AI technology is being used for breast cancer diagnosis and treatment

Local doctors are using artificial intelligence to make it easier to classify and find the target in the most complicated cases of breast cancer. This means that patients no longer have to get a needle stuck in them for a biopsy.

By studying enhanced magnetic resonance imaging (MRI), the AI ​​​​​​can precisely judge the types of breast cancer, helping doctors to plan accordingly.

Breast cancer is the most common cancer among women. Of all the types of breast cancer, the triple-negative variety is the most complicated to treat. With no targeted therapy, chemotherapy is the only treatment option.

About 15 percent to 20 percent of breast cancer patients have triple-negative breast cancer, which is often fatal.

Dr Shao Zhimin , from Shanghai Cancer Center, teamed up with experts in other fields of medicine, brain science, and information technology to come up with an intelligent plan for triple-negative breast cancer diagnosis.

"Clinical diagnosis plays an important role in cancer treatment," Shao said. "For the triple-negative variety, a clear and detailed diagnosis is the key for cancer classification and individualized treatments. There are four types of triple-negative breast cancer. Usually, the classification is based on pathological diagnosis and genetic tests after studying tumor tissues through puncture or surgery. However, such a procedure often causes trauma among patients, and the testing is a complicated and costly process.

"So we are looking at new ways of classification that are less traumatic to patients, less time consuming, and less costly," Shao said. "We hope to be able to do that with the help of clinical doctors and teams of experts in imaging, pathology, and AI."

Through enhanced MRI, experts developed the AI-based non-invasive system that quickly and accurately gave the classification of triple-negative breast cancer. Patients no longer needed to undergo puncture or surgery for a biopsy. By studying the edges of tumors, the system also gave a prognosis for treatment, doctors said.

Moreover, genetic testing is also a complicated molecular biological process. Doctors and scientists conducted digital scanning of a pathological section to accurately identify the mutated genes and treatment targets.

"With such technology, all cancer classification, key mutated genes, and treatment targets can be known in hours, providing huge convenience for follow-up treatment," Shao said. "These technologies will be promoted for clinical use soon."