information on rare diseases and open source information retrieval software (Apache Lucene Solr) tailored to the problem.

Orphan diseases affect more than 30 million Americans. 40% of patients are misdiagnosed or un-diagnosed. Traditional search engines need to be optimized to help diagnose rare diseases from symptoms. This is where FindZebra comes in! FindZebra is a symptom checker. Enter clinical and phenotype information, symptoms and findings and FindZebra returns potential diseases and genes involved. FindZebra is intended primarily for physicians and other professionals concerned with diagnosis of rare diseases.

FindZebra’s technology is spun out of Technical University of Denmark.


Ole Winther
DTU Compute
+45 45 25 38 95