乳にゅう癌がん 超ちょう音波おんぱ検査けんさで医師いしをサポートするAI活用かつよう機器ききが導入どうにゅう

It is said that one in nine Japanese women will develop breast cancer.

As early detection becomes increasingly important, equipment utilizing AI to assist doctors in diagnosis during ultrasound examinations has been developed and introduced at medical screening facilities in Tokyo.

Breast cancer is the most common cancer among Japanese women, with one in nine women affected. However, the overall survival rate is high, making early detection important.

The government recommends mammography screening using X-rays for breast cancer screening, but some people find the pain of breast compression to be burdensome.

On the other hand, while ultrasound examinations, which use ultrasound waves on the breast, place little burden on the body, a challenge has been the variability in accuracy depending on the physician performing the diagnosis.

Amidst this, Keio University and private companies have developed software that utilizes AI during ultrasound examinations to support doctors diagnoses, and its implementation has begun at medical checkup facilities in Tokyo.

AI has been trained on approximately 5,000 cases of breast cancer lesions, and when it determines that a lesion seen in an ultrasound image of the breast requires a detailed examination, it indicates this with a red frame.

Ultimately, whether or not a detailed examination will be conducted is determined by the doctor, but in pre-introduction tests, it was able to provide correct results at the same level as experienced physicians.

Professor Satoshi Hayashida of the Department of Breast Surgery at Keio University School of Medicine, who co-developed the technology, said, If breast cancer is detected at an early stage, the cure rate rises to 97 or 98%.

By utilizing AI, we hope that trust in the examination will increase, and that people who have avoided mammography due to pain will at least undergo an ultrasound examination.