Sounds like prime territory for machine learning, though I don't know how much variance there can be in the presentation of anomalies and how much training data you'd therefore need.

Do tumors from various cancers typically share any discernable features?