Deep Learning-Based Prediction of Post-treatment Survival in Hepatocellular Carcinoma Patients Using Pre-treatment CT Images and Clinical Data
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.This study developed a model using CT images and clinical data to predict survival in hepatocellular carcinoma (HCC) patients. The multimodal approach significantly improved prediction accuracy compared to using single data types.
Area Of Science
- Medical Imaging
- Oncology
- Artificial Intelligence
Background
- Hepatocellular carcinoma (HCC) survival prediction is challenging.
- Accurate prognostication is crucial for guiding treatment decisions.
Purpose Of The Study
- To develop and validate a predictive model for post-treatment survival in HCC patients.
- To integrate contrast-enhanced CT images with clinical and treatment data.
- To evaluate the model's performance using internal and external validation cohorts.
Main Methods
- A cascaded 3D convolutional neural network (CNN) model was developed.
- The model utilized pre-treatment CT scans and multimodal clinical information.
- Performance was evaluated using concordance index (C-index), mC/D AUC, and mBS.
Main Results
- The multimodal model integrating CT images and clinical data outperformed models using single data types.
- The optimized model achieved a C-index of 0.824 on the internal test cohort and 0.750 on the external cohort.
- The model demonstrated strong predictive capabilities for HCC patient survival.
Conclusions
- A CNN-based discrete-time survival prediction model integrating CT images and clinical information shows promise for HCC.
- This approach offers a valuable tool for predicting post-treatment survival in HCC patients.
- Multimodal data integration enhances the accuracy of survival prediction in oncology.
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
