EXoPERT EMERALD Study to Develop AI Diagnostic Software Using Blood Samples for Early Detection of Lung, Breast, Colorectal, Pancreatic, and Ovarian Cancers

What this Trial Is About

Researchers are conducting a multi-center study to create a clinical plasma repository and evaluate an artificial intelligence-based in vitro diagnostic software called ExoPred. This software analyzes surface-enhanced Raman spectroscopy SERS profiles of extracellular vesicles from human plasma to support early detection of multiple cancers including lung, breast, colorectal, pancreatic, and ovarian cancers. The study aims to develop, assess, and validate this diagnostic tool through a structured multi-phase process. The study includes three phases Phase 1 involves analyzing at least 500 samples to train the algorithm and set initial clinical cut-off values, followed by a database lock. Phase 2 tests a minimum of 50 new unknown samples to evaluate model robustness before the final algorithm is locked. Phase 3 is a blinded independent stage where at least 200 independent samples, excluded from prior phases, are tested to evaluate clinical sensitivity, specificity, and accuracy. All data is managed under strict quality procedures and regulatory compliance. Participants provide blood samples before any systemic or definitive cancer treatment. The study collects and analyzes these samples over up to 36 months to measure sufficient sample testing and diagnostic performance. Data handling follows standard operating procedures within a compliant electronic data capture system. Participants are assessed for eligibility based on cancer diagnosis and treatment history, with safety and data integrity monitored throughout the study duration.

EXoPERT EMERALD Clinical Study