Harvard Scientists Design AI Tool That Diagnoses Cancer and Predicts Patient Survival

In a latest development in AI-powered healthcare, scientists at Harvard Medical School have designed a versatile, ChatGPT-like AI model dubbed CHIEF (Clinical Histopathology Imaging Evaluation Foundation) to perform an array of diagnostic tasks across multiple forms of cancer.

The new model can perform a wide array of tasks and has been tested on 19 cancer types, giving it flexibility like that of large language models such as ChatGPT, according to the Harvard Gazette study.

The study further states that CHIEF was trained on 15 million unlabeled images chunked into sections of interest. The tool was then trained further on 60,000 whole-slide images of tissues including lung, breast, prostate, colorectal, stomach, esophageal, kidney, brain, liver, thyroid, pancreatic, cervical, uterine, ovarian, testicular, skin, soft tissue, adrenal gland, and bladder.

The model has been trained to look both at specific sections of an image and the whole image, allowing it to relate specific changes in one region to the overall context. Through this holistic interpretation, CHIEF achieved nearly 94 percent accuracy in cancer detection and significantly outperformed current AI approaches across 15 datasets containing 11 cancer types.

Following training, the team tested CHIEF’s performance on more than 19,400 whole-slide images from 32 independent datasets collected from 24 hospitals and patient cohorts across the globe.

“Our ambition was to create a nimble, versatile ChatGPT-like AI platform that can perform a broad range of cancer evaluation tasks,” said study senior author Kun-Hsing Yu, Assistant Professor of Biomedical Informatics in the Blavatnik Institute at Harvard Medical School. “Our model turned out to be very useful across multiple tasks related to cancer detection, prognosis, and treatment response across multiple cancers.”

“If validated further and deployed widely, our approach and approaches similar to ours could identify early on cancer patients who may benefit from experimental treatments targeting certain molecular variations, a capability that is not uniformly available across the world,” Yu said.

Overall, the study reveals that CHIEF has outperformed other state-of-the-art AI methods by up to 36 percent on the following tasks:

1. Cancer cell detection

2. Tumor origin identification

3. Predicting patient outcomes

4. Identifying the presence of genes and DNA patterns related to treatment response