PubMed:35207736 JSONTXT

{"project":"Test-merged","denotations":[{"id":"370","span":{"begin":107,"end":110},"obj":"Disease"},{"id":"371","span":{"begin":111,"end":119},"obj":"Species"},{"id":"372","span":{"begin":85,"end":90},"obj":"Disease"},{"id":"373","span":{"begin":414,"end":422},"obj":"Species"},{"id":"374","span":{"begin":326,"end":334},"obj":"Species"},{"id":"375","span":{"begin":1176,"end":1181},"obj":"Disease"},{"id":"376","span":{"begin":822,"end":827},"obj":"Disease"},{"id":"377","span":{"begin":321,"end":324},"obj":"Disease"},{"id":"378","span":{"begin":272,"end":277},"obj":"Disease"},{"id":"379","span":{"begin":1299,"end":1307},"obj":"Species"},{"id":"380","span":{"begin":295,"end":319},"obj":"Disease"},{"id":"457","span":{"begin":1176,"end":1181},"obj":"Disease"},{"id":"479","span":{"begin":85,"end":90},"obj":"Disease"},{"id":"554","span":{"begin":272,"end":277},"obj":"Disease"}],"attributes":[{"id":"A370","pred":"resolved_to","subj":"370","obj":"MESH:D006528"},{"id":"A371","pred":"resolved_to","subj":"371","obj":"9606"},{"id":"A372","pred":"resolved_to","subj":"372","obj":"MESH:D009369"},{"id":"A373","pred":"resolved_to","subj":"373","obj":"9606"},{"id":"A374","pred":"resolved_to","subj":"374","obj":"9606"},{"id":"A375","pred":"resolved_to","subj":"375","obj":"MESH:D009369"},{"id":"A376","pred":"resolved_to","subj":"376","obj":"MESH:D009369"},{"id":"A377","pred":"resolved_to","subj":"377","obj":"MESH:D006528"},{"id":"A378","pred":"resolved_to","subj":"378","obj":"MESH:D009369"},{"id":"A379","pred":"resolved_to","subj":"379","obj":"9606"},{"id":"A380","pred":"resolved_to","subj":"380","obj":"MESH:D006528"},{"id":"A457","pred":"resolved_to","subj":"457","obj":"MESH:D009369"},{"id":"A479","pred":"resolved_to","subj":"479","obj":"MESH:D009369"},{"id":"A554","pred":"resolved_to","subj":"554","obj":"MESH:D009369"}],"text":"Multi-Task Deep Learning Approach for Simultaneous Objective Response Prediction and Tumor Segmentation in HCC Patients with Transarterial Chemoembolization.\nThis study aimed to develop a deep learning-based model to simultaneously perform the objective response (OR) and tumor segmentation for hepatocellular carcinoma (HCC) patients who underwent transarterial chemoembolization (TACE) treatment. A total of 248 patients from two hospitals were retrospectively included and divided into the training, internal validation, and external testing cohort. A network consisting of an encoder pathway, a prediction pathway, and a segmentation pathway was developed, and named multi-DL (multi-task deep learning), using contrast-enhanced CT images as input. We compared multi-DL with other deep learning-based OR prediction and tumor segmentation methods to explore the incremental value of introducing the interconnected task into a unified network. Additionally, the clinical model was developed using multivariate logistic regression to predict OR. Results showed that multi-DL could achieve the highest AUC of 0.871 in OR prediction and the highest dice coefficient of 73.6% in tumor segmentation. Furthermore, multi-DL can successfully perform the risk stratification that the low-risk and high-risk patients showed a significant difference in survival (p = 0.006). In conclusion, the proposed method may provide a useful tool for therapeutic regime selection in clinical practice."}