A self-supervised classification model for endometrial diseases.

PURPOSE: Ultrasound imaging is the preferred method for the early diagnosis of endometrial diseases because of its non-invasive nature, low cost, and real-time imaging features. However, the accurate evaluation of ultrasound images relies heavily on the experience of radiologist. Therefore, a stable and objective computer-aided diagnostic model is crucial to assist radiologists in diagnosing endometrial lesions. METHODS: Transvaginal ultrasound images were collected from multiple hospitals in Quzhou city, Zhejiang province. The dataset comprised 1875 images from 734 patients, including cases of endometrial polyps, hyperplasia, and cancer. Here, we proposed a based self-supervised endometrial disease classification model (BSEM) that learns a joint unified task (raw and self-supervised tasks) and applies self-distillation techniques and ensemble strategies to aid doctors in diagnosing endometrial diseases. RESULTS: The performance of BSEM was evaluated using fivefold cross-validation. The experimental results indicated that the BSEM model achieved satisfactory performance across indicators, with scores of 75.1%, 87.3%, 76.5%, 73.4%, and 74.1% for accuracy, area under the curve, precision, recall, and F1 score, respectively. Furthermore, compared to the baseline models ResNet, DenseNet, VGGNet, ConvNeXt, VIT, and CMT, the BSEM model enhanced accuracy, area under the curve, precision, recall, and F1 score in 3.3-7.9%, 3.2-7.3%, 3.9-8.5%, 3.1-8.5%, and 3.3-9.0%, respectively. CONCLUSION: The BSEM model is an auxiliary diagnostic tool for the early detection of endometrial diseases revealed by ultrasound and helps radiologists to be accurate and efficient while screening for precancerous endometrial lesions.

Evaluation of Tumor Resection Effect of Color Doppler Ultrasound Positioning Guided Breast-Conserving Surgery Using Nano-Contrast Agent.

This study aimed to explore the clinical value of Fe3O4-based magnetic lipid nano-contrast agent in breast-conserving surgery for breast cancer using color doppler ultrasound positioning and to analyze the tumor resection effect of breast-conserving surgery. On account of Fe3O4 magnetic nanoparticles prepared by the chemical co-precipitation method, magnetic lipid ultrasonic microbubbles (MLU-MBs) were prepared by mechanical oscillation method after surface modification using polyethylene terephthalate (PET). Characterization and analysis of the prepared MLU-MBs were performed using scanning electron microscopy (SEM), energy spectrometer, transmission electron microscopy (TEM), and Fourier infrared spectroscopy (FIRS). A high-frequency alternating magnetic field was used to detect the heating of MLU-MBs and the color ultrasound machine was applied to observe the enhancement effect of the MLU-MBs on rabbit liver images. 92 patients undergoing breast-conserving surgery for breast cancer were taken as the research objects and were divided into a nano group (MLU-MB as contrast agent) and a control group (conventional contrast agent) according to the differences of intraoperative contrast agents, with 46 cases in each group. Before the surgery, both groups of patients were positioned and marked the tumor boundary under ultrasound. The differences in tumor volume (TV), amount of tissue removed, resection rate, and positive rate (PR) of resection margins were compared between the two groups. The results showed that the Fe3O4 magnetic nanoparticles were particles with an average particle size of about 15nm, and their iron and oxygen percentages were consistent with the content of Fe3O4. The MLU-MBs were spherical particles of about 1120nm, containing phosphorus (P), oxygen(O), and Ferrum (Fe). Under 30A and 220kHz of output current and frequency, the temperature rise of the MLU-MBs suspension with different concentrations was 10 ~ 60°C, and the temperature was constant after heating for 45 minutes. Compared with the rabbit liver parenchyma, the image was greatly enhanced. TV, the amount of tissue removed, the resection rate, and the PR of resection margins in the nano group were obviously lower than those in the control group (P<0.05). It showed that MLU-MBs with good image enhancement effect on account of Fe3O4 magnetic nanoparticles were successfully prepared and it could effectively reduce the PR of normal tissues and the positive margin of two-fold resection during breast-conserving surgery for breast cancer and showed good accuracy and stability.