Deep learning model for assessing survival benefits in hepatocellular carcinoma patients undergoing intra-arterial therapies based on proliferative subtype - Summary - MDSpire
Advertisement
Deep learning model for assessing survival benefits in hepatocellular carcinoma patients undergoing intra-arterial therapies based on proliferative subtype
To develop a multitask deep learning system for detecting proliferative hepatocellular carcinoma (HCC) and predicting survival outcomes after intra-arterial therapy (IAT).
Key Findings:
The model achieved AUCs of 0.825 and 0.792 for detecting proliferative HCC in training and testing sets, respectively, with 95% confidence intervals.
Prognostic nomograms combining radiomic and clinical variables outperformed traditional staging systems, with time-dependent AUCs ranging from 0.83 to 0.87 and an integrated Brier score of 0.12 versus 0.18–0.23 (all P < 0.001).
HAIC showed a significant survival benefit compared to TACE for high-risk patients (P < 0.001).
Interpretation:
The non-invasive deep learning method allows for preoperative identification of proliferative HCC, supporting personalized treatment choices and potentially improving outcomes in unresectable HCC, which is crucial for patient management.
Limitations:
The study is retrospective and may be subject to selection bias, which could affect the generalizability of the findings.
The model's performance may vary across different populations and imaging protocols.
Conclusion:
This study demonstrates the potential of deep learning in enhancing the diagnosis and treatment personalization for patients with proliferative HCC, which may lead to improved survival outcomes.
