Parathyroidectomy slows renal function decline in patients with primary hyperparathyroidism.

PURPOSE: Primary hyperparathyroidism has deleterious effects on health and causes nephrolithiasis and osteoporosis. However, it remains unclear whether parathyroidectomy benefits kidney function among patients with primary hyperparathyroidism. METHODS: In this retrospective study, patients with primary hyperparathyroidism receiving parathyroidectomy in a tertiary medical center between 2003 and 2017 were followed up until December 31 2017, death, or requiring renal replacement therapy. Impact of parathyroidectomy on kidney function was examined using longitudinal estimated glomerular filtration rate (eGFR) change scales: single, average, absolute difference, percent change, annual decline rate, and slope. We applied linear mixed-effect model to determine the effect of parathyroidectomy on kidney function. RESULTS: During study period, 167 patients with primary hyperparathyroidism were identified from 498 parathyroidectomized patients, and finally, 27 patients fulfilled our stringent criteria. Median follow-up duration was 1.50 years (interquartile range 1.05-1.81) before surgery and 2.47 years (1.37-6.43) after surgery. Although parathyroidectomy did not affect amount of proteinuria and distribution of eGFR, parathyroidectomy significantly slowed decline rate of eGFR compared with that before surgery (- 1.67 versus - 2.73 mL/min/1.73 m2/year, p < 0.001). More importantly, parathyroidectomy made more beneficial effects on kidney function in patients with age < 65 years and those without chronic kidney disease or hypertension. CONCLUSIONS: Our study showed that parathyroidectomy slows renal function decline irrespective of age or comorbidities, which offers novel insight into the revision of guidelines for surgical indications in primary hyperparathyroidism. Given small sample size, further large-scale controlled studies are warranted to confirm our findings.

Frequent loss of CD10 expression in follicular lymphoma with leukaemic presentation.

INTRODUCTION: Follicular lymphoma (FL) is usually a nodal lymphoma expressing CD10, rarely with leukaemic presentation (FL-LP). MATERIALS AND METHODS: We searched for FL-LP in our institution from 2000 to 2018 and characterised the neoplastic cells by flow cytometry, immunohistochemistry and fluorescence in situ hybridization. Thirteen (6.1%) of 212 FL cases were FL-LP, all de novo neoplasms. The leukaemic cells were small in 12 cases and large in one. All had concurrent FL, mostly (92%; 12/13) low-grade. The single case with large leukaemic cells had a concurrent primary splenic low-grade FL and a double-hit large B-cell lymphoma in the marrow. RESULTS: CD10 was expressed in the leukaemic cells in 38% (5/13) cases by flow cytometry and in 77% (10/13) cases in tumours (p= 0.0471). IGH/BCL2 reciprocal translocation was identified in 85% (11/13) cases. Most patients were treated with chemotherapy. In a median follow-up time of 36 months, nine patients were in complete remission. The 2- and 5-year survival rates were at 100% and 83%, respectively. In this study, we characterised a series of de novo FL-LP in Taiwan. All patients had concurrent nodal and/or tissue tumours, which might suggest that these patients seek medical help too late. CONCLUSION: The lower CD10 expression rate by flow cytometry than by immunohistochemistry might be due to different epitopes for these assays. Alternatively, loss of CD10 expression might play a role in the pathogenesis of leukaemic change. The clinical course of FL-LP could be aggressive, but a significant proportion of the patients obtained complete remission with chemotherapy.

Correction: A method to quantify mechanobiologic forces during zebrafish cardiac development using 4-D light sheet imaging and computational modeling.

[This corrects the article DOI: 10.1371/journal.pcbi.1005828.].

A method to quantify mechanobiologic forces during zebrafish cardiac development using 4-D light sheet imaging and computational modeling.

Blood flow and mechanical forces in the ventricle are implicated in cardiac development and trabeculation. However, the mechanisms of mechanotransduction remain elusive. This is due in part to the challenges associated with accurately quantifying mechanical forces in the developing heart. We present a novel computational framework to simulate cardiac hemodynamics in developing zebrafish embryos by coupling 4-D light sheet imaging with a stabilized finite element flow solver, and extract time-dependent mechanical stimuli data. We employ deformable image registration methods to segment the motion of the ventricle from high resolution 4-D light sheet image data. This results in a robust and efficient workflow, as segmentation need only be performed at one cardiac phase, while wall position in the other cardiac phases is found by image registration. Ventricular hemodynamics are then quantified by numerically solving the Navier-Stokes equations in the moving wall domain with our validated flow solver. We demonstrate the applicability of the workflow in wild type zebrafish and three treated fish types that disrupt trabeculation: (a) chemical treatment using AG1478, an ErbB2 signaling inhibitor that inhibits proliferation and differentiation of cardiac trabeculation; (b) injection of gata1a morpholino oligomer (gata1aMO) suppressing hematopoiesis and resulting in attenuated trabeculation; (c) weak-atriumm58 mutant (wea) with inhibited atrial contraction leading to a highly undeveloped ventricle and poor cardiac function. Our simulations reveal elevated wall shear stress (WSS) in wild type and AG1478 compared to gata1aMO and wea. High oscillatory shear index (OSI) in the grooves between trabeculae, compared to lower values on the ridges, in the wild type suggest oscillatory forces as a possible regulatory mechanism of cardiac trabeculation development. The framework has broad applicability for future cardiac developmental studies focused on quantitatively investigating the role of hemodynamic forces and mechanotransduction during morphogenesis.

Early improvement in PANSS-30, PANSS-8, and PANSS-6 scores predicts ultimate response and remission during acute treatment of schizophrenia.

OBJECTIVE: PANSS-8 and PANSS-6 are derived from the 30-item Positive and Negative Syndrome Scale (PANSS-30). We investigate whether PANSS-8 or PANSS-6 is a reliable, valid, sensitive to change measure, and scalable, and whether early improvement using them can predict response/remission. METHOD: Data were from 3 trials for 270 schizophrenia inpatients receiving antipsychotics. Internal consistency, validity, sensitivity to change, and scalability using PANSS-30, PANSS-8, and PANSS-6 at each assessment were examined. Early improvement was defined as at least 20% reduction of PANSS-30, PANSS-8, or PANSS-6 scores at week 2. Response was defined as at least 40% reduction of PANSS-30 and remission as a score of PANSS-8 ≤ 3 on each item at endpoint. Receiver operating characteristic analysis was used to determine which rating scale had better discriminative capacity. RESULTS: PANSS-8 and PANSS-6 showed acceptable internal consistency, were highly correlated with PANSS-30, and had sensitivity to change. PANSS-8 and PANSS-6 were scalable at each assessment, except for PANSS-6 at baseline. Early improvement using PANSS-8 or PANSS-6 had comparable predictive values with that of PANSS-30 for response/remission. CONCLUSION: PANSS-8 and PANSS-6 are clinically useful measures. Early improvement, regardless of whether PANSS-30, PANSS-8, or PANSS-6 is used, is a statistically significant predictor of response/remission.

DNA demethylation of the TIM-3 promoter is critical for its stable expression on T cells.

The T-cell immunoglobulin and mucin domain-containing protein 3 (TIM-3) is selectively expressed on terminally differentiated T helper 1 (Th1) cells and acts as a negative regulator that terminates Th1 responses. The dysregulation of TIM-3 expression on T cells is associated with several autoimmune phenotypes and with chronic viral infections; however, the mechanism of this regulation is unclear. In this study, we investigated the effect of DNA methylation on the expression of TIM-3. By analyzing the sequences of TIM-3 promoter regions in human and mouse, we identified a CpG island within the TIM-3 promoter and demonstrated that the promoter activity was controlled by DNA methylation. Furthermore, treatment with 5-aza-2'-deoxycytidine enhanced TIM-3 expression on mouse primary CD4(+) T cells under Th0-, Th1- or Th2-polarizing conditions. Finally, pyrosequencing analysis revealed that the methylation level of the TIM-3 promoter gradually decreased after each round of T-cell polarization, and this decrease was inversely correlated with TIM-3 expression. These data suggest that the DNA methylation of the TIM-3 promoter cooperates with lineage-specific transcription factors in the control of Th-cell development. In conclusion, DNA methylation-based regulation of TIM-3 may provide novel insights into understanding the dysregulation of TIM-3 expression under pathogenic conditions.

WemIQ: an accurate and robust isoform quantification method for RNA-seq data.

MOTIVATION: The deconvolution of isoform expression from RNA-seq remains challenging because of non-uniform read sampling and subtle differences among isoforms. RESULTS: We present a weighted-log-likelihood expectation maximization method on isoform quantification (WemIQ). WemIQ integrates an effective bias removal with a weighted expectation maximization (EM) algorithm to distribute reads among isoforms efficiently. The weight represents the oversampling or undersampling of sequence reads and is estimated through a generalized Poisson model without any presumption on the bias sources and formats. WemIQ significantly improves the quantification of isoform and gene expression as well as the derived exon inclusion rates. It provides robust expression estimates across different laboratories and protocols, which is valuable for the integrative analysis of RNA-seq. For the recent single-cell RNA-seq data, WemIQ also provides the opportunity to distinguish bias heterogeneity from true biological heterogeneity and uncovers smaller cell-to-cell expression variability.

PCa-RadHop: A transparent and lightweight feed-forward method for clinically significant prostate cancer segmentation.

Prostate Cancer is one of the most frequently occurring cancers in men, with a low survival rate if not early diagnosed. PI-RADS reading has a high false positive rate, thus increasing the diagnostic incurred costs and patient discomfort. Deep learning (DL) models achieve a high segmentation performance, although require a large model size and complexity. Also, DL models lack of feature interpretability and are perceived as "black-boxes" in the medical field. PCa-RadHop pipeline is proposed in this work, aiming to provide a more transparent feature extraction process using a linear model. It adopts the recently introduced Green Learning (GL) paradigm, which offers a small model size and low complexity. PCa-RadHop consists of two stages: Stage-1 extracts data-driven radiomics features from the bi-parametric Magnetic Resonance Imaging (bp-MRI) input and predicts an initial heatmap. To reduce the false positive rate, a subsequent stage-2 is introduced to refine the predictions by including more contextual information and radiomics features from each already detected Region of Interest (ROI). Experiments on the largest publicly available dataset, PI-CAI, show a competitive performance standing of the proposed method among other deep DL models, achieving an area under the curve (AUC) of 0.807 among a cohort of 1,000 patients. Moreover, PCa-RadHop maintains orders of magnitude smaller model size and complexity.

Vertical-cavity and randomly scattered lasing from different thicknesses of epitaxial ZnO films grown on Y2O3-buffered Si (111).

Two different types of lasing modes, vertical Fabry-Perot cavity and random lasing, were observed in ZnO epi-films of different thicknesses grown on Si (111) substrates. Under optical excitation at room temperature by a frequency tripled Nd:YVO4 laser with wavelength of 355 nm, the lasing thresholds are low due to high crystalline quality of the ZnO epitaxial films, which act as microresonators. For the thick ZnO layer (1,200 nm), its lasing action is originated from the random scattering due to the high density of crack networks developed in the thick ZnO film. However, the low crack density of the thin film (555 nm) fails to provide feedback loops essential for random scattering. Nevertheless, even the lower threshold lasing is achieved by the Fabry-Perot cavity formed by two interfaces of the thin ZnO film. The associated lasing modes of the thin ZnO film can be characterized as the transverse Gaussian modes attributed to the smooth curved surfaces.

GraphHop: An Enhanced Label Propagation Method for Node Classification.

A scalable semisupervised node classification method on graph-structured data, called GraphHop, is proposed in this work. The graph contains all nodes' attributes and link connections but labels of only a subset of nodes. Graph convolutional networks (GCNs) have provided superior performance in node label classification over the traditional label propagation (LP) methods for this problem. Nevertheless, current GCN algorithms suffer from a considerable amount of labels for training because of high model complexity or cannot be easily generalized to large-scale graphs due to the expensive cost of loading the entire graph and node embeddings. Besides, nonlinearity makes the optimization process a mystery. To this end, an enhanced LP method, called GraphHop, is proposed to tackle these problems. GraphHop can be viewed as a smoothening LP algorithm, in which each propagation alternates between two steps: label aggregation and label update. In the label aggregation step, multihop neighbor embeddings are aggregated to the center node. In the label update step, new embeddings are learned and predicted for each node based on aggregated results from the previous step. The two-step iteration improves the graph signal smoothening capacity. Furthermore, to encode attributes, links, and labels on graphs effectively under one framework, we adopt a two-stage training process, i.e., the initialization stage and the iteration stage. Thus, the smooth attribute information extracted from the initialization stage is consistently imposed in the propagation process in the iteration stage. Experimental results show that GraphHop outperforms state-of-the-art graph learning methods on a wide range of tasks in graphs of various sizes (e.g., multilabel and multiclass classification on citation networks, social graphs, and commodity consumption graphs).

Label Efficient Regularization and Propagation for Graph Node Classification.

An enhanced label propagation (LP) method called GraphHop was proposed recently. It outperforms graph convolutional networks (GCNs) in the semi-supervised node classification task on various networks. Although the performance of GraphHop was explained intuitively with joint node attribute and label signal smoothening, its rigorous mathematical treatment is lacking. In this paper, we propose a label efficient regularization and propagation (LERP) framework for graph node classification, and present an alternate optimization procedure for its solution. Furthermore, we show that GraphHop only offers an approximate solution to this framework and has two drawbacks. First, it includes all nodes in the classifier training without taking the reliability of pseudo-labeled nodes into account in the label update step. Second, it provides a rough approximation to the optimum of a subproblem in the label aggregation step. Based on the LERP framework, we propose a new method, named the LERP method, to solve these two shortcomings. LERP determines reliable pseudo-labels adaptively during the alternate optimization and provides a better approximation to the optimum with computational efficiency. Theoretical convergence of LERP is guaranteed. Extensive experiments are conducted to demonstrate the effectiveness and efficiency of LERP. That is, LERP outperforms all benchmarking methods, including GraphHop, consistently on five common test datasets, two large-scale networks, and an object recognition task at extremely low label rates (i.e., 1, 2, 4, 8, 16, and 20 labeled samples per class).

L-BGNN: Layerwise Trained Bipartite Graph Neural Networks.

Learning low-dimensional representations of bipartite graphs enables e-commerce applications, such as recommendation, classification, and link prediction. A layerwise-trained bipartite graph neural network (L-BGNN) embedding method, which is unsupervised, efficient, and scalable, is proposed in this work. To aggregate the information across and within two partitions of a bipartite graph, a customized interdomain message passing (IDMP) operation and an intradomain alignment (IDA) operation are adopted by the proposed L-BGNN method. Furthermore, we develop a layerwise training algorithm for L-BGNN to capture the multihop relationship of large bipartite networks and improve training efficiency. We conduct extensive experiments on several datasets and downstream tasks of various scales to demonstrate the effectiveness and efficiency of the L-BGNN method as compared with state-of-the-art methods. Our codes are publicly available at https://github.com/TianXieUSC/L-BGNN.

Perceptually Weighted Rate Distortion Optimization for Video-Based Point Cloud Compression.

Dynamic point cloud is a volumetric visual data representing realistic 3D scenes for virtual reality and augmented reality applications. However, its large data volume has been the bottleneck of data processing, transmission, and storage, which requires effective compression. In this paper, we propose a Perceptually Weighted Rate-Distortion Optimization (PWRDO) scheme for Video-based Point Cloud Compression (V-PCC), which aims to minimize the perceptual distortion of reconstructed point cloud at the given bit rate. Firstly, we propose a general framework of perceptually optimized V-PCC to exploit visual redundancies in point clouds. Secondly, a multi-scale Projection based Point Cloud quality Metric (PPCM) is proposed to measure the perceptual quality of 3D point cloud. The PPCM model comprises 3D-to-2D patch projection, multi-scale structural distortion measurement, and fusion model. Approximations and simplifications of the proposed PPCM are also presented for both V-PCC integration and low complexity. Thirdly, based on the simplified PPCM model, we propose a PWRDO scheme with Lagrange multiplier adaptation, which is incorporated into the V-PCC to enhance the coding efficiency. Experimental results show that the proposed PPCM models can be used as standalone quality metrics, and they are able to achieve higher consistency with the human subjective scores than the state-of-the-art objective visual quality metrics. Also, compared with the latest V-PCC reference model, the proposed PWRDO-based V-PCC scheme achieves an average bit rate reduction of 13.52%, 8.16%, 10.56% and 9.54%, respectively, in terms of four objective visual quality metrics for point clouds. It is significantly superior to the state-of-the-art coding algorithms. The computational complexity of the proposed PWRDO increases by 1.71% and 0.05% on average to the V-PCC encoder and decoder, respectively, which is negligible. The source codes of the PPCM and PWRDO schemes are available at https://github.com/VVCodec/PPCM-PWRDO.

SUCCESSIVE SUBSPACE LEARNING FOR CARDIAC DISEASE CLASSIFICATION WITH TWO-PHASE DEFORMATION FIELDS FROM CINE MRI.

Cardiac cine magnetic resonance imaging (MRI) has been used to characterize cardiovascular diseases (CVD), often providing a noninvasive phenotyping tool. While recently flourished deep learning based approaches using cine MRI yield accurate characterization results, the performance is often degraded by small training samples. In addition, many deep learning models are deemed a "black box," for which models remain largely elusive in how models yield a prediction and how reliable they are. To alleviate this, this work proposes a lightweight successive subspace learning (SSL) framework for CVD classification, based on an interpretable feedforward design, in conjunction with a cardiac atlas. Specifically, our hierarchical SSL model is based on (i) neighborhood voxel expansion, (ii) unsupervised subspace approximation, (iii) supervised regression, and (iv) multi-level feature integration. In addition, using two-phase 3D deformation fields, including end-diastolic and end-systolic phases, derived between the atlas and individual subjects as input offers objective means of assessing CVD, even with small training samples. We evaluate our framework on the ACDC2017 database, comprising one healthy group and four disease groups. Compared with 3D CNN-based approaches, our framework achieves superior classification performance with 140× fewer parameters, which supports its potential value in clinical use.

Differential association of proinflammatory cytokines with left ventricular diastolic dysfunction in subjects with and without continuous ambulatory peritoneal dialysis.

BACKGROUND AND AIMS: The association between inflammation and left ventricular (LV) diastolic dysfunction in continuous ambulatory peritoneal dialysis (CAPD) and non-CAPD patients is not established. The objective of this study was to test the above association and whether inflammation interacts with CAPD to increase LV diastolic dysfunction risks. METHODS AND RESULTS: 120 subjects with normal creatinine levels and 101 CAPD patients were recruited. Echocardiographic parameters were assessed in all patients. The participants were classified as having LV diastolic dysfunction by echocardiographic findings including mitral inflow E/A ratio < 1, deceleration time > 220 cm/s, or decreased peak annular early diastolic velocity in tissue Doppler imaging. Blood was sampled at the baseline for measurement of inflammation markers, including tissue necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6). Subjects with LV diastolic dysfunction had higher proinflammation cytokines levels in both groups. Inflamed markers correlated significantly with echocardiography parameters for LV diastolic dysfunction in patients receiving CAPD. In a multivariate regression analysis adjusting for all the factors associated with LV diastolic dysfunction, inflammation is still significantly associated with left ventricular diastolic dysfunction (TNF-alpha, OR: 2.6, 95% CI: 2.0-3.35, p < 0.001; IL-6, OR: 1.26, 95% CI: 1.25-1.26, p = 0.01). In addition, the interaction of CAPD and inflammation significantly contributed to the development of LV diastolic dysfunction (CAPD∗ TNF-α: OR: 1.45, 95% CI: 1.13-1.79, P = 0.004). CONCLUSION: We found inflammation plays a vital role for LV diastolic dysfunction especially in CAPD patients. A synergistic effect between CAPD and inflammation, especially TNF-α, would further aggravate LV diastolic dysfunction.

Detection of Rickettsia spp. and host and habitat associations of fleas (Siphonaptera) in eastern Taiwan.

Rickettsia typhi and Rickettsia felis (Rickettsiales: Rickettsiaceae) are two rickettsiae principally transmitted by fleas, but the detection of either pathogen has rarely been attempted in Taiwan. Of 2048 small mammals trapped in eastern Taiwan, Apodemus agrarius Pallas (24.5%) and Mus caroli Bonhote (24.4%) (both: Rodentia: Muridae) were the most abundant, and M. caroli hosted the highest proportion of fleas (63.9% of 330 fleas). Two flea species were identified: Stivalius aporus Jordan and Rothschild (Siphonaptera: Stivaliidae), and Acropsylla episema Rothschild (Siphonaptera: Leptopsyllidae). Nested polymerase chain reaction targeting parts of the ompB and gltA genes showed six fleas to be positive for Rickettsia spp. (3.8% of 160 samples), which showed the greatest similarity to R. felis, Rickettsia japonica, Rickettsia conorii or Rickettsia sp. TwKM01. Rickettsia typhi was not detected in the fleas and Rickettsia co-infection did not occur. Both flea species were more abundant during months with lower temperatures and less rainfall, and flea abundance on M. caroli was not related to soil hardness, vegetative height, ground cover by litter or by understory layer, or the abundance of M. caroli. Our study reveals the potential circulation of R. felis and other rickettsiae in eastern Taiwan, necessitating further surveillance of rickettsial diseases in this region. This is especially important because many novel rickettsioses are emerging worldwide.

R-PointHop: A Green, Accurate, and Unsupervised Point Cloud Registration Method.

Inspired by the recent PointHop classification method, an unsupervised 3D point cloud registration method, called R-PointHop, is proposed in this work. R-PointHop first determines a local reference frame (LRF) for every point using its nearest neighbors and finds local attributes. Next, R-PointHop obtains local-to-global hierarchical features by point downsampling, neighborhood expansion, attribute construction and dimensionality reduction steps. Thus, point correspondences are built in hierarchical feature space using the nearest neighbor rule. Afterwards, a subset of salient points with good correspondence is selected to estimate the 3D transformation. The use of the LRF allows for invariance of the hierarchical features of points with respect to rotation and translation, thus making R-PointHop more robust at building point correspondence, even when the rotation angles are large. Experiments are conducted on the 3DMatch, ModelNet40, and Stanford Bunny datasets, which demonstrate the effectiveness of R-PointHop for 3D point cloud registration. R-PointHop's model size and training time are an order of magnitude smaller than those of deep learning methods, and its registration errors are smaller, making it a green and accurate solution. Our codes are available on GitHub (https://github.com/pranavkdm/R-PointHop).

Geometrical Interpretation and Design of Multilayer Perceptrons.

The multilayer perceptron (MLP) neural network is interpreted from the geometrical viewpoint in this work, that is, an MLP partition an input feature space into multiple nonoverlapping subspaces using a set of hyperplanes, where the great majority of samples in a subspace belongs to one object class. Based on this high-level idea, we propose a three-layer feedforward MLP (FF-MLP) architecture for its implementation. In the first layer, the input feature space is split into multiple subspaces by a set of partitioning hyperplanes and rectified linear unit (ReLU) activation, which is implemented by the classical two-class linear discriminant analysis (LDA). In the second layer, each neuron activates one of the subspaces formed by the partitioning hyperplanes with specially designed weights. In the third layer, all subspaces of the same class are connected to an output node that represents the object class. The proposed design determines all MLP parameters in a feedforward one-pass fashion analytically without backpropagation. Experiments are conducted to compare the performance of the traditional backpropagation-based MLP (BP-MLP) and the new FF-MLP. It is observed that the FF-MLP outperforms the BP-MLP in terms of design time, training time, and classification performance in several benchmarking datasets. Our source code is available at https://colab.research.google.com/drive/1Gz0L8A-nT4ijrUchrhEXXsnaacrFdenn?usp = sharing.

Brain MR Atlas Construction Using Symmetric Deep Neural Inpainting.

Modeling statistical properties of anatomical structures using magnetic resonance imaging is essential for revealing common information of a target population and unique properties of specific subjects. In brain imaging, a statistical brain atlas is often constructed using a number of healthy subjects. When tumors are present, however, it is difficult to either provide a common space for various subjects or align their imaging data due to the unpredictable distribution of lesions. Here we propose a deep learning-based image inpainting method to replace the tumor regions with normal tissue intensities using only a patient population. Our framework has three major innovations: 1) incompletely distributed datasets with random tumor locations can be used for training; 2) irregularly-shaped tumor regions are properly learned, identified, and corrected; and 3) a symmetry constraint between the two brain hemispheres is applied to regularize inpainted regions. Henceforth, regular atlas construction and image registration methods can be applied using inpainted data to obtain tissue deformation, thereby achieving group-specific statistical atlases and patient-to-atlas registration. Our framework was tested using the public database from the Multimodal Brain Tumor Segmentation challenge. Results showed increased similarity scores as well as reduced reconstruction errors compared with three existing image inpainting methods. Patient-to-atlas registration also yielded better results with improved normalized cross-correlation and mutual information and a reduced amount of deformation over the tumor regions.

BERTHop: An Effective Vision-and-Language Model for Chest X-ray Disease Diagnosis.

Vision-and-language (V&L) models take image and text as input and learn to capture the associations between them. These models can potentially deal with the tasks that involve understanding medical images along with their associated text. However, applying V&L models in the medical domain is challenging due to the expensiveness of data annotations and the requirements of domain knowledge. In this paper, we identify that the visual representation in general V&L models is not suitable for processing medical data. To overcome this limitation, we propose BERTHop, a transformer-based model based on PixelHop++ and VisualBERT for better capturing the associations between clinical notes and medical images. Experiments on the OpenI dataset, a commonly used thoracic disease diagnosis benchmark, show that BERTHop achieves an average Area Under the Curve (AUC) of 98.12% which is 1.62% higher than state-of-the-art while it is trained on a 9× smaller dataset.