Harvard boffins build multimodal AI system to predict 14 types of cancer

Multimodal AI models, trained on numerous types of data, could help doctors screen patients at risk of developing multiple different cancers more accurately.. Researchers from the Brigham and Women's Hospital part of Harvard University's medical school developed a deep learning model capable of identifying 14 types of cancer. …

[Roland6]

A good test for this system would be to run the individual test subjects through the final system to see how many previously undetected (or wrongly classified) cases are found.

[John Smith 19]

a single overarching system, made up of numerous deep-learning based algorithms

IOW multiple multi-layer ANN's

With multiple biases dependent on the training sets they have been fed.

Which won't be able to tell you how they did what they did.

On the surface an excellent result

Underneath. Not so much.