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.

HIF1α/PD-L1 axis mediates hypoxia-induced cell apoptosis and tumor progression in follicular thyroid carcinoma.

BACKGROUND: Hypoxia-inducible factor 1α (HIF-1α) and programmed cell death-1 protein ligand 1 (PD-L1) are implicated in the metastasis and progression processes of multiple cancers. Hypoxia selectively elevates PD-L1 expression via HIF1α activation in several solid tumors; however, the regulatory effect of HIF1α on PD-L1 in the pathogenesis of follicular thyroid cancer (FTC) remains unclear. This study aims to investigate the regulatory effect of HIF1α on PD-L1 and their potential roles in FTC pathogenesis. METHODS: Spearman correlation analysis was performed to clarify the relationships between HIF1α and PD-L1 expressions and the clinicopathologic characteristics. The expressions of HIF1α and PD-L1 at mRNA and protein levels were analyzed by qRT-PCR and Western blot. Hypoxia induction and cell transfection were conducted in FTC cells. TUNEL and Annexin V staining were used to detect the cell apoptosis. FTC xenograft tumor models were generated to evaluate the roles of HIF1α and PD-L1 in vivo. RESULTS: Here, we found that the expressions of HIF1α and PD-L1 were significantly increased in FTC tissues and were correlated with the FTC clinicopathologic features, such as the tumor size, T stage, TNM staging, and metastasis. In FTC cells, hypoxia-induced increased HIF1α and PD-L1 expression. Knockdown of HIF1α inhibits hypoxia-induced PD-L1 expression and cells apoptosis. Moreover, inhibition of HIF1α or PD-L1 significantly delays tumor growth and metastasis in vivo. CONCLUSION: Hypoxia could promote FTC progression by upregulating HIF1α and PD-L1, which could serve as the molecular targets for FTC treatment.