ProMapp: Prospectiveclinical trial utilizing machine learning for more efficient diagnosis and stratification of pheochromocytoma and paraganglioma

Principal Investigators

Scientific Staff

Angelos Filippatos – Groupleader
Georg Pommer  – Student
Manuel Schulze – TA
Carola Kunath – TA

Project Description

We have recently established machine learning (ML) models for streamlining diagnostic stratification of patients with suspected pheochromocytomas/paragangliomas (PGGLs) and further identification of those with synchronous metastatic disease or at high risk for developing metastases. Promapp is an international multicenter prospective cohort study with randomized intervention that will actually establish whether our ML models can be used to more efficiently and effectively to streamline the processes for both diagnostic screening of patients with suspected PPGL and follow-up of confirmed cases for identifying presence of or future development of metastatic disease.

Aims

(I) Establish whether provision of machine learning-based interpretations to standard laboratory results (i.e., plasma metanephrines and methoxytyramine) improves the efficiency and effectiveness of confirmation or exclusion of PPGL (Phase 1).

(II) Establish whether provision of machine learning-based interpretations to standard routine care improves the efficiency and effectiveness of clinicians for diagnosis of metastatic disease (Phase 2).

(III) To prospectively validate and improve ML algorithm-based models through learning-based adjustments under real time diagnostic conditions in a population based diagnostic approach where the prevalence of a PPGL is expected at around 1% and the prevalence of metastatic disease for patients with PPGL at around 15-20%.

Publications (CRC/TRR 205 included)