What You Need to Know About a Healthy Calorie Deficit

AI tool can reduce the need for invasive diagnostic procedures in ulcerative colitis

Many individuals choose non-invasive procedures over invasive and perhaps unpleasant ones when seeing a doctor. Fortunately, researchers at Tokyo Medical and Dental University (TMDU) have created a technology that can eliminate the necessity for invasive diagnostic tests in ulcerative colitis patients.

Endoscopic and histologic examinations are both useful in the diagnosis of many disorders. In an endoscopic examination, a long flexible tube containing a light and a camera is directly put into a patient's body to evaluate a specific organ or tissue. A biopsy, or a sample of tissue from the patient's body, is taken and analyzed outside the body for a histological examination. TMDU researchers previously created a deep neural network system called DNUC to assess ulcerative colitis, a chronic condition occurring in the large intestine characterized by persistent inflammation in the colon lining. This artificial intelligence program can evaluate photographs of tissues to identify and quantify regions of inflammation and illness without the need for biopsies.

ulcerative colitis

In their most recent work, the researchers expanded the use of DNUC from still pictures to live colonoscopy footage of ulcerative colitis patients. The prospective, multicenter trial comprised 770 patients in total. The researchers proved that DNUC can detect the presence or absence of inflammation in real-time, with good agreement between DNUC results and expert diagnosis. DNUC was also capable of accurately predicting cases of remission.

"We verified that DNUC can detect regions of inflammation and offer an endoscopic score for those locations."

The DNUC scores were compared to expert scores and showed a high degree of agreement, demonstrating the correctness of the DNUC algorithm.

This artificial intelligence technology has the potential to deliver several improvements to the medical industry. "The use of DNUC can decrease the need for biopsies, saving time and money for both patients and clinicians," says Mamoru Watanabe, the study's principal author. The DNUC system may also be able to analyze photos and video data faster than a physician. Furthermore, endoscopy necessitates training, and interpretations of endoscopic data can be subjective, differing from endoscopist to endoscopist. DNUC can provide more quantitative standards in assessments, addressing present difficulties with unpredictability and bias in medical diagnosis.

This method may be used with commercially accessible colonoscopy systems, making it easier to use in clinical practice. DNUC might also help with junior gastroenterology training. Overall, this study demonstrates how artificial intelligence has the potential to improve existing medical treatment.