Evaluation of a deep learning-based computer-aided diagnosis system for distinguishing benign from malignant thyroid nodules in ultrasound images.

PURPOSE: Computer-aided diagnosis (CAD) systems assist in solving subjective diagnosis problems that typically rely on personal experience. A CAD system has been developed to differentiate malignant thyroid nodules from benign thyroid nodules in ultrasound images based on deep learning methods. The diagnostic performance was compared between the CAD system and the experienced attending radiologists. METHODS: The ultrasound image dataset for training the CAD system included 651 malignant nodules and 386 benign nodules while the database for testing included 422 malignant nodules and 128 benign nodules. All the nodules were confirmed by pathology results. In the proposed CAD system, a support vector machine (SVM) is used for classification and fused features which combined the deep features extracted by a convolutional neural network (CNN) with the hand-crafted features such as the histogram of oriented gradient (HOG), local binary patterns (LBP), and scale invariant feature transform (SIFT) were obtained. The optimal feature subset was formed by selecting these fused features based on the maximum class separation distance and used as the training sample for the SVM. RESULTS: The accuracy, sensitivity, and specificity of the CAD system were 92.5%, 96.4%, and 83.1%, respectively, which were higher than those of the experienced attending radiologists. The areas under the ROC curves of the CAD system and the attending radiologists were 0.881 and 0.819, respectively. CONCLUSIONS: The CAD system for thyroid nodules exhibited a better diagnostic performance than experienced attending radiologists. The CAD system could be a reliable supplementary tool to diagnose thyroid nodules using ultrasonography. Macroscopic features in ultrasound images, such as the margins and shape of thyroid nodules, could influence the diagnostic efficiency of the CAD system.

Comprehensive evaluation of medullary thyroid carcinoma before surgery.

BACKGROUND: Medullary thyroid carcinoma (MTC) is a rare disease, but it exhibits more aggressive behaviors. The aim of this study was to improve the diagnostic accuracy of MTC before surgery by analyzing the clinical and ultrasonic data of patients with MTC. METHODS: The study included 71 patients (96 lesions) with histopathologically proven MTC between April 2011 and September 2016 in the Cancer Hospital of Chinese Academy of Medical Sciences and Peking Union Medical College. The clinical characteristics and sonographic findings were retrospectively reviewed and compared between the ultrasonic correct diagnosis group and the ultrasonic misdiagnosis group with the t test or Mann-Whitney U test for quantitative parameters and the χ test or Fisher exact test for qualitative parameters. RESULTS: Compared with the ultrasonic correct diagnosis group, the proportion of the cystic change in the ultrasonic misdiagnosed group was high (25.0% vs. 4.2%), the uncircumscribed margin and irregular shape proportions were low (20.8%, 58.3% vs. 74.7%, 87.3%), calcification was relatively rare (20.8% vs. 56.3%), and rich vascularity was relatively rare (25.0% vs. 78.9%). CONCLUSIONS: In the case of atypical MTC, such as cystic change, circumscribed margin, regular shape, no calcification, no rich vascularity, and normal cervical lymph nodes, MTC is easily misdiagnosed as benign by ultrasound. Therefore, ultrasound, cytology and serum calcitonin should be comprehensively evaluated for a preoperative diagnosis of MTC.

[Evaluation of the efficacy and the limitation of ultrasound-guided core-needle biopsy, core-needle aspiration and fine-needle aspiration in micro-nodules of thyroid].

OBJECTIVE: To evaluate the efficacy and the limitation of ultrasound-guided core-needle biopsy, ultrasound-guided core-needle aspiration and ultrasound-guided fine-needle aspiration in micro-nodules of thyroid. METHODS: A retrospective was performed in 92 patients with suspectable malignent micro-nodules in thyroid. Of them, 52 patients underwent US-CNB and US-FNA and 40 patients underwent US-CNA and US-FNA. The diagnoses for the micro-nodules were identified by histopathlogical examination after surgery. RESULT: Among 52 cases with both US-CNB and US-FNA, 41 got nondiagnostic US-CNB and 11 cases successfully got the correct diagnoses of US-CNB; 6 cases got the incorrect diagnosis of US-FNA and 46 cases got the correct diagnosis of US-FNA. Of 40 cases with US-CNA and US-FNA, unsatisfactory specimen of US-CNA occurred in 14 cases and satisfactory specimen of US-CNA were got in 26 cases; unsatisfactory specimen of US-FNA occurred in 4 cases and satisfactory specimen of US-FNA. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of US-FNA in 92 cases for the diagnosis of malignancy were 93.4%, 86.7%, 97.3%, 72.2% and 92.3%, respectively. CONCLUSIONS: US-FNA is the most valuable method for the diagnosis of suspectable malignent micro-nodules in thyroid before operation.