Key result
Dual-branch deep learning outperforms baseline with ~0.93 AUC for pulmonary embolism detection on low-dose CTPA.
Why the study?
Conventional CTPA protocols for diagnosing PE require relatively high iodine contrast and radiation doses, raising concerns about renal injury and radiation exposure.
Does a deep learning-based framework improve diagnostic performance for pulmonary embolism under low-iodine and low-radiation CTPA conditions compared to single-branch baselines?
Observational (n=191)
No
Does a deep learning-based framework improve diagnostic performance for pulmonary embolism under low-iodine and low-radiation CTPA conditions compared to single-branch baselines?
Absolute Event Rate: 0.928% vs 0.814%
p-value: p=<0.001
A deep learning framework integrating image enhancement and classification enables accurate pulmonary embolism diagnosis on low-iodine, low-radiation CTPA with an AUC of 0.928.
May enable low-iodine low-radiation CTPA for PE diagnosis; extends deep learning beyond single-branch baselines.
Pulmonary embolism (PE) is a life-threatening condition for which computed tomography pulmonary angiography (CTPA) is the standard diagnostic modality. However, conventional CTPA protocols require relatively high iodine contrast and radiation doses, raising concerns about renal injury and radiation exposure. In this study, we propose a deep learning-based framework for PE diagnosis under low-iodine and low-radiation CTPA conditions. The proposed two-stage framework integrates image enhancement and classification by jointly leveraging original low-exposure images and their super-resolved counterparts. We further construct and publicly release a low-iodine, low-radiation CTPA dataset developed in collaboration with a clinical institution to support reproducible research in safe imaging. Experimental results demonstrate that the proposed method substantially improves diagnostic performance compared with single-branch baselines, achieving an area under the ROC curve (AUC) of 0.928 while maintaining balanced sensitivity and specificity. These findings suggest that the proposed framework enables accurate and safer PE diagnosis under reduced contrast and radiation exposure, offering a practical solution for improving diagnostic safety in clinical CTPA imaging.
No takes yet. Share an insight, caveat, or question.
Hong et al. (2026) conducted an observational in Pulmonary embolism (n=191). Dual-Branch Classification Network (DBCN) with Frequency-Aware Super-Resolution Network (FASRN) vs. Baseline model using original low-exposure images was evaluated on Area under the ROC curve (AUC) for pulmonary embolism classification (95% CI 0.904-0.949, p=<0.001). The dual-branch deep learning framework achieved an AUC of 0.928 for pulmonary embolism detection under low-iodine and low-radiation CTPA conditions, significantly outperforming the baseline model.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: