Clinical Report: A Radiomic Nomogram Utilizing Deep Learning from Visceral Fat

Overview

This study presents a deep-learning-based radiomics nomogram (DLRN) for preoperative risk stratification of gastrointestinal stromal tumors (GISTs) using visceral fat features from non-contrast CT scans.

Background

Gastrointestinal stromal tumors (GISTs) are rare but exhibit significant heterogeneity, necessitating reliable preoperative risk stratification for effective treatment planning. Traditional methods rely on postoperative indicators, which can be limited by specimen availability. This study explores the potential of using non-contrast CT imaging to assess visceral fat as a predictive tool for GIST risk levels.

Data Highlights

Key Findings

• The DLRN achieved an AUC of 0.936 in the derivation cohort and 0.862 in the external test cohort.
• Accuracy rates were 0.873 in the derivation cohort and 0.925 in the external test cohort.
• Sensitivity and specificity were 0.811 and 0.904 in the derivation cohort, respectively.

Clinical Implications

The DLRN offers a non-invasive method for preoperative risk stratification of GISTs.

Conclusion

The DLRN integrating visceral fat-derived features may enhance preoperative risk assessment of GISTs.

Related Resources & Content

1. British Sarcoma Group, Gastrointestinal stromal tumour (GIST): clinical practice guidelines, 2024/2025 -- Guidelines for GIST management
2. PMC, Placebo-Controlled Randomized Trial of Adjuvant Imatinib Mesylate Following the Resection of Localized, Primary Gastrointestinal Stromal Tumor (GIST), 2024 -- Study on adjuvant therapy
3. ScienceDirect, Diagnostic accuracy of radiomics in risk stratification of gastrointestinal stromal tumors: A systematic review and meta-analysis, 2025 -- Review on radiomics in GIST
4. Frontiers in Oncology — Development and validation of a multiparametric MRI-based radiomics nomogram for the tripartite discrimination of primary benign, primary malignant, and metastatic lumbar spinal tumors
5. Frontiers in Oncology — A deep learning and radiomics fusion model enhances endoscopic ultrasonography diagnosis of gastric tumors
6. npj Digital Medicine — Digital Multimodal Biopsy for Predicting Occult Peritoneal Metastasis Prior to Surgery in Gastric Cancer Patients
7. Integrating Transrectal Ultrasound with a Radiomics Approach to Assess Neoadjuvant Chemoradiotherapy Outcomes in Locally Advanced Rectal Cancer
8. Development and validation of a multiparametric MRI-based radiomics nomogram for the tripartite discrimination of primary benign, primary malignant, and metastatic lumbar spinal tumors
9. A deep learning and radiomics fusion model enhances endoscopic ultrasonography diagnosis of gastric tumors
10. Digital Multimodal Biopsy for Predicting Occult Peritoneal Metastasis Prior to Surgery in Gastric Cancer Patients
11. Gastrointestinal stromal tumour (GIST): British Sarcoma Group clinical practice guidelines
12. Placebo-Controlled Randomized Trial of Adjuvant Imatinib Mesylate Following the Resection of Localized, Primary Gastrointestinal Stromal Tumor (GIST) - PMC
13. Diagnostic accuracy of radiomics in risk stratification of gastrointestinal stromal tumors: A systematic review and meta-analysis - ScienceDirect