Background: Ultrasound imaging is the most frequently performed for the patients with chronic hepatitis or liver cirrhosis. However, ultrasound imaging is highly operator dependent and interpretation of ultrasound images is subjective, thus well-trained radiologist is required for evaluation. Automated classification of liver fibrosis could alleviate the shortage of skilled radiologist especially in low-to-middle income countries. The purposed of this study is to evaluate deep convolutional neural networks (DCNNs) for classifying the degree of liver fibrosis according to the METAVIR score using US images. Methods: We used ultrasound (US) images from two tertiary university hospitals. A total of 7920 US images from 933 patients were used for training/validation of DCNNs. All patient were underwent liver biopsy or hepatectomy, and liver fibrosis was categorized based on pathology results using the METAVIR score. Five well-established DCNNs (VGGNet, ResNet, DenseNet, EfficientNet and ViT) was implemented to predict the METAVIR score. The performance of DCNNs for five-level (F0/F1/F2/F3/F4) classification was evaluated through area under the receiver operating characteristic curve (AUC) with 95% confidential interval, accuracy, sensitivity, specificity, positive and negative likelihood ratio. Results: Similar mean AUC values were achieved for five models; VGGNet (0.96), ResNet (0.96), DenseNet (0.95), EfficientNet (0.96), and ViT (0.95). The same mean accuracy (0.94) and specificity values (0.96) were yielded for all models. In terms of sensitivity, EffcientNet achieved highest mean value (0.85) while the other models produced slightly lower values range from 0.82 to 0.84. Conclusion: In this study, we demonstrated that DCNNs can classify the staging of liver fibrosis according to METAVIR score with high performance using conventional B-mode images. Among them, EfficientNET that have fewer parameters and computation cost produced highest performance. From the results, we believe that DCNNs based classification of liver fibrosis may allow fast and accurate diagnosis of liver fibrosis without needs of additional equipment for add-on test and may be powerful tool for supporting radiologists in clinical practice.

Background: Although surgical resection is the only potentially curative treatment for biliary tract cancer, the prognosis remains poor after a major operation such as pancreatoduodenectomy or hepatectomy. We aimed to investigate the impact of preoperative body compositions on long‐term survival of patients undergoing resection of biliary tract cancer.Methods: We analysed data of patients diagnosed with biliary tract cancer who underwent surgery from 2009 to 2015. Skeletal muscle area, skeletal muscle radiation attenuation, and visceral and subcutaneous adipose tissue areas were measured from the computed tomography images at L3 vertebral levels obtained before resection of cancer. Patients were divided into two groups based on the sex‐specific median values for each parameter, and long‐term survival was compared between the groups.Results: A total of 371 patients (women, 39.6%; mean age, 66.2 ± 9.6 years) were finally included in the analysis. Patients with low skeletal muscle index (SMI) had significantly shorter median survival than those with high SMI (29 vs. 39 months; P = 0.026). Patients with low skeletal muscle attenuation (SMA) also showed reduced survival compared with those with high SMA (median survival 25 vs. 60 months; P = 0.002). Combining these two factors, survival was highest in the high SMI/high SMA group (reference) and lowest in the low SMI/low SMA group (hazard ratio, 2.18; 95% confidence interval, 1.44–3.30). Visceral and subcutaneous adipose tissue areas were not associated with long‐term survival.Conclusions: Low SMI and low SMA on computed tomography scan have a negative impact on survival after resection of biliary tract cancer. They can be used in preoperative risk assessment to assist in treatment decision making.

A commercial MRI-based deep learning algorithm for prostate cancer detection showed greater positive predictive value, despite its lower sensitivity, potentially allowing it to assist radiologists in biopsy planning.

Ultrasonography (US) is recommended as a standard surveillance tool for patients with a high risk of developing hepatocellular carcinoma (HCC). However, the low sensitivity of US for small HCC can lead to surveillance failure, resulting in advanced stage tumor presentations. For the early detection of HCC in high-risk patients and to improve survival and prognosis, a new efficient imaging tool with a high sensitivity for HCC detection is needed. The purpose of this study is to evaluate and compare the feasibility and efficacy of non-contrast magnetic resonance imaging (MRI) with US as a surveillance tool for HCC in patients with liver cirrhosis. MAGNUS-HCC is a prospective, multicenter clinical trial with a crossover design for a single arm of patients. This study was approved by six Institutional Review Boards, and informed consent was obtained from all participants. All patients will undergo liver US every 6 months and non-contrast liver MRI every 12 months during a follow-up period of 3 years. If a focal liver lesion suspected of harboring HCC is detected, dynamic liver computed tomography (CT) will be performed to confirm the diagnosis. After the last surveillance round, patients without suspicion of HCC or who are not diagnosed with HCC will be evaluated with a dynamic liver CT to exclude false-negative findings. The primary endpoint is to compare the rate of detection of HCC by US examinations performed at 6-month intervals with that of yearly non-contrast liver MRI studies during a 3-year follow-up. The secondary endpoint is the survival of the patients who developed HCC within the 3-year follow-up period. MAGNUS-HCC is the first study to compare the feasibility of non-contrast MRI with US as a surveillance tool for the detection of HCC in high-risk patients. We anticipate that the evidence presented in this study will establish the efficacy of non-contrast MRI as a surveillance tool for HCC in high-risk patients. The date of trial registration ( NCT02551250 ) in this study was September 15, 2015, and follow-up is still ongoing.