Domain adaptation via Wasserstein distance and discrepancy metric for chest X-ray image classification.

Deep learning technology can effectively assist physicians in diagnosing chest radiographs. Conventional domain adaptation methods suffer from inaccurate lesion region localization, large errors in feature extraction, and a large number of model parameters. To address these problems, we propose a novel domain-adaptive method WDDM to achieve abnormal identification of chest radiographic images by combining Wasserstein distance and difference measures. Specifically, our method uses BiFormer as a multi-scale feature extractor to extract deep feature representations of data samples, which focuses more on discriminant features than convolutional neural networks and Swin Transformer. In addition, based on the loss minimization of Wasserstein distance and contrast domain differences, the source domain samples closest to the target domain are selected to achieve similarity and dissimilarity across domains. Experimental results show that compared with the non-transfer method that directly uses the network trained in the source domain to classify the target domain, our method has an average AUC increase of 14.8% and above. In short, our method achieves higher classification accuracy and better generalization performance.

Multi-Branch Attention Learning for Bone Age Assessment with Ambiguous Label.

Bone age assessment (BAA) is a typical clinical technique for diagnosing endocrine and metabolic diseases in children's development. Existing deep learning-based automatic BAA models are trained on the Radiological Society of North America dataset (RSNA) from Western populations. However, due to the difference in developmental process and BAA standards between Eastern and Western children, these models cannot be applied to bone age prediction in Eastern populations. To address this issue, this paper collects a bone age dataset based on the East Asian populations for model training. Nevertheless, it is laborious and difficult to obtain enough X-ray images with accurate labels. In this paper, we employ ambiguous labels from radiology reports and transform them into Gaussian distribution labels of different amplitudes. Furthermore, we propose multi-branch attention learning with ambiguous labels network (MAAL-Net). MAAL-Net consists of a hand object location module and an attention part extraction module to discover the informative regions of interest (ROIs) based only on image-level labels. Extensive experiments on both the RSNA dataset and the China Bone Age (CNBA) dataset demonstrate that our method achieves competitive results with the state-of-the-arts, and performs on par with experienced physicians in children's BAA tasks.

Knockdown of circ_NEK6 Decreased 131I Resistance of Differentiated Thyroid Carcinoma via Regulating miR-370-3p/MYH9 Axis.

Radioresistance is a crucial factor for the failure of iodine 131 (131I)-based radiotherapy for differentiated thyroid carcinoma (DTC). This study aimed to explore the effect of circ_NEK6 on the development of 131I resistance in DTC and its potential mechanism. In this study, we demonstrated that circ_NEK6 expression was significantly elevated in 131I-resistant DTC tissues and cell lines. Knockdown of circ_NEK6 significantly repressed 131I resistance via inhibiting cell proliferation, migration, invasion abilities, and inducing cell apoptosis and DNA damage in 131I-resistant DTC cells. Mechanistically, knockdown of circ_NEK6 suppressed 131I resistance in DTC by upregulating the inhibitory effect of miR-370-3p on the expression of myosin heavy chain 9 (MYH9). In vivo experiments showed that circ_NEK6 inhibition aggravated 131I radiation-induced inhibition of xenograft tumor growth. Taken together, knockdown of circ_NEK6 repressed 131I resistance in DTC cells by regulating the miR-370-3p/MYH9 axis, indicating that circ_NEK6 may act as a potential biomarker and therapeutic target for DTC patients with 131I resistance.

Aptamer recognition and proximity-induced entropy-driven circuit for enzyme-free and rapid amplified detection of platelet-derived growth factor-BB.

Platelet-derived growth factor-BB (PDGF-BB) is currently used as a biomarker protein for cancer early diagnosis and clinical treatment. Herein, we reported a robust and enzyme-free strategy based on aptamer recognition and proximity-induced entropy-driven circuits (AR-PEDC) for homogeneous and rapid detection of platelet-derived growth factor BB (PDGF-BB) without any washing steps or thermocycling. The proximity probes specifically recognize target protein to form the completed trigger (CT). Then, the CT reacts with three-strand complex to form intermediate, which subsequently binds to fuel strand to release reporter strand, assistant strand and the CT. The revised proximity probes exhibit significantly improved signal-to-background ratio and faster association rate. Moreover, target protein/proximity probes interaction can specifically initiate entropy-driven circuits, thus providing immense signal amplification for ultrasensitive detection of PDGF-BB with low detection limit of 9.6 pM. The practical ability of the developed strategy is demonstrated by detection of PDGF-BB in human serum with satisfactory results. In addition, this method is flexible and can be conveniently extended to a variety of targets by simply substituting the target specific sequence. Thus, this strategy presents a rapid, low background and versatile amplification mechanism for the detection of protein biomarkers and offers a promising alternative platform for clinical diagnosis.

[Papillon-Lefèvre syndrome: a case report].

Papillon-Lefèvre syndrome (PLS) is a very rare syndrome of autosomal recessive inheritance characterized by palmoplantar hyperkeratosis and a severe destructive periodontitis, leading to premature loss of both primary and permanent dentitions. This article reported a boy who was diagnosed as having PLS.