MRI-based deep learning model for early TACE response prediction in HCC: multicenter validation with biological insights
Bibliographic record
Abstract
BACKGROUND: Transarterial chemoembolization (TACE) remains a cornerstone treatment for hepatocellular carcinoma (HCC), yet heterogeneous treatment response poses significant clinical challenges. This multicenter study aimed to develop and validate a deep learning model that leverages pretreatment MRI to predict objective response to initial TACE, while exploring imaging-biological correlations. METHODS: We utilized retrospective data from 3 institutions, which included HCC patients who underwent TACE. A deep learning algorithm (hereinafter, DLTR) was developed for predicting TACE response by comparing various deep learning algorithms. A multilayer perceptron was then employed to integrate potential clinical factors into the model (hereinafter, DLTRMLP) classifier. Performance was evaluated by the area under the receiver operating characteristic curve (AUC) in internal and external cohorts. Survival differences were assessed using log-rank test in two external test sets. RNA-sequencing data from the Cancer Image Archive (TCIA) were used to link imaging signatures to biological pathways. RESULTS: DLTRMLP achieved higher AUC than DLTR and clinical models in predicting TACE efficacy in two external test cohorts (AUC: 0.8 vs. 0.649, 0.648; 0.818 vs. 0.629, 0.659) and effectively stratified patients by progression-free survival (P = 0.035). Deep learning features correlated with 149 genes (P < 0.05), which were notably enriched in angiogenesis, EMT, hypoxia, and TGF-β Signalling pathways. CONCLUSION: The DLTRMLP model, combining MRI-based deep learning and clinical variables, robustly predicts TACE response and reveals imaging signatures linked to tumour proliferation biology. Its potential integration into MRI workflows could help optimize treatment decision-making for HCC.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".