Objective: The risk factors associated with hematoma formation following ultrasound (US)-guided vacuum-assisted excision (US-VAE) of breast fibroadenoma (FA), particularly those related to the US features of the lesions, remain unclear. This study aimed to investigate the US features influencing hematoma formation after US-VAE of FAs. Materials and methods: This retrospective analysis included 137 patients with 145 FA nodules who underwent US-VAE at our hospital from January 2018 to August 2024, and subsequently underwent US follow-up to monitor hematoma formation and absorption. The US features of the lesions, including size, number of nodules removed, location, depth, distance from the nipple, peripheral tissue structure, and color Doppler flow features, were retrospectively analyzed to identify the potential risk factors associated with hematoma formation. Results: Hematomas formed in 78 nodules (53.8%) 24–48 h after the US-VAE procedure and disappeared within 6 months in 77 cases (98.7%). In the multivariate logistic regression analysis, FA depth ≥ 1.5 cm or <0.7 cm (OR = 2.490, 95% CI: 1.056–5.872; p < 0.05), surrounding tissue mainly composed of fat (OR = 2.471, 95% CI: 1.043–5.857; p < 0.05), and rich blood supply (OR = 3.164, 95% CI: 1.444–6.934; p < 0.01) were independent risk factors for hematoma formation. Conclusion: Most hematomas that formed after US-VAE for breast FA disappeared within 6 months. FA depth ≥ 1.5 cm or <0.7 cm, surrounding tissue mainly composed of fat, and rich blood supply indicated the possibility of hematoma formation after US-VAE.

Objectives: The preoperative differentiation of thyroid nodules into benign thyroid nodules (BTN), non-metastatic malignant thyroid nodules (NMTN), and metastatic malignant thyroid nodules (MMTN) is critical for guiding clinical management strategies. Ultrasound (US) examinations frequently exhibit diagnostic inconsistencies due to operator-dependent variability. Computer-assisted diagnosis (CAD), an artificial intelligence (AI) model based on convolutional neural networks (CNNs), can help overcome inconsistencies in US examination outcomes by leveraging large-scale ultrasound imaging datasets to improve classification accuracy. Our study aimed to establish and validate this AI-powered ultrasound diagnostic model for precise preoperative discrimination among BTN, NMTN, and MMTN. Methods: A total of 209 patients (BTN = 66, NMTN = 15, and MMTN = 128) were consecutively identified and enrolled from a multi-center database. A subset of 195 patients (BTN = 60, NMTN = 15, and MMTN = 120) was selected for final analysis. These patients were divided into two groups: a training set (BTN = 50, NMTN = 11, and MMTN = 100) and a testing set (BTN = 10, NMTN = 4, and MMTN = 20). A total of 3,537 ultrasound images from the 195 patients were preprocessed by normalizing grayscale values and reducing noise. The processed images were then input into the AI model, which was trained to classify thyroid nodules. The model’s performance was evaluated using the testing set and assessed through receiver operating characteristic (ROC) curve analysis and the confusion matrix. Finally, the diagnostic accuracy of the AI model was compared with that of radiologists to determine its clinical utility in ultrasound-based diagnosis. Results: Compared to junior and senior radiologists, the AI model achieved near-perfect AUC values of 0.97 (BTN), 0.99 (NMTN), and 0.96 (MMTN), significantly outperforming the senior radiologist’s AUCs (0.88 for NMTN) and the junior radiologist’s weaker discrimination. In addition, the accuracy of this model was higher than all ultrasound radiologists (95% vs. 73 and 84% for the junior radiologist and senior radiologist, respectively). Conclusion: The AI-based ultrasound imaging diagnostic model showed excellent performance in differentiating BTN, NMTN, and MMTN, supporting its value as a diagnostic tool for the clinical decision-making process.