Objective: Ovarian cancer is the most deadly deadliest gynecological tumor in the female reproductive system. Therefore, the present study sought to determine the diagnostic performance of International Ovarian Tumor Analysis Simple Rules (IOTA SR), the Ovarian-Adnexal Reporting and Data System (O-RADS), and Cancer Antigen 125 (CA125) in discriminating benign and malignant ovarian tumors. The study also assessed whether a combination of the two ultrasound categories systems and CA125 can improve the diagnostic performance. Methods: A total of 453 patients diagnosed with ovarian tumors were retrospectively enrolled from Fujian Cancer Hospital between January 2017 and September 2020. The data collected from patients included age, maximum lesion diameter, location, histopathology, levels of CA125, and detailed ultrasound reports. Additionally, all ultrasound images were independently assessed by two ultrasound physicians with more than 5 years of experience in the field, according to the IOTA simple rules and O-RADS guidelines. Furthermore, the area under the curve (AUC), sensitivity, and specificity of the above mentioned predictors were calculated using the receiver operating characteristic curve. Results: Out of the 453 patients, 184 had benign lesions, while 269 had malignant ovarian tumors. In addition, the AUCs of IOTA SR, O-RADS, and CA125 in the overall population were 0.831, 0.804, and 0.812, respectively, and the sensitivities of IOTA SR, O-RADS, and CA125 were 94.42, 94.42, and 80.30%, respectively. On the other hand, the AUCs of IOTA SR combined with CA125, O-RADS combined with CA125, and IOTA SR plus O-RADS combined with CA125 were 0.900, 0.891, and 0.909, respectively. The findings also showed that the AUCs of a combination of the three approaches were significantly higher than those of individual strategies (p0.05). Conclusion: The findings showed that a combination of IOTA SR or O-RADS in combination with CA125 may improve the ability to distinguish benign from malignant ovarian tumors.

Purpose: This study aimed to investigate the value of shear wave elastography (SWE) for characterization of breast masses in a Chinese population. Patients and methods: Two thousand two hundred seventy-three women consented to be prospectively enrolled for the characterization of breast masses with ultrasound and SWE. Breast masses were known from symptoms, palpability, and/or previous imaging screening with mammography and/or ultrasound. Correlation of SWE qualitative and quantitative features with malignancy risk and impact on diagnostic performance of combining SWE features were assessed, and the Breast Imaging Reporting and Data System (BI-RADS) scoring was calculated using histopathology as reference. Results: Data of 2,262 masses (median size: 13 mm; range: 1.3–50) from 2,262 patients (median age: 43 years; range: 18–91) were investigated, of which 752 (33.3%) were malignant. Sensitivity and specificity of BI-RADS diagnosis were 97.5% (733/752) and 54.8% (827/1,510), respectively. By logistic regression, the combination of maximum elasticity (E_max) measurements with BI-RADS assessments increased the area under the receiver operating characteristic curve from 0.908 (95% CI: 0.896–0.920) to 0.954 (95% CI: 0.944–0.962). Using E_max of 30 kPa or lower to selectively downgrade BI-RADS 4a masses to follow-up, and E_max of 160 kPa or higher to selectively upgrade BI-RADS 3 lesions to biopsy, specificity significantly increased from 54.8% (827/1,510) to 66.1% (998/1,510) (P<0.001) while sensitivity decreased nonsignificantly from 97.5% (733/752) to 96.9% (729/752) (P=0.2891). Positive predictive value for biopsy recommendation increased from 51.7% (733/1,417) to 58.7% (729/1,241) (P<0.001). Conclusion: Adding SWE maximum stiffness to BI-RADS 3 and BI-RADS 4a breast masses in a Chinese population increased significantly the specificity of breast ultrasonography, without significant change in sensitivity.

Objectives. Evidence suggests that about 80 of all salivary gland tumors involve the parotid glands, with approximately 20 of parotid gland tumors (PGTs) being malignant. Discriminating benign and malignant parotid gland lesions preoperatively is vital for selecting the appropriate treatment strategy. This study explored the diagnostic performance of deep learning system for discriminating benign and malignant PGTs in ultrasonography images and compared it with radiologists. Methods. A total of 251 consecutive patients with surgical resection and proven parotid gland malignant or benign tumors who underwent preoperative ultrasound examinations were enrolled in this study between January 2014 and November 2020. Next, we compared the diagnostic accuracy of deep learning methods (ViT-B\16, EfficientNetB3, DenseNet121, and ResNet50) and radiologists in parotid gland tumor. In addition, the area under the curve (AUC), specificity, sensitivity, positive predictive value, and negative predictive value were calculated. Results. Among the 251 patients, 176/251 were the training set, whereas 75/251 were the validation set. Results showed that 74/251 patients had malignant tumor. Deep learning models achieved good performance in differentiating benign from malignant tumors, with the diagnostic accuracy and AUCs of ViT-B\16, EfficientNetB3, DenseNet121, and ResNet50 model being 81 and 0.81, 80 and 0.82, 77 and 0.81, and 79 and 0.80, respectively. On the other hand, the diagnostic accuracy and AUCs of radiologists were 77-81 and 0.68-0.75, respectively. It was evident that the diagnostic accuracy of deep learning methods was higher than that of inexperienced radiologists, but there was no significant difference between deep learning methods and experienced radiologists. Conclusions. This study shows that the deep learning system can be used for diagnosing parotid tumors. The findings also suggest that the deep learning system may improve the diagnosis performance of inexperienced radiologists.

Purpose: To develop a deep learning (DL) model for differentiating between benign and malignant ovarian tumors of Ovarian-Adnexal Reporting and Data System Ultrasound (O-RADS US) Category 4 lesions, and validate its diagnostic performance. Methods: A retrospective analysis of 1619 US images obtained from three centers from December 2014 to March 2023. DeepLabV3 and YOLOv8 were jointly used to segment, classify, and detect ovarian tumors. Precision and recall and area under the receiver operating characteristic curve (AUC) were employed to assess the model performance. Results: A total of 519 patients (including 269 benign and 250 malignant masses) were enrolled in the study. The number of women included in the training, validation, and test cohorts was 426, 46, and 47, respectively. The detection models exhibited an average precision of 98.68% (95% CI: 0.95–0.99) for benign masses and 96.23% (95% CI: 0.92–0.98) for malignant masses. Moreover, in the training set, the AUC was 0.96 (95% CI: 0.94–0.97), whereas in the validation set, the AUC was 0.93(95% CI: 0.89–0.94) and 0.95 (95% CI: 0.91–0.96) in the test set. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive values for the training set were 0.943,0.957,0.951,0.966, and 0.936, respectively, whereas those for the validation set were 0.905,0.935, 0.935,0.919, and 0.931, respectively. In addition, the sensitivity, specificity, accuracy, positive predictive value, and negative predictive value for the test set were 0.925, 0.955, 0.941, 0.956, and 0.927, respectively. Conclusion: The constructed DL model exhibited high diagnostic performance in distinguishing benign and malignant ovarian tumors in O-RADS US category 4 lesions.

Purpose: To evaluate a classification model of contrast-enhanced ultrasound (CEUS) and examine the characteristics of patients with false-negative diagnosis. Patients and methods: A retrospective secondary analysis of a multicenter trial of CEUS for breast cancer diagnosis (from August 2015 to April 2017) was undertaken. Patients (n=1,023) with Breast Imaging Reporting and Data System 4–5 lesions on B-mode ultrasound underwent CEUS. Pathological diagnoses were available from surgical or biopsy specimens for correlation. Lesion maximum diameter (LMD), distance to the papilla (DtP), distance from the superficial edge of the lesion to the skin (DtS), distance from the deep edge of the lesion to the pectoralis muscle (DtPM), and body mass index (BMI) were evaluated. Results: Median age and BMI were 48.0 and 41.2 years and 23.2 and 22.4 kg/m² for patients with malignant and benign lesions, respectively. Overall sensitivity, specificity, and accuracy of CEUS for malignancy were 89.4%, 65.3%, and 75.8%, respectively. The patients with true-positive and false-negative diagnosis (ie, with malignant lesion) were older than those with false-positive and true-negative diagnosis (ie, with benign lesion). Patients with true-positive and false-positive diagnoses had higher BMI than patients with true-negative and false-negative diagnoses (P=0.004). Patients with true-positive and false-negative diagnoses had larger LMD and DtP, as well as smaller DtS and DtPM. Conclusion: Older age, higher BMI, larger LMD and DtP, and smaller DtS and DtPM were associated with malignant lesions on CEUS. Patients with these characteristics should undergo further imaging.

Background: We aimed to investigate the influence of patient and lesion characteristics on our diagnostic model for contrast-enhanced ultrasound (CEUS) of the breast, comparing its accuracy with that of histopathology. Methods: Conducting a study with eight medical centers, we compared 1,023 breast lesions categorized as BI-RADS 4 or 5 with the score from our newly-established CEUS-based diagnostic model, comparing the results with pathological outcomes. Univariate and multivariate logistic regression analyses were conducted to determine the influence of clinicopathological characteristics on the performance of this CEUS model. Results: Logistic regression analysis showed that patients’ age, maximum lesion diameter, and distance from the lesion’s deep edge to the pectoralis major were significant independent influencing factors. The model’s diagnostic accuracy was greater for patients >35 y (P=0.005), for maximum lesion diameter >20 mm, and for distance from the lesion’s deep edge to the pectoralis major <3.05 mm. There was no significant difference in accuracy between lesions with maximum lesion diameter 10–20 and <10 mm (P=0.393). Conclusions: The diagnostic performance of the proposed CEUS model for breast lesions is influenced by patients’ age, maximum lesion diameter, and distance from the lesion’s deep edge to the pectoralis major. Consideration of influencing factors is required to optimize clinical use of the CEUS model.