According to a story from The Japan Times, a team of Japanese researchers from Riken and the National Cancer Center were able to use artificial intelligence to diagnose stomach cancer in its early stages with a high degree of accuracy. The success of the experiment could result in better outcomes for patients, as beginning treatment earlier often results in a more powerful response.
About Stomach Cancer
Stomach cancer is a form of cancer that can appear in the stomach lining. The causes of stomach cancer are varied; many cases have been linked to infection from the bacteria Helicobacter pylori. HIV/AIDS, Epstein-Barr virus, smoking, family history, and a diet high in pickled vegetables, processed meat, red meat, smoked foods, and salt are considered potential risk factors. This cancer rarely presents symptoms in the early stages, meaning it is not often discovered until it has reached an advanced stage. Symptoms include loss of appetite, heartburn, diarrhea, nausea, vomiting, bloating, fatigue, weakness, difficulty swallowing, and jaundice. It occurs more frequently in men and is most common in East Asia and Eastern Europe. Treatment may involve surgery, radiation, chemo, and targeted therapy. The global five year survival rate is less than 10 percent, although in the US it is 28 percent. To learn more about stomach cancer, click here.
Study Findings
In the study, an AI module could distinguish between an endoscope image that showed healthy tissue and stomach cancer in less than a second. It could identify cancer with 80 percent accuracy and healthy tissue with 95 percent accuracy. While these accuracy rates could certainly improve, they are comparable to the accuracy of experienced specialists. The team recommends that AI should be implemented as a diagnostic support for doctors as soon as possible.
Diagnosing cancer early is essential in successful treatment. Cancer identified in stage I or II has a far greater chance of being cured, and at the very least, the patient is more likely to be able to survive long-term.
The experiment is just one example of the ways in which AI can help improve medical diagnosis and treatment. At the CHAIN Cup, which was held in Beijing last June, AI was able to demonstrate its superiority over human doctors in detecting disease in MRI images.
