Accurate diagnosis of thyroid nodules using ultrasonography is a highly valuable, but challenging task. With the emergence of artificial intelligence, deep learning based methods can provide assistance to radiologists, whose performance heavily depends on the quantity and quality of training data, but current ultrasound image datasets for thyroid nodule either directly utilize the TI-RADS assessments as labels or are not publicly available. Faced with these issues, an open-access ultrasound image dataset for thyroid nodule detection and classification is proposed, i.e. the TN5000, which comprises 5,000 B-mode ultrasound images of thyroid nodule, as well as complete annotations and biopsy confirmations by expert radiologists. Additionally, the statistical characteristics of this proposed dataset have been analyzed clearly, some baseline methods for the detection and classification of thyroid nodules are recommended as the benchmark, along with their evaluation results. To our best knowledge, TN5000 is the largest open-access ultrasound image dataset of thyroid nodule with professional labeling, and is the first ultrasound image dataset designed both for the thyroid nodule detection and classification. These kinds of images with annotations can contribute to analyze the intrinsic properties of thyroid nodules and to determine the necessity of FNA biopsy, which are crucial in ultrasound diagnosis.

Ultrasonography is a valuable diagnosis method for thyroid nodules. Automatically discriminating benign and malignant nodules in the ultrasound images can provide aided diagnosis suggestions, or increase the diagnosis accuracy when lack of experts. The core problem in this issue is how to capture appropriate features for this specific task. Here, we propose a feature extraction method for ultrasound images based on the convolution neural networks (CNNs), try to introduce more meaningful and specific features to the classification. A CNN model trained with ImageNet data is transferred to the ultrasound image domain, to generate semantic deep features under small sample condition. Then, we combine those deep features with conventional features such as Histogram of Oriented Gradient (HOG) and Scale Invariant Feature Transform (SIFT) together to form a hybrid feature space. Furthermore, to make the general deep features more pertinent to our problem, a feature subset selection process is employed for the hybrid nodule classification, followed by a detailed discussion on the influence of feature number and feature composition method. Experimental results on 1037 images show that the accuracy of our proposed method is 0.929, which outperforms other relative methods by over 10%.

Background: B7-H3, also known as CD276, an important immune checkpoint member of the B7-CD28 family, is confirmed as a promising target after PD-L1 in clinical trials. Although the overexpression of B7-H3 has been associated with invasive metastatic potential and poor prognosis in multiple types of cancer, nothing is known regarding the expression profiles of B7-H3 in papillary thyroid carcinoma (PTC). In this study, we carried out a large-scale analysis of B7-H3 expression in PTC patients and evaluated the potential clinical significance of B7-H3.Methods: In total, data from 1,210 samples, including 867 cases from TCGA and four GEO datasets, were collected for B7-H3–related transcriptome analyses, and 343 postoperative, whole-tumor sections were collected from patients with PTC at our institute for B7-H3–specific immunohistochemistry (IHC) staining. The statistical analysis was primarily accomplished using the R project for statistical computing.Results: B7-H3 positivity was found in 84.8% of PTC patients (291/343), and the mRNA and protein expression levels of B7-H3 in PTC were markedly higher than those of para-tumor tissues (p < 0.001), demonstrating that B7-H3 can serve as a potential diagnostic biomarker for PTC. The significant upregulation of B7-H3 in PTC is caused by distinct patterns of CNVs and CpG DNA methylation. Functional enrichment analysis confirmed that high B7-H3 expression was significantly associated with specific immune features and angiogenesis. High B7-H3 protein expression was associated with tumor size (p = 0.022), extrathyroidal extension (ETE) (p = 0.003), and lymph node metastasis (LNM) (p < 0.001). More importantly, multivariate analysis confirmed that B7-H3 was an independent predictor of relapse-free survival (RFS) (p < 0.05). In the subgroup analysis, positive B7-H3 staining was associated with worse RFS in patients with primary tumor size ≥2 cm (p < 0.05), age ≥55 years (p < 0.05), LNM (p = 0.07), multifocality (p < 0.05), and ETE (p < 0.05). In addition, Circos plots indicated that B7-H3 was significantly associated with other immune checkpoints in the B7-CD28 family.Conclusion: This is the first comprehensive study to elucidate the expression profile of B7-H3 in PTC. Our observations revealed that B7-H3 is a novel independent biomarker for predicting LNM and disease recurrence for PTC patients, and it thus may serve as an indicator that could be used to improve risk-adapted therapeutic strategies and a novel target for immunotherapy strategies for patients who undergo an aggressive disease course.