Accurate segmentation of medical images is essential for clinical decision-making, and deep learning techniques have shown remarkable results in this area. However, existing segmentation models that combine transformer and convolutional neural networks often use skip connections in U-shaped networks, which may limit their ability to capture contextual information in medical images. To address this limitation, we propose a coordinated mobile and residual transformer UNet (MRC-TransUNet) that combines the strengths of transformer and UNet architectures. Our approach uses a lightweight MR-ViT to address the semantic gap and a reciprocal attention module to compensate for the potential loss of details. To better explore long-range contextual information, we use skip connections only in the first layer and add MR-ViT and RPA modules in the subsequent downsampling layers. In our study, we evaluated the effectiveness of our proposed method on three different medical image segmentation datasets, namely, breast, brain, and lung. Our proposed method outperformed state-of-the-art methods in terms of various evaluation metrics, including the Dice coefficient and Hausdorff distance. These results demonstrate that our proposed method can significantly improve the accuracy of medical image segmentation and has the potential for clinical applications. Illustration of the proposed MRC-TransUNet. For the input medical images, we first subject them to an intrinsic downsampling operation and then replace the original jump connection structure using MR-ViT. The output feature representations at different scales are fused by the RPA module. Finally, an upsampling operation is performed to fuse the features to restore them to the same resolution as the input image.

In view of the major depressive disorder characteristics such as high mortality as well as high recurrence, it is important to explore an objective and effective detection method for major depressive disorder. Considering the advantages complementary of different machine learning algorithms in information mining process, as well as the fusion complementary of different information, in this study, the spatial-temporal electroencephalography fusion framework using neural network is proposed for major depressive disorder detection. Since electroencephalography is a typical time series signal, we introduce recurrent neural network embedded in long short-term memory unit for extract temporal domain features to solve the problem of long-distance information dependence. To reduce the volume conductor effect, the temporal electroencephalography data are mapping into a spatial brain functional network using phase lag index, then the spatial domain features were extracted from brain functional network using 2D convolutional neural networks. Considering the complementarity between different types of features, the spatial-temporal electroencephalography features are fused to achieve data diversity. The experimental results show that spatial-temporal features fusion can improve the detection accuracy of major depressive disorder with a highest of 96.33%. In addition, our research also found that theta, alpha, and full frequency band in brain regions of left frontal, left central, right temporal are closely related to MDD detection, especially theta frequency band in left frontal region. Only using single-dimension EEG data as decision basis, it is difficult to fully explore the valuable information hidden in the data, which affects the overall detection performance of MDD. Meanwhile, different algorithms have their own advantages for different application scenarios. Ideally, different algorithms should use their respective advantages to jointly address complex problems in engineering fields. To this end, we propose a computer-aided MDD detection framework based on spatial-temporal EEG fusion using neural network, as shown in Fig. 1. The simplified process is as follows: (1) Raw EEG data acquisition and preprocessing. (2) The time series EEG data of each channel are input as recurrent neural network (RNN), and RNN is used to process and extract temporal domain (TD) features. (3) The BFN among different EEG channels is constructed, and CNN is used to process and extract the spatial domain (SD) features of the BFN. (4) Based on the theory of information complementarity, the spatial-temporal information is fused to realize efficient MDD detection. Fig. 1 MDD detection framework based on spatial-temporal EEG fusion.

Soft subspace clustering (SSC), which analyzes high-dimensional data and applies various weights to each cluster class to assess the membership degree of each cluster to the space, has shown promising results in recent years. This method of clustering assigns distinct weights to each cluster class. By introducing spatial information, enhanced SSC algorithms improve the degree to which intraclass compactness and interclass separation are achieved. However, these algorithms are sensitive to noisy data and have a tendency to fall into local optima. In addition, the segmentation accuracy is poor because of the influence of noisy data. In this study, an SSC approach that is based on particle swarm optimization is suggested with the intention of reducing the interference caused by noisy data. The particle swarm optimization method is used to locate the best possible clustering center. Second, increasing the amount of geographical membership makes it possible to utilize the spatial information to quantify the link between different clusters in a more precise manner. In conclusion, the extended noise clustering method is implemented in order to maximize the weight. Additionally, the constraint condition of the weight is changed from the equality constraint to the boundary constraint in order to reduce the impact of noise. The methodology presented in this research works to reduce the amount of sensitivity the SSC algorithm has to noisy data. It is possible to demonstrate the efficacy of this algorithm by using photos with noise already present or by introducing noise to existing photographs. The revised SSC approach based on particle swarm optimization (PSO) is demonstrated to have superior segmentation accuracy through a number of trials; as a result, this work gives a novel method for the segmentation of noisy images.

CD47/SIRPα pathway is a new breakthrough in the field of tumor immunity after PD-1/PD-L1. While current monoclonal antibody therapies targeting CD47/SIRPα have demonstrated some anti-tumor effectiveness, there are several inherent limitations associated with these formulations. In the paper, we developed a predictive model that combines next-generation phage display (NGPD) and traditional machine learning methods to distinguish CD47 binding peptides. First, we utilized NGPD biopanning technology to screen CD47 binding peptides. Second, ten traditional machine learning methods based on multiple peptide descriptors and three deep learning methods were used to build computational models for identifying CD47 binding peptides. Finally, we proposed an integrated model based on support vector machine. During the five-fold cross-validation, the integrated predictor demonstrated specificity, accuracy, and sensitivity of 0.755, 0.764, and 0.772, respectively. Furthermore, an online bioinformatics tool called CD47Binder has been developed for the integrated predictor. This tool is readily accessible on http://i.uestc.edu.cn/CD47Binder/cgi-bin/CD47Binder.pl .

Studying the formation and stability of atherosclerotic plaques in the hemodynamic field is essential for understanding the growth mechanism and preventive treatment of atherosclerotic plaques. In this paper, based on a multiplayer porous wall model, we established a two-way fluid-solid interaction with time-varying inlet flow. The lipid-rich necrotic core (LRNC) and stress in atherosclerotic plaque were described for analyzing the stability of atherosclerotic plaques during the plaque growth by solving advection-diffusion-reaction equations with finite-element method. It was found that LRNC appeared when the lipid levels of apoptotic materials (such as macrophages, foam cells) in the plaque reached a specified lower concentration, and increased with the plaque growth. LRNC was positively correlated with the blood pressure and was negatively correlated with the blood flow velocity. The maximum stress was mainly located at the necrotic core and gradually moved toward the left shoulder of the plaque with the plaque growth, which increases the plaque instability and the risk of the plaque shedding. The computational model may contribute to understanding the mechanisms of early atherosclerotic plaque growth and the risk of instability in the plaque growth.

Recently developed single-cell RNA-seq (scRNA-seq) technology has given researchers the chance to investigate single-cell level of disease development. Clustering is one of the most essential strategies for analyzing scRNA-seq data. Choosing high-quality feature sets can significantly enhance the outcomes of single-cell clustering and classification. But computationally burdensome and highly expressed genes cannot afford a stabilized and predictive feature set for technical reasons. In this study, we introduce scFED, a feature-engineered gene selection framework. scFED identifies prospective feature sets to eliminate the noise fluctuation. And fuse them with existing knowledge from the tissue-specific cellular taxonomy reference database (CellMatch) to avoid the influence of subjective factors. Then present a reconstruction approach for noise reduction and crucial information amplification. We apply scFED on four genuine single-cell datasets and compare it with other techniques. According to the results, scFED improves clustering, decreases dimension of the scRNA-seq data, improves cell type identification when combined with clustering algorithms, and has higher performance than other methods. Therefore, scFED offers certain benefits in scRNA-seq data gene selection.

Classification of glomerular pathology based on histology sections is the key to diagnose the type and degree of kidney diseases. To address problems in the classification of glomerular lesions in children, a deep learning-based complete glomerular classification framework was designed to detect and classify glomerular pathology. A neural network integrating Resnet and Senet (RS-INet) was proposed and a glomerular classification algorithm implemented to achieve high-precision classification of glomerular pathology. SE-Resnet was applied with improvement by transforming the convolutional layer of the original Resnet residual block into a convolutional block with smaller parameters as well as reduced network parameters on the premise of ensuring network performance. Experimental results showed that our algorithm had the best performance in differentiating mesangial proliferative glomerulonephritis (MsPGN), crescent glomerulonephritis (CGN), and glomerulosclerosis (GS) from normal glomerulus (Normal) compared with other classification algorithms. The accuracy rates were 0.960, 0.940, 0.937, and 0.968, respectively. This suggests that the classification algorithm proposed in the present study is able to identify glomerular lesions with a higher precision, and distinguish similar glomerular pathologies from each other.

Breast cancer is commonly diagnosed with mammography. Using image segmentation algorithms to separate lesion areas in mammography can facilitate diagnosis by doctors and reduce their workload, which has important clinical significance. Because large, accurately labeled medical image datasets are difficult to obtain, traditional clustering algorithms are widely used in medical image segmentation as an unsupervised model. Traditional unsupervised clustering algorithms have limited learning knowledge. Moreover, some semi-supervised fuzzy clustering algorithms cannot fully mine the information of labeled samples, which results in insufficient supervision. When faced with complex mammography images, the above algorithms cannot accurately segment lesion areas. To address this, a semi-supervised fuzzy clustering based on knowledge weighting and cluster center learning (WSFCM_V) is presented. According to prior knowledge, three learning modes are proposed: a knowledge weighting method for cluster centers, Euclidean distance weights for unlabeled samples, and learning from the cluster centers of labeled sample sets. These strategies improve the clustering performance. On real breast molybdenum target images, the WSFCM_V algorithm is compared with currently popular semi-supervised and unsupervised clustering algorithms. WSFCM_V has the best evaluation index values. Experimental results demonstrate that compared with the existing clustering algorithms, WSFCM_V has a higher segmentation accuracy than other clustering algorithms, both for larger lesion regions like tumor areas and for smaller lesion areas like calcification point areas. Figure. The principle of Euclidean distance weighting for unlabeled samples.

PURPOSE: Studies relating to the right ventricle (RV) are inadequate, and specific diagnostic algorithms still need to be improved. This essay is designed to make exploration and verification on an algorithm of deep learning based on imaging and clinical data to detect RV abnormalities. METHODS: The Automated Cardiac Diagnosis Challenge dataset includes 20 subjects with RV abnormalities (an RV cavity volume which is higher than 110 mL/m2 or RV ejection fraction which is lower than 40%) and 20 normal subjects who suffered from both cardiac MRI. The subjects were separated into training and validation sets in a ratio of 7:3 and were modeled by utilizing a nerve net of deep-learning and six machine-learning algorithms. Eight MRI specialists from multiple centers independently determined whether each subject in the validation group had RV abnormalities. Model performance was evaluated based on the AUC, accuracy, recall, sensitivity and specificity. Furthermore, a preliminary assessment of patient disease risk was performed based on clinical information using a nomogram. RESULTS: The deep-learning neural network outperformed the other six machine-learning algorithms, with an AUC value of 1 (95% confidence interval: 1-1) on both training group and validation group. This algorithm surpassed most human experts (87.5%). In addition, the nomogram model could evaluate a population with a disease risk of 0.2-0.8. CONCLUSIONS: A deep-learning algorithm could effectively identify patients with RV abnormalities. This AI algorithm developed specifically for right ventricular abnormalities will improve the detection of right ventricular abnormalities at all levels of care units and facilitate the timely diagnosis and treatment of related diseases. In addition, this study is the first to validate the algorithm's ability to classify RV abnormalities by comparing it with human experts.

Kidney stone disease is one of the most common and serious health problems in much of the world, leading to many hospitalizations with severe pain. Detecting small stones is difficult and time-consuming, so an early diagnosis of kidney disease is needed to prevent the loss of kidney failure. Recent advances in artificial intelligence (AI) found to be very successful in the diagnosis of various diseases in the biomedical field. However, existing models using deep networks have several problems, such as high computational cost, long training time, and huge parameters. Providing a low-cost solution for diagnosing kidney stones in a medical decision support system is of paramount importance. Therefore, in this study, we propose "StoneNet", a lightweight and high-performance model for the detection of kidney stones based on MobileNet using depthwise separable convolution. The proposed model includes a combination of global average pooling (GAP), batch normalization, dropout layer, and dense layers. Our study shows that using GAP instead of flattening layers greatly improves the robustness of the model by significantly reducing the parameters. The developed model is benchmarked against four pre-trained models as well as the state-of-the-art heavy model. The results show that the proposed model can achieve the highest accuracy of 97.98%, and only requires training and testing time of 996.88 s and 14.62 s. Several parameters, such as different batch sizes and optimizers, were considered to validate the proposed model. The proposed model is computationally faster and provides optimal performance than other considered models. Experiments on a large kidney dataset of 1799 CT images show that StoneNet has superior performance in terms of higher accuracy and lower complexity. The proposed model can assist the radiologist in faster diagnosis of kidney stones and has great potential for deployment in real-time applications.

Accurate segregation of retinal blood vessels network plays a crucial role in clinical assessments, treatments, and rehabilitation process. Owing to the presence of acquisition and instrumentation anomalies, precise tracking of vessels network is challenging. For this, a new fundus image segmentation framework is proposed by combining deep neural networks, and hidden Markov model. It has three main modules: the Atrous spatial pyramid pooling-based encoder, the decoder, and hidden Markov model vessel tracker. The encoder utilized modified ResNet18 deep neural networks model for low-and-high-levels features extraction. These features are concatenated in module-II by the decoder to perform convolution operations to obtain the initial segmentation. Previous modules detected the main vessel structure and overlooked some small capillaries. For improved segmentation, hidden Markov model vessel tracker is integrated with module-I and-II to detect overlooked small capillaries of the vessels network. In last module, final segmentation is obtained by combining multi-oriented sub-images using logical OR operation. This novel framework is validated experimentally using two standard DRIVE and STARE datasets. The developed model offers high average values of accuracy, area under the curve, and sensitivity of 99.8, 99.0, and 98.2%, respectively. Analysis of the results revealed that the developed approach offered enhanced performance in terms of sensitivity 18%, accuracy 3%, and specificity 1% over the state-of-the-art approaches. Owing to better learning and generalization capability, the developed approach tracked blood vessels network efficiently and automatically compared to other approaches. The proposed approach can be helpful for human eye assessment, disease diagnosis, and rehabilitation process.

Long non-coding RNAs (lncRNAs) are important regulators of biological processes. It has recently been shown that some lncRNAs include small open reading frames (sORFs) that can encode small peptides of no more than 100 amino acids. However, existing methods are commonly applied to human and animal datasets and still suffer from low feature representation capability. Thus, accurate and credible prediction of sORFs with coding ability in plant lncRNAs is imperative. This paper proposes a new method termed sORFPred, in which we design a model named MCSEN by combining multi-scale convolution and Squeeze-and-Excitation Networks to fully mine distinct information embedded in sORFs, integrate and optimize multiple sequence-based and physicochemical feature descriptors, and built a two-layer prediction classifier based on Bayesian optimization algorithm and Extra Trees. sORFPred has been evaluated on sORFs datasets of three species and experimentally validated sORFs dataset. Results indicate that sORFPred outperforms existing methods and achieves 97.28% accuracy, 97.06% precision, 97.52% recall, and 97.29% F1-score on Arabidopsis thaliana, which shows a significant improvement in prediction performance compared to various conventional shallow machine learning and deep learning models.

Differentiation of ductal carcinoma in situ (DCIS, a precancerous lesion of the breast) from fibroadenoma (FA) using ultrasonography is significant for the early prevention of malignant breast tumors. Radiomics-based artificial intelligence (AI) can provide additional diagnostic information but usually requires extensive labeling efforts by clinicians with specialized knowledge. This study aims to investigate the feasibility of differentially diagnosing DCIS and FA using ultrasound radiomics-based AI techniques and further explore a novel approach that can reduce labeling efforts without sacrificing diagnostic performance. We included 461 DCIS and 651 FA patients, of whom 139 DCIS and 181 FA patients constituted a prospective test cohort. First, various feature engineering-based machine learning (FEML) and deep learning (DL) approaches were developed. Then, we designed a difference-based self-supervised (DSS) learning approach that only required FA samples to participate in training. The DSS approach consists of three steps: (1) pretraining a Bootstrap Your Own Latent (BYOL) model using FA images, (2) reconstructing images using the encoder and decoder of the pretrained model, and (3) distinguishing DCIS from FA based on the differences between the original and reconstructed images. The experimental results showed that the trained FEML and DL models achieved the highest AUC of 0.7935 (95% confidence interval, 0.7900-0.7969) on the prospective test cohort, indicating that the developed models are effective for assisting in differentiating DCIS from FA based on ultrasound images. Furthermore, the DSS model achieved an AUC of 0.8172 (95% confidence interval, 0.8124-0.8219), indicating that our model outperforms the conventional radiomics-based AI models and is more competitive.

Drug-drug interaction refers to taking the two drugs may produce certain reaction which may be a threat to patients' health, or enhance the efficacy helpful for medical work. Therefore, it is necessary to study and predict it. In fact, traditional experimental methods can be used for drug-drug interaction prediction, but they are time-consuming and costly, so we prefer to use more accurate and convenient calculation methods to predict the unknown drug-drug interaction. In this paper, we proposed a deep learning framework called MSResG that considers multi-sources features of drugs and combines them with Graph Auto-Encoder to predicting. Firstly, the model obtains four feature representations of drugs from the database, namely, chemical substructure, target, pathway and enzyme, and then calculates the Jaccard similarity of the drugs. To balance different drug features, we perform similarity integration by finding the mean value. Then we will be comprehensive similarity network combined with drug interaction network, and encodes and decodes it using the graph auto-encoder based on residual graph convolution network. Encoding is to learn the potential feature vectors of drugs, which contain similar information and interaction information. Decoding is to reconstruct the network to predict unknown drug-drug interaction. The experimental results show that our model has advanced performance and is superior to other existing advanced methods. Case study also shows that MSResG has practical significance.

Accurately detecting linear B-cell epitopes (BCEs) makes great sense in vaccine design, immunodiagnostic test, antibody production, disease prevention and treatment. Wet-lab experiments for determining linear BCEs are both expensive and laborious, which are not able to meet the recognition needs of modern massive protein sequence data. Instead, computational methods can efficiently identify linear BCEs with low cost. Although several computational methods are available, the performance is still not satisfactory. Thus, we propose a new method, LBCE-XGB, to forecast linear BCEs based on XGBoost algorithm. To represent the biological information concealed in peptide sequences, the embeddings of the residues were obtained from a pre-trained domain-specific BERT model. In addition, the other five types of attributes comprising amino acid composition, amino acid antigenicity scale were also extracted. The best feature combination was determined according to the cross-validation results. Against the models developed by other deep learning and machine learning algorithms, LBCE-XGB achieves the top performance with an AUROC of 0.845 for fivefold cross-validation. The results on the independent test set show that our model attains an AUROC of 0.838 which is substantially higher than other state-of-the-art methods. The outcomes indicate that the representations of BERT could be an effective feature in predicting linear BCEs and we believe that LBCE-XGB could be a useful medium for detecting linear B cell epitopes with high accuracy and low cost.

This review article summarizes the recent advances in the diagnostic studies of autism spectrum disorders (ASDs) considering some of the most influential research articles from the last two decades. ASD is a heterogeneous neurodevelopmental disorder characterized by abnormalities in social interaction, communication, and behavioral patterns as well as some unique strengths and differences. The current diagnosis systems are based on autism diagnostic observation schedule (ADOS) or autism diagnostic interview-revised (ADI-R), but biological markers are also important for an effective diagnosis of ASDs. The amalgamation of neuroimaging techniques, such as structural and functional magnetic resonance imaging (sMRI and fMRI), with machine-learning and deep-learning approaches helps throw new light on typical biological markers of ASDs at the early stage of life. To assess the performance of a deep neural network, we develop a light-weighted CNN model for ASD classification. The overall accuracy, precision, and F1-score of the proposed model are 99.92%, 99.93% and 99.92%, respectively. All the neuroimaging studies we have reviewed can be divided into 3 categories, viz. thickness, volume and functional connectivity-based studies. We conclude with a discussion of the major findings of considered studies and promising directions for future research in this field.

As a well-established technique, DNA synthesis offers interesting possibilities for designing multifunctional nanodevices. The micro-processing system of modern semiconductor circuits is dependent on strategies organized on silicon chips to achieve the speedy transmission of substances or information. Similarly, spatially localized structures allow for fixed DNA molecules in close proximity to each other during the synthesis of molecular circuits, thus providing a different strategy that of opening up a remarkable new area of inquiry for researchers. Herein, the Visual DSD (DNA strand displacement) modeling language was used to design and analyze the spatially organized DNA reaction network. The execution rules depend on the hybridization reaction caused by directional complementary nucleotide sequences. A series of DNA strand displacement calculations were organized on the locally coded travel track, and autonomous movement and addressing operations are gradually realized. The DNA nanodevice operates in this manner follows the embedded "molecular program", which improves the reusability and scalability of the same sequence domain in different contexts. Through the communication between various building blocks, the DNA device-carrying the target molecule moves in a controlled manner along the programmed track. In this way, a variety of molecular functional group transport and specific partition storage can be realized. The simulation results of the visual DSD tool provide qualitative and quantitative proof for the operation of the system.

Stroke is still the World's second major factor of death, as well as the third major factor of death and disability. Ischemic stroke is a type of stroke, in which early detection and treatment are the keys to preventing ischemic strokes. However, due to the limitation of privacy protection and labeling difficulties, there are only a few studies on the intelligent automatic diagnosis of stroke or ischemic stroke, and the results are unsatisfactory. Therefore, we collect some data and propose a 3D carotid Computed Tomography Angiography (CTA) image segmentation model called CA-UNet for fully automated extraction of carotid arteries. We explore the number of down-sampling times applicable to carotid segmentation and design a multi-scale loss function to resolve the loss of detailed features during the process of down-sampling. Moreover, based on CA-Unet, we propose an ischemic stroke risk prediction model to predict the risk in patients using their 3D CTA images, electronic medical records, and medical history. We have validated the efficacy of our segmentation model and prediction model through comparison tests. Our method can provide reliable diagnoses and results that benefit patients and medical professionals.

DNA methylation is an epigenetic alteration that plays a fundamental part in governing gene regulatory processes. The DNA methylation mechanism affixes methyl groups to distinct cytosine residues, influencing chromatin architectures. Multiple studies have demonstrated that DNA methylation's regulatory effect on genes is linked to the beginning and progression of several disorders. Researchers have recently uncovered thousands of phenotype-related methylation sites through the epigenome-wide association study (EWAS). However, combining the methylation levels of several sites within a gene and determining the gene-level DNA methylation remains challenging. In this study, we proposed the supervised UMAP Assisted Gene-level Methylation method (sUAGM) for disease prediction based on supervised UMAP (Uniform Manifold Approximation and Projection), a manifold learning-based method for reducing dimensionality. The methylation values at the gene level generated using the proposed method are evaluated by employing various feature selection and classification algorithms on three distinct DNA methylation datasets derived from blood samples. The performance has been assessed employing classification accuracy, F-1 score, Mathews Correlation Coefficient (MCC), Kappa, Classification Success Index (CSI) and Jaccard Index. The Support Vector Machine with the linear kernel (SVML) classifier with Recursive Feature Elimination (RFE) performs best across all three datasets. From comparative analysis, our method outperformed existing gene-level and site-level approaches by achieving 100% accuracy and F1-score with fewer genes. The functional analysis of the top 28 genes selected from the Parkinson's disease dataset revealed a significant association with the disease.

Complex diseases are generally caused by disorders of biological networks and/or mutations in multiple genes. Comparisons of network topologies between different disease states can highlight key factors in their dynamic processes. Here, we propose a differential modular analysis approach that integrates protein-protein interactions with gene expression profiles for modular analysis, and introduces inter-modular edges and date hubs to identify the "core network module" that quantifies the significant phenotypic variation. Then, based on this core network module, key factors, including functional protein-protein interactions, pathways, and driver mutations, are predicted by the topological-functional connection score and structural modeling. We applied this approach to analyze the lymph node metastasis (LNM) process in breast cancer. The functional enrichment analysis showed that both inter-modular edges and date hubs play important roles in cancer metastasis and invasion, and in metastasis hallmarks. The structural mutation analysis suggested that the LNM of breast cancer may be the outcome of the dysfunction of rearranged during transfection (RET) proto-oncogene-related interactions and the non-canonical calcium signaling pathway via an allosteric mutation of RET. We believe that the proposed method can provide new insights into disease progression such as cancer metastasis.