ProMUSE Reduces Alzheimer Diagnosis Cost by Deferring MRI/PET to Uncertain Cases

ProMUSE uses a Dirichlet-based evidential classifier on low-cost clinical data first, only escalating to MRI and PET when uncertainty exceeds a threshold, reducing expensive modality acquisition while maintaining classification accuracy for Alzheimer's staging.