Researchers developed Oncoformer, a multimodal transformer that analyzes longitudinal health records and chest X-rays to classify cancer, predict diagnoses up to one year in advance, infer tumor stage, and forecast treatment response and recurrence risk. Validation across large independent cohorts showed strong performance, although prospective trials in diverse, average-risk populations are needed before clinical use.
UCLA researchers have identified a promising strategy to make a new class of targeted cancer therapies more effective against one of the deadliest forms of prostate cancer, potentially overcoming treatment resistance that has limited their success.
Risk calculators that doctors use to estimate a person's risk of a heart attack or stroke work about as well for predicting certain cancers, according to a new study by UCL (University College London) researchers.
When a cancer cell leaves the original tumor to spread to a new, far away location, it first has to survive what most cells cannot: periodic high force moments in the vessels and heart of up to 5,000 dynes per centimeter squared as it circulates throughout the body.
Researchers combined microfluidic enrichment, AI-assisted holography, and fluorescence profiling to identify circulating tumor cell candidates in 13 men with metastatic castration-resistant prostate cancer and 8 healthy donors. The platform found higher median candidate counts in patients, while nearly two-thirds of PSMA-positive candidates lacked EpCAM, suggesting EpCAM-dependent assays could miss some cells.
Why did you receive this email?
You are receiving this email because you subscribed to updates from AZoNetwork UK Ltd. on one of
our websites
and requested to be notified of additional information.
