Towards AI-aided diagnosis

Helping medical experts to make the right decisions, faster.

Nowadays, the vast majority of medical diagnosis and decisions are supported by data. Be it lab results of a blood test, or the radiology analysis of a computed tomography image set. Medical experts have learned to interpret this data, and extract diagnostically relevant aspects from it with astonishing precision.

Yet, the treasure of combined medical images and data records can yield even more insights, when it is combined and evaluated using modern machine learning-techniques. This is our mission.

Tumor diagnosis and care

A multi-disciplinary attempt.

Tumor diagnosis, prognostication and treatment has always been a multi-disciplinary art. Oncologists, pathologists, and surgeons - to only name the main disciplines - combine their skills to derive the individually most suitable therapy for a patient.

As computer scientists skilled in creating AI solutions, we join these disciplines. It's our goal to find out how to best support each discipline, during diagnosis, therapy, and follow-up.

Based on our strong local network in Ingolstadt and Erlangen, we collaborate with specialists from medicine and computer sciences worldwide.

Multimodal

Combining volumetric and other data

Every lab test and imaging modality has it's strenghts and weaknesses in diagnosis. We therefore combine multiple modalities like magnetic-resonance imaging or ultrasound with pathology images and findings to create AI algorithms that incorporate all this knowledge and aim to help in a more precise diagnosis and treatment planning.

Open Datasets

Bigger is better

It has been often said that data is the new oil in the information age. This is certainly true for machine learning. We think that data sets should not only be available to big companies, who can afford acquisition and labor for labeling.

We therefore keep releasing open data sets, many of which represent the largest in their field.