Ultrasound model: an impressive tool in liver cancer early diagnosis

Dr. Yeun-Yoon Kim from Yonsei University in South Korea will unveil a groundbreaking ultrasound model that integrates clinical liver cancer early diagnosis. This research showcases the model's superior performance compared to previously reported risk scoring systems in the field.

Anticipating and effectively managing the risk of hepatocellular carcinoma is pivotal for triaging and developing optimal treatment strategies. Emerging models, incorporating ultrasound features, aim to revolutionize risk prediction in this context, with a specific focus on leveraging AI in oncology.

Dr. Kim and his colleagues embarked on the journey to develop and validate a model specifically designed to predict a super-high-risk group of incident HCC in patients with chronic hepatitis B or C under ultrasound surveillance. The incorporation of AI in oncology adds a powerful dimension to the model's capabilities in early diagnosis.

The team leveraged a developmental dataset comprising 7,918 patients and an internal validation dataset with 3,393 patients. Clinical characteristics and ultrasound features, such as the presence of cirrhosis, fatty liver, splenomegaly, ascites, and cirrhotic nodules, were crucial components in shaping the model. To assess its efficacy, the group compared the model's performance against existing risk models.

Results from the study revealed five-year cumulative incidence rates of hepatocellular carcinoma at 7.6%, 7.4%, and 4.3% in the development, validation, and external test datasets, respectively.

Regression analysis identified several clinical and ultrasound features independently associated with cancer risk, including age, sex, diabetes mellitus, serum albumin and alanine aminotransferase levels, platelet counts, cirrhotic parenchymal echotexture on imaging, and multiple cirrhotic nodules on imaging. The integration of AI in cancer treatment enhances the precision of these risk associations.

The team reported five-year cumulative cancer incidence rates in the validation and external test datasets, standing at 24.1% and 15.5%, respectively. AI in oncology diagnosis, particularly in early diagnosis, proves instrumental in achieving these groundbreaking outcomes. "In light of limited alternative imaging modalities in cancer surveillance, our prediction model serves to focus on the super-imaging modalities," the team noted. This advancement holds significant promise for refining hepatocellular carcinoma risk assessment and tailoring preventive measures for improved patient outcomes. The integration of AI in cancer treatment and oncology diagnosis marks a paradigm shift in the approach to early diagnosis and proactive management of hepatocellular carcinoma.

-------------------------------------

The application of AI in oncology is not uncommon nowadays. Doctors have been using the power of smart computers to identify and analyze cancer, automatically diagnosing the probability of having cancer instantly.

Introducing DrAid™ CT Liver Cancer D&T, an AI technology that was built by VinBrain, served as a machine-learning technology for cancer detection. Artificial intelligence in healthcare systems is designed to assist radiologists in automatically detecting abnormal lesions through CT imaging, provide clinical solutions to aid in the early identification of liver cancer, and assist oncologists in developing treatment plans.

Learn more about DrAid™ CT Liver Cancer D&T now to see how the future of liver cancer diagnostics works!