International Consensus Histopathological Criteria for Subtyping Idiopathic Multicentric Castleman Disease Based on Machine Learning Analysis.

Summary: Imagine trying to sort different types of apples, but they all look almost exactly the same on the outside. That is what doctors face with a rare illness called Castleman disease. To fix this, researchers used a smart computer program (machine learning) to look closely at lymph node tissue samples from over 100 patients. The computer found three specific clues inside the cells that help tell the different types of the disease apart. This new system was over 90% accurate! Now, doctors have a much better tool to diagnose patients correctly and choose the right treatments.