Objective:To explore the feasibility of deep learning-based restoration of obscured thyroid ultrasound images.Methods:A total of 358 images of thyroid nodules were retropectively collected from January 2020 to October 2021 at Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, and the images were randomly masked and restored using DeepFillv2. The difference in grey values between the images before and after restoration was compared, and 6 sonographers (2 chief physicians, 2 attending physicians, 2 residents) were invited to compare the rate of correctness of judgement and detection of image discrepancies. The ultrasound features of thyroid nodules (solid composition, microcalcifications, markedly hypoechoic, ill-defined or irregular margins, or extrathyroidal extensions, vertical orientation and comet-tail artifact) were extracted according to the Chinese Thyroid Imaging Reporting and Data System (C-TIRADS). The consistency of ultrasound features of thyroid nodules before and after restoration were compared.Results:The mean squared error of the images before and after restoration ranged from 0.274 to 0.522, and there were significant differences in the rate of correctness of judgement and detection of image discrepancies between physicians of different groups(all P<0.001). The overall accuracy rate was 51.95%, the overall detection rate was 1.79%, there were significant differences also within the chief physicians and resident groups (all P<0.001). The agreement rate of all ultrasound features of the nodules before and after image restoration was higher than 70%, over 90% agreement rate for features such as solid composition and comet-tail artifact. Conclusions:The algorithm can effectively repair obscured thyroid ultrasound images while preserving image features, which is expected to expand the deep learning image database, and promote the development of deep learning in the field of ultrasound images.

BACKGROUND:Several fetal cardiovascular structural defects may alter the hemodynamics of the cardiac chambers resulting in changes in chamber sizes. Quantitative measurements of the sizes of cardiac chambers can augment the diagnostic power of fetal echocardiography.AIMS:Using a new left atrial volume tracking (LAVT) method, time-left atrial volume curves (TLAVCs) can be automatically obtained. The goal of this study was to examine whether this method can be used to evaluate left atrial volume (LAV) and provide reference values for LAV and indices of left atrial function in normal human fetuses.METHODS:Two hundred and four normal human fetuses were enrolled. Using LAVT, the maximal left atrial volume (LAVmax) and minimal left atrial volume (LAVmin) were measured from TLAVCs. Left atrial ejection fraction (EF) was calculated. The maximal left atrial area (LAAmax) and minimal left atrial area (LAAmin) were measured using manual method tracing.RESULTS:Between 21 and 40 weeks, mean LAVmax increased from 0.27 ml to 4.15 ml, and mean LAVmin increased from 0.13 ml to 2.26 ml, respectively, while the EF remained stable at around 0.43. From 21 to 40 weeks, mean LAAmax increased from 0.61 cm2 to 2.64 cm2, and mean LAAmin increased from 0.34 cm2 to 1.53 cm2.CONCLUSIONS:This study establishes reference values for fetal LAV during the second half of gestation. The LAVT method appears to be feasible in estimating fetal LAV and shows potential for assessing left atrial function.

Objective To evaluate the correlation between ultrasound features of papillary thyroid carcinoma(PTC)and lymph node metastasis by preoperative ultrasound elemental observation of thyroid nodules.Methods Three hundred and seventy-six patients who underwent primary thyroid surgery and confirmed by ultrasound and pathological data as single-focal PTC from Jannary to December 2017 in Sir Run Run Shaw Hospital of Zhejiang Univbersity College of Medicine were retrospectively analyzed. According to the presence or absence of lymph node metastasis,they were divided into central and lateral lymph node metastasis group and non-metastasis group.Independent risk factors for central lymph node metastasis (CLNM) and lateral lymph node metastasis (LLNM) were analyzed byχ2 test and multivariate Logistic regression.Results Multivariate analysis showed that the posterior margin of the cancer was <0.25 cm from the posterior wall of the thyroid gland as an independent risk factor for CLNM(P=0.025). Compared with the tumor volume ≤0.38 cm3 ,the cancer volume >0.38 cm3(P=0.000),was more prone to CLNM.And multivariate analysis showed that the anterior margin of the cancer was <0.17 cm(P =0.006)from the anterior thyroid capsule and the inner wall of the foci was <0.26 cm (P =0.014) as independent risk factors for LLNM.Compared with the maximum diameter of the tumor lesion ≤1 cm,the maximum diameter >2 cm (P =0.001) group was more prone to LLNM.Compared with the tumor volume ≤0.38 cm3 ,the tumor volume >0.38 cm3(P =0.000)was more prone to LLNM.ConclusionsThe larger volume of single focal PTC carcinoma and the closer to the posterior thyroid capsule are independent risk factors for CLNM.The larger volume and diameter of single focal PTC,and the closer to the anterior and medial wall capsule are independent risk factors for LLNM.