The University of Tokyo · Medicine
Professor Masaya Sato's research lab focuses on advancing the understanding and diagnosis of liver diseases, particularly hepatocellular carcinoma (HCC), through innovative integration of clinical data, molecular biomarkers, and artificial intelligence. The lab specializes in developing machine learning and deep learning models that combine multimodal data—such as medical imaging (ultrasonography), patient background factors, and blood-based biomarkers—to improve diagnostic accuracy and predict disease progression. A key research direction involves elucidating the roles of bioactive lipids, like sphingosine 1-phosphate (S1P), in liver fibrosis and inflammation, as well as the impact of genetic polymorphisms (e.g., IL28B, PNPLA3) on disease outcomes in chronic hepatitis C and HCC. The lab aims to translate these findings into clinically applicable tools for early detection and personalized management of liver diseases.
Figures are computed from collected data and may differ slightly.
Because of its multifactorial nature, predicting the presence of cancer using a single biomarker is difficult. We aimed to establish a novel machine-learning model for predicting hepatocellular carcinoma (HCC) using real-world data obtained during clinical practice. To establish a predictive model, we developed a machine-learning framework which developed optimized classifiers and their respective hyperparameter, depending on the nature of the data, using a grid-search method. We applied the cur
The role of sphingosine 1-phosphate (S1P) in liver fibrosis or inflammation was not fully examined in human. Controversy exists which S1P receptors, S1P1 and S1P3 vs S1P2, would be importantly involved in its mechanism. To clarify these matters, 80 patients who received liver resection for hepatocellular carcinoma and 9 patients for metastatic liver tumor were enrolled. S1P metabolism was analyzed in background, non-tumorous liver tissue. mRNA levels of sphingosine kinase 1 (SK1) but not SK2 wer
IL28B polymorphisms appeared to modify the natural course of disease in patients with CHC. Disease progression seems to be promoted in patients with the rs12979860 CC and rs8099917 TT genotypes.
Integration of patient background and blood biomarkers in addition to US image using multimodal representation learning outperformed the CNN model using US images. We expect that the deep multimodal representation model could be a feasible and acceptable tool for the definitive diagnosis of liver tumors using B-mode US.
Despite recent improvements in therapeutic interventions, hepatocellular carcinoma is still associated with a poor prognosis in patients with an advanced disease at diagnosis. Recently, significant progress has been made in image recognition through advances in the field of artificial intelligence (AI) (or machine learning), especially deep learning. AI is a multidisciplinary field that draws on the fields of computer science and mathematics for developing and implementing computer algorithms ca
The PNPLA3 genotype GG may be associated with accelerated hepatocarcinogenesis in CHC patients through increased steatosis in the liver.
Patient age, CP class, CCI, and duration of anesthesia were identified as important risk factors for predicting postoperative mortality in cirrhotic patients. The ADOPT-LC score effectively predicts in-hospital mortality following elective surgery and may assist decisions regarding surgical procedures in cirrhotic patients based on a quantitative risk assessment.
DTx for NASH was found to be highly efficacious and well-tolerated. Further evaluation of the DTx intervention for NASH in a phase 3 trial is warranted.
Open papers in the app to read, cite, and organize with AI.
