Researchers at Kyung Hee University Hospital at Gangdong said Monday that they have developed a diagnostic model that can predict the metastasis of colorectal cancer to lymph nodes.

The stage and prognosis of colorectal cancer have been confirmed through ultrasound, computed tomography, and pathological examinations. However, it was difficult to predict metastasis due to the method's practical limitations, said Professor Kwak Min-seob of the Department of Gastroenterology, who led the research.

Pathological diagnosis is considered the most important indicator. Still, the inability to carefully screen an enormous amount of images, the limitations of the qualitative assessment scale, and the difference of doctors' judgments acted as obstacles in prediction.

The research team has developed a system predicting lymph node metastasis in colorectal cancer by combining the latest computer image analysis technology with artificial intelligence. Researchers conducted the study on 164 patients with stage 1 to 3 colorectal cancer registered in the Cancer Genome Atlas program launched by the National Institutes of Health.

Ninety-eight patients, or 59.8 percent of total subjects, did not show lymph node metastasis, while 66, or 40.2 percent, showed metastasis.

The research team expects to maximize the therapeutic effect by conveniently and accurately identifying the risk of lymph node metastasis in colorectal cancer patients.

"We can accurately predict metastasis and establish a treatment and follow-up observation method for each patient with the recently developed model," Kwak said. "I believe follow-ups and more in-depth AI studies will lead to the precision medicine."

The study was published in the latest issue of Frontiers in Oncology.