Combining Supervised and Unsupervised Methods to Support Early Diagnosis of Hepatocellular Carcinoma

The early diagnosis of Hepatocellular Carcinoma (HCC) is extremely important for effective treatment and improvements in diagnosis are indispensable, particularly concerning the differentiation between “early” HCC and non neoplastic nodules. In this paper, we reconsidered the results obtained previously and compared them with the results of an unsupervised method to achieve a deep knowledge on uncertain lesions. This analysis agreed with the predictions on DNs obtained by the supervised system, providing pathologists with reliable information to support their diagnostic process