Sang Il Choi

Background: High false-positive rates remain a significant challenge in the automated detection of pulmonary embolism (PE) using Computed Tomography Pulmonary Angiography (CTPA). This study evaluated the additional value of a False-Positive Reduction (FPR) module integrated into a Region Proposal Network (RPN). Methods: A retrospective analysis of 303 CTPA scans (163 PE-positive and 140 PE-negative) was conducted from a single tertiary institution. Both models were additionally validated on an independent external cohort of 100 CTPA scans (50 PE-positive and 50 PE-negative) from the RSNA PE Challenge dataset. The diagnostic performance of the one-stage RPN-only model was compared with that of a two-stage Modified Mask R-CNN (Region-based Convolutional Neural Network) incorporating the FPR module. Results: The Modified Mask R-CNN exhibited significant improvement in terms of specificity. The false-positive rate per scan decreased by 31% in comparison to the RPN-only model. Although there was a slight reduction in patient-level sensitivity, the Positive Predictive Value significantly increased by 10.5%. Additionally, patient-level specificity for emboli with a volume ≥ 1000 mm³ increased, reflecting a 7.4% relative improvement in detecting clinically significant emboli. Conclusions: The Modified Mask R-CNN significantly reduced false positives while maintaining high sensitivity over a region proposal network.

Diffuse sclerosing variant papillary thyroid carcinoma (DSV-PTC) is a rare and aggressive subtype of PTC. Preoperative diagnosis is challenging owing to the overlapping of imaging characteristics with those of other thyroid conditions. This study aimed to evaluate the accuracy of preoperative ultrasound (US) in predicting DSV-PTC and to identify significant diagnostic factors. This retrospective study analyzed 34 patients who were preoperatively suspected of having DSV-PTC based on US findings and later underwent thyroidectomy. Patients were divided into 2 groups based on the final histopathological diagnosis: DSV-PTC and non-DSV-PTC. Demographic, radiological, and pathological characteristics were also compared. Only 32.4% of patients initially suspected of having DSV-PTC were confirmed postoperatively. Among the US features, the Korean Thyroid Imaging Reporting and Data System (K-TIRADS) score 5 was significantly associated with DSV-PTC (P = 0.038), whereas other radiological factors, including echogenicity and microcalcifications, were not. The histopathological features, such as tumor size, BRAF and TERT mutations, vascular invasion, and lymph node metastasis, showed no significant differences between the groups. Preoperative US has limited accuracy (32.4%) in diagnosing DSV-PTC. Because of the aggressive treatment recommendations based on preoperative suspicion, clinicians should carefully consider the limitations of imaging. Further studies incorporating fine-needle aspiration or core needle biopsy are required to improve diagnostic accuracy.

Choi SI. Cardiovasc Imaging Asia. 2025 Oct;9(4):39-40. https://doi.org/10.22468/cvia.2025.00122

Oh S, et al. Investig Magn Reson Imaging. 2025 Jun;29(2):84-95. https://doi.org/10.13104/imri.2025.0006

최상일, et al. J Korean Soc Radiol. 2025 May;86(3):381-388. https://doi.org/10.3348/jksr.2025.0017

Lim H, et al. Korean J Radiol. 2025 Apr;26:e53. https://doi.org/10.3348/kjr.2025.0230

Jeong S, et al. Investig Magn Reson Imaging. 2025 Mar;29(1):51-58. https://doi.org/10.13104/imri.2025.0003

Jang S, et al. Cardiovasc Imaging Asia. 2025 Jan;9(1):9-14. https://doi.org/10.22468/cvia.2025.00017

Kim C, et al. Cardiovasc Imaging Asia. 2024 Apr;8(2):21-31. https://doi.org/10.22468/cvia.2024.00059