Better Leukemia Diagnostics Through AI (BELUGA): A Case-Control Study to Evaluate Artificial Intelligence for Blood Cancer Diagnosis

What this Trial Is About

Researchers are evaluating the usefulness of deep learning-based artificial intelligence (AI) algorithms for diagnosing hematologic malignancies such as leukemia and lymphoma. This study, called BELUGA, aims to prospectively compare AI-guided diagnostics with current human-based standard methods at the Munich Leukemia Laboratory. The goal is to assess if AI can improve diagnostic sensitivity, specificity, and speed, as well as impact clinical decision-making in real-world hematology settings. The study uses an unprecedented collection of 25,000 digitalized blood smears and 25,000 flow cytometry data points collected over 15 years to train a deep neural network. This AI system will analyze blood and bone marrow samples from patients with suspected hematological malignancies. The study involves three parts: training the AI with historical data, prospectively testing it on new samples, and evaluating how AI-assisted diagnosis affects downstream testing like genetic analysis. Participants provide blood or bone marrow samples that undergo both routine human evaluation and AI-based analysis. Researchers will measure diagnostic accuracy, time to diagnosis, and clinical consequences of AI use compared to standard methods. The study includes quality control for samples and will track multiple outcomes from August 2020 to July 2021. Participation involves no treatment but focuses on diagnostic testing and improving patient care through enhanced diagnostic tools.

Better Leukemia Diagnostics Through AI (BELUGA)