Deep learning and radiomics-driven algorithm for automated identification of May-Thurner syndrome in Iliac CTV imaging

Abstract

Objective: This research aimed to create a dataset of Iliac CTV scans for automated May-Thurner syndrome (MTS) detection using deep learning and radiomics. In addition, it sought to establish an automated segmentation model for Iliac Vein CTV scans and construct a radiomic signature for MTS diagnosis.

Methods: We collected a dataset of 490 cases meeting specific inclusion and exclusion criteria, anonymized to comply with HIPAA regulations. Iliac Vein CTV scans were prepared with contrast agent administration, followed by image acquisition and evaluation. A deep learning-based segmentation model, UPerNet, was employed using 10-fold cross-validation. Radiomic features were extracted from the scans and used to construct a diagnostic radiomic signature. Statistical analysis, including Dice values and ROC analysis, was conducted to evaluate segmentation and diagnostic performance.

Results: The dataset consisted of 201 positive cases of MTS and 289 negative cases. The UPerNet segmentation model exhibited remarkable accuracy in identifying MTS regions. A Dice coefficient of 0.925 (95% confidence interval: 0.875-0.961) was observed, indicating the precision and reliability of our segmentation model. Radiomic analysis produced a diagnostic radiomic signature with significant clinical potential. ROC analysis demonstrated promising results, underscoring the efficacy of the developed model in distinguishing MTS cases. The radiomic signature demonstrated strong diagnostic capabilities for MTS. Within the training dataset, it attained a notable area under the curve (AUC) of 0.891, with a 95% confidence interval ranging from 0.825 to 0.956, showcasing its effectiveness. This diagnostic capability extended to the validation dataset, where the AUC remained strong at 0.892 (95% confidence interval: 0.793-0.991). These results highlight the accuracy of our segmentation model and the diagnostic value of our radiomic signature in identifying MTS cases.

Conclusion: This study presents a comprehensive approach to automate MTS detection from Iliac CTV scans, combining deep learning and radiomics. The results suggest the potential clinical utility of the developed model in diagnosing MTS, offering a non-invasive and efficient alternative to traditional methods.

Keywords: Computed Tomography Venography; Convolutional Neural Networks; May-Thurner syndrome; deep learning; iliac vein compression.

Figure 1

Representative CTV cases: (A) Included scan meeting criteria (complete contrast, no artifacts); (B,C) Excluded cases; (D) Post-stent exclusion.

Figure 2

(A) Represents a visual output obtained from Digital Subtraction Angiography (DSA) and also the gold standard for our experimental grouping, indicating that the patient’s iliac vein is almost occluded. (B,C) The axial and coronal images of CTV, respectively. (D) 3D reconstruction of the iliac vein.

Figure 3

The procedure for both training and validating the UPerNet model aimed at segmentation is outlined, subsequently leading to the development and assessment of a radiomic signature for the diagnosis of MTS. The terms used include LASSO, this refers to various methodologies utilized in statistical analysis and machine learning: LASSO, which denotes the least absolute shrinkage and selection operator; mRMR, signifying minimum redundancy maximum relevance; ROC, representing receiver operating characteristic; and SHAP, which is an acronym for SHapley Additive exPlanations.

Figure 4

(a) Illustrates the loss curve throughout the training process, while panel (b) depicts the variations in the Dice coefficient measured within the internal structures.

Figure 5

The features derived from radiomic analysis, along with their associated coefficients, were employed to establish the radiomic signature.

Figure 6

The evaluation of the diagnostic efficacy of the radiomic signature was conducted through a comparison of ROC curves, assessing both the training and external validation datasets. ROC refers to the receiver operating characteristic, while AUC signifies the area under the curve.

Figure 7

Summary plots utilizing SHapley Additive exPlanations (SHAP) were created for the radiomic signature to illustrate the significance of individual features as well as the aggregated contributions of these features to the overall diagnostic efficacy.