Artificial intelligence for Pediatric Rare Diseases

Background

​

Children with rare diseases, the orphans of medicine, are disadvantaged in many respects: Odyssees from doctor to doctor, wrong diagnoses and clinical mismanagement determine their lives. For +95% of rare diseases, there is no tested or approved medication.

​

At the same time medical progress now enables the collection of incredible amounts of comprehensive medical data. This offers great potential for finding new ways for medical diagnostics. However, there is so much newly generated data that it is impossible for a doctor to keep track of it, analyze it and integrate it into the practice. And this is where Machine and Artificial intelligence can provide a long-awaited help since AI/MI is able to analyze and structure such data and make it useful for personalized medicine, in particular for diagnosing and treating rare diseases for children.

​

While children with rare diseases have the potential to be cured, pathological tendencies can be filtered out in adults by analyzing data with AI. Biomarkers highlighting risks for neurological or cardiological events can be detected at an early stage. Application of AI will support the planning of prevention and pharmaceutical care in a more targeted way.

​

​

Project Goal

 

Together with Care for Rare Foundation and Dr. von Haunerschen Children's Hospital (Ludwig Maximilian University of Munich) we develop an AI platform to support, diagnose and treat children patients in a more efficient, targeted and individualized way. Because helping children with various rare diseases will require creation of hundreds and thousands of AI models, we will open this platform to the community of volunteering Data Scientists and AI Experts who want to help these diseased children. To preserve the highest standards of data security while working with very sensitive medical data, our platform will be decentralized and designed in such a way that actual data remains in clinics, nobody except medical staff can access it and training of AI models takes place on the clinic’s servers. 

​

​

Expected Impact

 

Early detection of rare children’s diseases and minimization time to diagnosis. Decreasing costs for unnecessary testing and treatment and minimization of the use of incorrect medication and possible side effects. And finally saving Children’s Lives!

​

​

Project Phases

​

• Phase 1: Preparation of the data and gathering of requirements for the platform

• Phase 2: MVP creation

• Phase 3: First AI detected diagnoses with the help of data science community

• Phase 4: Scaling and saving further Childrens’ Lives

​

​

Current Status

​

Phase 1 is running.

​

​