Gated CNN-Transformer Network for Automatic Cardiovascular Diagnosis using 12-lead Electrocardiogram.

12-lead electrocardiogram (ECG) is a widely used method in the diagnosis of cardiovascular disease (CVD). With the increase in the number of CVD patients, the study of accurate automatic diagnosis methods via ECG has become a research hotspot. The use of deep learning-based methods can reduce the influence of human subjectivity and improve the diagnosis accuracy. In this paper, we propose a 12-lead ECG automatic diagnosis method based on channel features and temporal features fusion. Specifically, we design a gated CNN-Transformer network, in which the CNN block is used to extract signal embeddings to reduce data complexity. The dual-branch transformer structure is used to effectively extract channel and temporal features in low-dimensional embeddings, respectively. Finally, the features from the two branches are fused by the gating unit to achieve automatic CVD diagnosis from 12-lead ECG. The proposed end-to-end approach has more competitive performance than other deep learning algorithms, which achieves an overall diagnostic accuracy of 85.3% in the 12-lead ECG dataset of CPSC-2018.

USBDAN: Unsupervised Scale-aware and Boundary-aware Domain Adaptive Network for Gastric Tumor Segmentation.

Accurate segmentation of gastric tumors from computed tomography (CT) images provides useful image information for guiding the diagnosis and treatment of gastric cancer. Researchers typically collect datasets from multiple medical centers to increase sample size and representation, but this raises the issue of data heterogeneity. To this end, we propose a new cross-center 3D tumor segmentation method named unsupervised scale-aware and boundary-aware domain adaptive network (USBDAN), which includes a new 3D neural network that efficiently bridges an Anisotropic neural network and a Transformer (AsTr) for extracting multi-scale features from the CT images with anisotropic resolution, and a scale-aware and boundary-aware domain alignment (SaBaDA) module for adaptively aligning multi-scale features between two domains and enhancing tumor boundary drawing based on location-related information drawn from each sample across all domains. We evaluate the proposed method on an in-house CT image dataset collected from four medical centers. Our results demonstrate that the proposed method outperforms several state-of-the-art methods.

Adversarial learning-based multi-level dense-transmission knowledge distillation for AP-ROP detection.

The Aggressive Posterior Retinopathy of Prematurity (AP-ROP) is the major cause of blindness for premature infants. The automatic diagnosis method has become an important tool for detecting AP-ROP. However, most existing automatic diagnosis methods were with heavy complexity, which hinders the development of the detecting devices. Hence, a small network (student network) with a high imitation ability is exactly needed, which can mimic a large network (teacher network) with promising diagnostic performance. Also, if the student network is too small due to the increasing gap between teacher and student networks, the diagnostic performance will drop. To tackle the above issues, we propose a novel adversarial learning-based multi-level dense knowledge distillation method for detecting AP-ROP. Specifically, the pre-trained teacher network is utilized to train multiple intermediate-size networks (i.e., teacher-assistant networks) and one student network by dense transmission mode, where the knowledge from all upper-level networks is transmitted to the current lower-level network. To ensure that two adjacent networks can distill the abundant knowledge, the adversarial learning module is leveraged to enforce the lower-level network to generate the features that are similar to those of the upper-level network. Extensive experiments demonstrate that our proposed method can realize the effective knowledge distillation from the teacher to student networks. We achieve a promising knowledge distillation performance for our private dataset and a public dataset, which can provide a new insight for devising lightweight detecting systems of fundus diseases for practical use.

LAC-GAN: Lesion attention conditional GAN for Ultra-widefield image synthesis.

Automatic detection of retinal diseases based on deep learning technology and Ultra-widefield (UWF) images plays an important role in clinical practices in recent years. However, due to small lesions and limited data samples, it is not easy to train a detection-accurate model with strong generalization ability. In this paper, we propose a lesion attention conditional generative adversarial network (LAC-GAN) to synthesize retinal images with realistic lesion details to improve the training of the disease detection model. Specifically, the generator takes the vessel mask and class label as the conditional inputs, and processes the random Gaussian noise by a series of residual block to generate the synthetic images. To focus on pathological information, we propose a lesion feature attention mechanism based on random forest (RF) method, which constructs its reverse activation network to activate the lesion features. For discriminator, a weight-sharing multi-discriminator is designed to improve the performance of model by affine transformations. Experimental results on multi-center UWF image datasets demonstrate that the proposed method can generate retinal images with reasonable details, which helps to enhance the performance of the disease detection model.

Gene-related Parkinson's disease diagnosis via feature-based multi-branch octave convolution network.

Parkinson's disease (PD) is a common neurodegenerative disease in the elderly population. PD is irreversible and its diagnosis mainly relies on clinical symptoms. Hence, its effective diagnosis is vital. PD has the related gene mutation called gene-related PD, which can be diagnosed not only in the specific PD patients, but also in the healthiest people without clinical symptoms of PD. Since mutations in PD-related genes can affect healthy people, and unaffected PD-related gene carriers can develop into PD patients, it is very necessary to distinguish gene-related PD diseases. The magnetic resonance imaging (MRI) has a lot of information about brain tissue, which can distinguish gene-related PD diseases. However, the limited amount of the gene-related cohort in PD is a challenge for further diagnosis. Therefore, we develop a joint learning framework called feature-based multi-branch octave convolution network (FMOCNN), which uses MRI data for gene-related cohort PD diagnosis. FMOCNN performs sample-feature selection to learn discriminative samples and features and contains a deep neural network to obtain high-level feature representation from various feature types. Specifically, we first train a cardinality constrained sample-feature selection (CCSFS) model to select informative samples and features. We then establish a multi-branch octave convolution neural network (MBOCNN) to jointly train multiple feature inputs. High/low-frequency learning in MBOCNN is exploited to reduce redundant feature information and enhance the feature expression ability. Our method is validated on the publicly available Parkinson's Progression Markers Initiative (PPMI) dataset. Experiments demonstrate that our method achieves promising classification performance and outperforms similar algorithms.

Unsupervised Domain Adaptation Based Image Synthesis and Feature Alignment for Joint Optic Disc and Cup Segmentation.

Due to the discrepancy of different devices for fundus image collection, a well-trained neural network is usually unsuitable for another new dataset. To solve this problem, the unsupervised domain adaptation strategy attracts a lot of attentions. In this paper, we propose an unsupervised domain adaptation method based image synthesis and feature alignment (ISFA) method to segment optic disc and cup on fundus images. The GAN-based image synthesis (IS) mechanism along with the boundary information of optic disc and cup is utilized to generate target-like query images, which serves as the intermediate latent space between source domain and target domain images to alleviate the domain shift problem. Specifically, we use content and style feature alignment (CSFA) to ensure the feature consistency among source domain images, target-like query images and target domain images. The adversarial learning is used to extract domain-invariant features for output-level feature alignment (OLFA). To enhance the representation ability of domain-invariant boundary structure information, we introduce the edge attention module (EAM) for low-level feature maps. Eventually, we train our proposed method on the training set of the REFUGE challenge dataset and test it on Drishti-GS and RIM-ONE_r3 datasets. On the Drishti-GS dataset, our method achieves about 3% improvement of Dice on optic cup segmentation over the next best method. We comprehensively discuss the robustness of our method for small dataset domain adaptation. The experimental results also demonstrate the effectiveness of our method. Our code is available at https://github.com/thinkobj/ISFA.

Parkinson's Disease Classification and Clinical Score Regression via United Embedding and Sparse Learning From Longitudinal Data.

Parkinson's disease (PD) is known as an irreversible neurodegenerative disease that mainly affects the patient's motor system. Early classification and regression of PD are essential to slow down this degenerative process from its onset. In this article, a novel adaptive unsupervised feature selection approach is proposed by exploiting manifold learning from longitudinal multimodal data. Classification and clinical score prediction are performed jointly to facilitate early PD diagnosis. Specifically, the proposed approach performs united embedding and sparse regression, which can determine the similarity matrices and discriminative features adaptively. Meanwhile, we constrain the similarity matrix among subjects and exploit the l2,p norm to conduct sparse adaptive control for obtaining the intrinsic information of the multimodal data structure. An effective iterative optimization algorithm is proposed to solve this problem. We perform abundant experiments on the Parkinson's Progression Markers Initiative (PPMI) data set to verify the validity of the proposed approach. The results show that our approach boosts the performance on the classification and clinical score regression of longitudinal data and surpasses the state-of-the-art approaches.

Multi-center sparse learning and decision fusion for automatic COVID-19 diagnosis.

The coronavirus disease 2019 (COVID-19) pandemic caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has led to a sharp increase in hospitalized patients with multi-organ disease pneumonia. Early and automatic diagnosis of COVID-19 is essential to slow down the spread of this epidemic and reduce the mortality of patients infected with SARS-CoV-2. In this paper, we propose a joint multi-center sparse learning (MCSL) and decision fusion scheme exploiting chest CT images for automatic COVID-19 diagnosis. Specifically, considering the inconsistency of data in multiple centers, we first convert CT images into histogram of oriented gradient (HOG) images to reduce the structural differences between multi-center data and enhance the generalization performance. We then exploit a 3-dimensional convolutional neural network (3D-CNN) model to learn the useful information between and within 3D HOG image slices and extract multi-center features. Furthermore, we employ the proposed MCSL method that learns the intrinsic structure between multiple centers and within each center, which selects discriminative features to jointly train multi-center classifiers. Finally, we fuse these decisions made by these classifiers. Extensive experiments are performed on chest CT images from five centers to validate the effectiveness of the proposed method. The results demonstrate that the proposed method can improve COVID-19 diagnosis performance and outperform the state-of-the-art methods.

Latent Space Learning and Feature Learning using Multi-template for Multi-classification of Alzheimer's Disease.

Alzheimer's disease (AD) is a common brain disease in the elderly that leads to thinking, memory, and behavior disorders. As the population ages, the proportion of AD patients is also increasing. Accordingly, computer-aided diagnosis of AD attracts more and more attention recently. In this paper, we propose a novel model combining latent space learning and feature learning using features extracted from multiple templates for AD multi-classification. Specifically, latent space learning is employed to obtain the inter-relationship between multiple templates, and feature learning is performed to explore the intrinsic relation in feature space. Finally, the most discriminative features are selected to boost the multi-classification performance. Our proposed model uses the data from the Alzheimer's disease neuroimaging initiative dataset. Furthermore, a series of comparative experiments indicate that our proposed model is quite competitive.

Dual feedback based bipolar current source with high stability for driving voice coil motors in wide temperature ranges.

Bipolar current sources with a stability better than 0.1% in the temperature range of -30 to +70 °C are demanded for driving voice coil motors applied in a new ultra-quiet satellite platform, but almost none of the existing designs satisfy the harsh requirements. This paper presents a possible solution, which is essentially a floating-load, bipolar current source circuit with a dual feedback path. The key circuit is a composite amplifier (co-amp) composed of a high precision amplifier for error correction and a high power amplifier for load driving. The first feedback path comprises a specially designed four-wire current-sense resistor for current-to-voltage conversion and a discrete instrumentation amplifier for amplifying the converted voltage and closing the loop. The second feedback path is a proposed compensation network for loop stability. Error budgets for evaluating current stability and choosing key components of the circuit are comprehensively studied based on a derived rigorous current equation. Loop-stability problems attributable to the inductive load and the high open-loop gain of the co-amp are analyzed, and the proposed dual feedback compensation method is verified by theory, simulation, and measurement. All these contributions are demonstrated by three implemented prototypes with an output of up to ±2 A. The measured results agree well with theoretical predictions. The best and the worst stability performances of the three prototypes at +2 and -2 A are, respectively, 394 and 986 ppm in the temperature range of -30 to +70 °C, which are close to the theoretical value of 776 ppm.

Cross-attention multi-branch network for fundus diseases classification using SLO images.

Fundus diseases classification is vital for the health of human beings. However, most of existing methods detect diseases by means of single angle fundus images, which lead to the lack of pathological information. To address this limitation, this paper proposes a novel deep learning method to complete different fundus diseases classification tasks using ultra-wide field scanning laser ophthalmoscopy (SLO) images, which have an ultra-wide field view of 180-200˚. The proposed deep model consists of multi-branch network, atrous spatial pyramid pooling module (ASPP), cross-attention and depth-wise attention module. Specifically, the multi-branch network employs the ResNet-34 model as the backbone to extract feature information, where the ResNet-34 model with two-branch is followed by the ASPP module to extract multi-scale spatial contextual features by setting different dilated rates. The depth-wise attention module can provide the global attention map from the multi-branch network, which enables the network to focus on the salient targets of interest. The cross-attention module adopts the cross-fusion mode to fuse the channel and spatial attention maps from the ResNet-34 model with two-branch, which can enhance the representation ability of the disease-specific features. The extensive experiments on our collected SLO images and two publicly available datasets demonstrate that the proposed method can outperform the state-of-the-art methods and achieve quite promising classification performance of the fundus diseases.

Attention-Guided Multi-Branch Convolutional Neural Network for Mitosis Detection From Histopathological Images.

Mitotic count is an important indicator for assessing the invasiveness of breast cancers. Currently, the number of mitoses is manually counted by pathologists, which is both tedious and time-consuming. To address this situation, we propose a fast and accurate method to automatically detect mitosis from the histopathological images. The proposed method can automatically identify mitotic candidates from histological sections for mitosis screening. Specifically, our method exploits deep convolutional neural networks to extract high-level features of mitosis to detect mitotic candidates. Then, we use spatial attention modules to re-encode mitotic features, which allows the model to learn more efficient features. Finally, we use multi-branch classification subnets to screen the mitosis. Compared to existing related methods in literature, our method obtains the best detection results on the dataset of the International Pattern Recognition Conference (ICPR) 2012 Mitosis Detection Competition. Code has been made available at: https://github.com/liushaomin/MitosisDetection.

AMD-GAN: Attention encoder and multi-branch structure based generative adversarial networks for fundus disease detection from scanning laser ophthalmoscopy images.

The scanning laser ophthalmoscopy (SLO) has become an important tool for the determination of peripheral retinal pathology, in recent years. However, the collected SLO images are easily interfered by the eyelash and frame of the devices, which heavily affect the key feature extraction of the images. To address this, we propose a generative adversarial network called AMD-GAN based on the attention encoder (AE) and multi-branch (MB) structure for fundus disease detection from SLO images. Specifically, the designed generator consists of two parts: the AE and generation flow network, where the real SLO images are encoded by the AE module to extract features and the generation flow network to handle the random Gaussian noise by a series of residual block with up-sampling (RU) operations to generate fake images with the same size as the real ones, where the AE is also used to mine features for generator. For discriminator, a ResNet network using MB is devised by copying the stage 3 and stage 4 structures of the ResNet-34 model to extract deep features. Furthermore, the depth-wise asymmetric dilated convolution is leveraged to extract local high-level contextual features and accelerate the training process. Besides, the last layer of discriminator is modified to build the classifier to detect the diseased and normal SLO images. In addition, the prior knowledge of experts is utilized to improve the detection results. Experimental results on the two local SLO datasets demonstrate that our proposed method is promising in detecting the diseased and normal SLO images with the experts labeling.

Adaptive sparse learning using multi-template for neurodegenerative disease diagnosis.

Neurodegenerative diseases are excessively affecting millions of patients, especially elderly people. Early detection and management of these diseases are crucial as the clinical symptoms take years to appear after the onset of neuro-degeneration. This paper proposes an adaptive feature learning framework using multiple templates for early diagnosis. A multi-classification scheme is developed based on multiple brain parcellation atlases with various regions of interest. Different sets of features are extracted and then fused, and a feature selection is applied with an adaptively chosen sparse degree. In addition, both linear discriminative analysis and locally preserving projections are integrated to construct a least square regression model. Finally, we propose a feature space to predict the severity of the disease by the guidance of clinical scores. Our proposed method is validated on both Alzheimer's disease neuroimaging initiative and Parkinson's progression markers initiative databases. Extensive experimental results suggest that the proposed method outperforms the state-of-the-art methods, such as the multi-modal multi-task learning or joint sparse learning. Our method demonstrates that accurate feature learning facilitates the identification of the highly relevant brain regions with significant contribution in the prediction of disease progression. This may pave the way for further medical analysis and diagnosis in practical applications.

Protein-Protein Interactions Prediction via Multimodal Deep Polynomial Network and Regularized Extreme Learning Machine.

Predicting the protein-protein interactions (PPIs) has played an important role in many applications. Hence, a novel computational method for PPIs prediction is highly desirable. PPIs endow with protein amino acid mutation rate and two physicochemical properties of protein (e.g., hydrophobicity and hydrophilicity). Deep polynomial network (DPN) is well-suited to integrate these modalities since it can represent any function on a finite sample dataset via the supervised deep learning algorithm. We propose a multimodal DPN (MDPN) algorithm to effectively integrate these modalities to enhance prediction performance. MDPN consists of a two-stage DPN, the first stage feeds multiple protein features into DPN encoding to obtain high-level feature representation while the second stage fuses and learns features by cascading three types of high-level features in the DPN encoding. We employ a regularized extreme learning machine to predict PPIs. The proposed method is tested on the public dataset of H. pylori, Human, and Yeast and achieves average accuracies of 97.87%, 99.90%, and 98.11%, respectively. The proposed method also achieves good accuracies on other datasets. Furthermore, we test our method on three kinds of PPI networks and obtain superior prediction results.

Parkinson's Disease Diagnosis via Joint Learning From Multiple Modalities and Relations.

Parkinson's disease (PD) is a neurodegenerative progressive disease that mainly affects the motor systems of patients. To slow this disease deterioration, early and accurate diagnosis of PD is an effective way, which alleviates mental and physical sufferings by clinical intervention. In this paper, we propose a joint regression and classification framework for PD diagnosis via magnetic resonance and diffusion tensor imaging data. Specifically, we devise a unified multitask feature selection model to explore multiple relationships among features, samples, and clinical scores. We regress four clinical variables of depression, sleep, olfaction, cognition scores, as well as perform the classification of PD disease from the multimodal data. The multitask model explores the relationships at the level of clinical scores, image features, and subjects, to select the most informative and diseased-related features for diagnosis. The proposed method is evaluated on the public Parkinson's progression markers initiative dataset. The extensive experimental results show that the multitask framework can effectively boost the performance of regression and classification and outperforms other state-of-the-art methods. The computerized predictions of clinical scores and label for PD diagnosis may offer quantitative reference for decision support as well.

An Improved Object Detection Method for Mitosis Detection.

Breast cancer grading is important for patient prognosis, and the mitosis count is one of the most important indicators for breast cancer grading. Traditional methods use handcraft features and deep learning based methods to detect mitosis in a classified model. These methods are time-consuming and difficult for practical clinical practice application. For this reason, this paper proposes an improved object detection method for automatic mitosis detection from histological images. First, we use a convolutional neural network (CNN) to automatically extract mitosis features. Then, we use the region proposed network (RPN) to locate a set of class-agnostic mitosis proposals. Finally, we use the improved R-CNN subnet to screen for mitosis from these proposals. Our approach achieved the best results in the ICPR2012 mitosis detection competition test dataset. Additionally, our proposed method is fast enough to be potentially used in clinical and health centers.

Sparse feature learning for multi-class Parkinson's disease classification.

This paper solves the multi-class classification problem for Parkinson's disease (PD) analysis by a sparse discriminative feature selection framework. Specifically, we propose a framework to construct a least square regression model based on the Fisher's linear discriminant analysis (LDA) and locality preserving projection (LPP). This framework utilizes the global and local information to select the most relevant and discriminative features to boost classification performance. Differing in previous methods for binary classification, we perform a multi-class classification for PD diagnosis. Our proposed method is evaluated on the public available Parkinson's progression markers initiative (PPMI) datasets. Extensive experimental results indicate that our proposed method identifies highly suitable regions for further PD analysis and diagnosis and outperforms state-of-the-art methods.

Joint regression and classification via relational regularization for Parkinson's disease diagnosis.

It is known that the symptoms of Parkinson's disease (PD) progress successively, early and accurate diagnosis of the disease is of great importance, which slows the disease deterioration further and alleviates mental and physical suffering. In this paper, we propose a joint regression and classification scheme for PD diagnosis using baseline multi-modal neuroimaging data. Specifically, we devise a new feature selection method via relational learning in a unified multi-task feature selection model. Three kinds of relationships (e.g., relationships among features, responses, and subjects) are integrated to represent the similarities among features, responses, and subjects. Our proposed method exploits five regression variables (depression, sleep, olfaction, cognition scores and a clinical label) to jointly select the most discriminative features for clinical scores prediction and class label identification. Extensive experiments are conducted to demonstrate the effectiveness of the proposed method on the Parkinson's Progression Markers Initiative (PPMI) dataset. Our experimental results demonstrate that multi-modal data can effectively enhance the performance in class label identification compared with single modal data. Our proposed method can greatly improve the performance in clinical scores prediction and outperforms the state-of-art methods as well. The identified brain regions can be recognized for further medical analysis and diagnosis.

Prediction of protein-protein interactions by label propagation with protein evolutionary and chemical information derived from heterogeneous network.

Prediction of protein-protein interactions (PPIs) is of great significance. To achieve this, we propose a novel computational method for PPIs prediction based on a similarity network fusion (SNF) model for integrating the physical and chemical properties of proteins. Specifically, the physical and chemical properties of protein are the protein amino acid mutation rate and its hydrophobicity, respectively. The amino acid mutation rate is extracted using a BLOSUM62 matrix, which puts the protein sequence into block substitution matrix. The SNF model is exploited to fuse protein physical and chemical features of multiple data by iteratively updating each original network. Finally, the complementary features from the fused network are fed into a label propagation algorithm (LPA) for PPIs prediction. The experimental results show that the proposed method achieves promising performance and outperforms the traditional methods for the public dataset of H. pylori, Human, and Yeast. In addition, our proposed method achieves average accuracy of 76.65%, 81.98%, 84.56%, 84.01% and 84.38% on E. coli, C. elegans, H. sapien, H. pylori and M. musculus datasets, respectively. Comparison results demonstrate that the proposed method is very promising and provides a cost-effective alternative for predicting PPIs. The source code and all datasets are available at http://pan.baidu.com/s/1dF7rp7N.