Myocardial infarction detection method based on the continuous T-wave area feature and multi-lead-fusion deep features.

OBJECTIVE: Myocardial infarction (MI) is one of the most threatening cardiovascular diseases. This paper aims to explore a method for using an algorithm to autonomously classify myocardial infarction based on the electrocardiogram (ECG). APPROACH: A detection method of MI that fuses continuous T-wave area (C_TWA) feature and ECG deep features is proposed. This method consists of three main parts: (1) The onset of MI is often accompanied by changes in the shape of the T-wave in the ECG, thus the area of the T-wave displayed on different heartbeats will be quite different. The adaptive sliding window method is used to detect the start and end of the T-wave, and calculate the C_TWA on the same ECG record. Additionally, the coefficient of variation (CV) of C_TWA is defined as the C_TWA feature of the ECG. (2) The multi lead fusion convolutional neural network (Multi-lead-fusion CNN) was implemented to extract the deep features of the ECG. (3) The C_TWA feature and deep features of the ECG were fused by soft attention, and then inputted into the multi-layer perceptron to obtain the detection result. RESULTS: According to the interpatient paradigm, the proposed method reached a 97.67% accuracy, 96.59% precision, and 98.96% recall on the PTB dataset, whereas the proposed method reached 93.15% accuracy, 93.20% precision, and 95.14% recall on the clinical dataset. SIGNIFICANCE: The proposed method accurately extracts the feature of the C_TWA, and combines the deep features of the signal, thereby improving the detection accuracy and achieving ideal results on clinical datasets.

Time-gated single-pixel imaging of Cherenkov emission from a medical linear accelerator.

Cherenkov imaging is an ideal tool for real-time in vivo verification of a radiation therapy dose. Given that radiation is pulsed from a medical linear accelerator (LINAC) together with weak Cherenkov emissions, time-gated high-sensitivity imaging is required for robust measurements. Instead of using an expensive camera system with limited efficiency of detection in each pixel, a single-pixel imaging (SPI) approach that maintains promising sensitivity over the entire spectral band could be used to provide a low-cost and viable alternative. A prototype SPI system was developed and demonstrated here in Cherenkov imaging of LINAC dose delivery to a water tank. Validation experiments were performed using four regular fields and an intensity-modulated radiotherapy (IMRT) delivery plan. The Cherenkov image-based projection percent depth dose curves (pPDDs) were compared to pPDDs simulated by the treatment planning system (TPS), with an overall average error of 0.48, 0.42, 0.65, and 1.08% for the 3, 5, 7, and 9 cm square beams, respectively. The composite image of the IMRT plan achieved a 85.9% pass rate using 3%/3 mm gamma index criteria, in comparing Cherenkov intensity and TPS dose. This study validates the feasibility of applying SPI to the Cherenkov imaging of radiotherapy dose for the first time to our knowledge.

Microbial Metagenomes and Host Transcriptomes Reveal the Dynamic Changes of Rumen Gene Expression, Microbial Colonization and Co-Regulation of Mineral Element Metabolism in Yaks from Birth to Adulthood.

Yaks are the main pillar of plateau animal husbandry and the material basis of local herdsmen's survival. The level of mineral elements in the body is closely related to the production performance of yaks. In this study, we performed a comprehensive analysis of rumen epithelial morphology, transcriptomics and metagenomics to explore the dynamics of rumen functions, microbial colonization and functional interactions in yaks from birth to adulthood. Bacteria, eukaryotes, archaea and viruses colonized the rumen of yaks from birth to adulthood, with bacteria being the majority. Bacteroidetes and Firmicutes were the dominant phyla in five developmental stages, and the abundance of genus Lactobacillus and Fusobacterium significantly decreased with age. Glycoside hydrolase (GH) genes were the most highly represented in five different developmental stages, followed by glycosyltransferases (GTs) and carbohydrate-binding modules (CBMs), where the proportion of genes coding for CBMs increased with age. Integrating host transcriptome and microbial metagenome revealed 30 gene modules related to age, muscle layer thickness, nipple length and width of yaks. Among these, the MEmagenta and MEturquoise were positively correlated with these phenotypic traits. Twenty-two host genes involved in transcriptional regulation related to metal ion binding (including potassium, sodium, calcium, zinc, iron) were positively correlated with a rumen bacterial cluster 1 composed of Alloprevotella, Paludibacter, Arcobacter, Lactobacillus, Bilophila, etc. Therefore, these studies help us to understand the interaction between rumen host and microorganisms in yaks at different ages, and further provide a reliable theoretical basis for the development of feed and mineral element supplementation for yaks at different ages.

scCompressSA: dual-channel self-attention based deep autoencoder model for single-cell clustering by compressing gene-gene interactions.

BACKGROUND: Single-cell clustering has played an important role in exploring the molecular mechanisms about cell differentiation and human diseases. Due to highly-stochastic transcriptomics data, accurate detection of cell types is still challenged, especially for RNA-sequencing data from human beings. In this case, deep neural networks have been increasingly employed to mine cell type specific patterns and have outperformed statistic approaches in cell clustering. RESULTS: Using cross-correlation to capture gene-gene interactions, this study proposes the scCompressSA method to integrate topological patterns from scRNA-seq data, with support of self-attention (SA) based coefficient compression (CC) block. This SA-based CC block is able to extract and employ static gene-gene interactions from scRNA-seq data. This proposed scCompressSA method has enhanced clustering accuracy in multiple benchmark scRNA-seq datasets by integrating topological and temporal features. CONCLUSION: Static gene-gene interactions have been extracted as temporal features to boost clustering performance in single-cell clustering For the scCompressSA method, dual-channel SA based CC block is able to integrate topological features and has exhibited extraordinary detection accuracy compared with previous clustering approaches that only employ temporal patterns.

Data governance and Gensini score automatic calculation for coronary angiography with deep-learning-based natural language extraction.

With the widespread adoption of electronic health records, the amount of stored medical data has been increasing. Clinical data, often in the form of semi-structured or unstructured electronic medical records (EMRs), contains rich patient information. However, due to the use of natural language by physicians when composing these records, the effectiveness of traditional methods such as dictionaries, rule matching, and machine learning in the extraction of information from these unstructured texts falls short of clinical standards. In this paper, a novel deep-learning-based natural language extraction method is proposed to overcome current shortcomings in data governance and Gensini score automatic calculation in coronary angiography. A pre-trained model called bidirectional encoder representation from transformers (BERT) with strong text feature representation capabilities is employed as the feature representation layer. It is combined with bidirectional long short-term memory (BiLSTM) and conditional random field (CRF) models to extract both global and local features from the text. The study included an evaluation of the model on a dataset from a hospital in China and it was compared with another model to validate its practical advantages. Hence, the BiLSTM-CRF model was employed to automatically extract relevant coronary angiogram information from EMR texts. The achieved F1 score was 91.19, which is approximately 0.87 higher than the BERT-BiLSTM-CRF model.

Incongruence between confirmed and suspected clinical cases of Japanese encephalitis virus infection.

Background: Japanese encephalitis (JE) is a notifiable infectious disease in China. Information on every case of JE is reported to the superior health administration department. However, reported cases include both laboratory-confirmed and clinically diagnosed cases. This study aimed to differentiate between clinical and laboratory-confirmed cases of Japanese encephalitis virus (JEV) infection, and improve the accuracy of reported JE cases by analyzing the acute-phase serum and cerebrospinal fluid of all reported JE cases in the Sichuan province from 2012 to 2022. Methods: All acute-phase serum and/or cerebrospinal fluid samples of the reported JE cases were screened for IgM(ImmunoglobulinM)to JEV using the enzyme-linked immunosorbent assay (ELISA), and the detection of the viral genes of JEV and 9 other pathogens including enterovirus (EV), using reverse transcription PCR was attempted. Epidemiological analyses of JE and non-JE cases based on sex, age, onset time, and geographical distribution were also performed. Results: From 2012 to 2022, 1558 JE cases were reported in the Sichuan province. The results of serological (JEV-specific IgM) and genetic testing for JEV showed that 81% (1262/1558) of the reported cases were confirmed as JEV infection cases (laboratory-confirmed cases). Among the 296 cases of non-JEV infection, 6 viruses were detected in the cerebrospinal fluid in 62 cases, including EV and the Epstein-Barr virus (EBV), constituting 21% (62/296) of all non-JE cases. Among the 62 non-JEV infection cases with confirmed pathogens, infections with EV and EBV included 17 cases each, herpes simplex virus (HSV-1/2) included 14 cases, varicella- zoster virus included 6 cases, mumps virus included 2 cases, and human herpes viruses-6 included 1 case. Additionally, there were five cases involving mixed infections (two cases of EV/EBV, one case of HSV-1/HSV-2, one case of EBV/HSV-1, and one case of EV/herpes viruses-6). The remaining 234 cases were classified as unknown viral encephalitis cases. Our analysis indicated that those aged 0-15 y were the majority of the patients among the 1558 reported JE cases. However, the incidence of laboratory-confirmed JE cases in the >40 y age group has increased in recent years. The temporal distribution of laboratory-confirmed cases of JE revealed that the majority of cases occurred from May to September each year, with the highest incidence in August. Conclusion: The results of this study indicate that there is a certain discrepancy between clinically diagnosed and laboratory-confirmed cases of JE. Each reported case should be based on laboratory detection results, which is of great importance in improving the accuracy of case diagnosis and reducing misreporting. Our results are not only important for addressing JE endemic to the Sichuan province, but also provide a valuable reference for the laboratory detection of various notifiable infectious diseases in China and other regions outside China.

Transcriptome profiling in rumen, reticulum, omasum, and abomasum tissues during the developmental transition of pre-ruminant to the ruminant in yaks.

The development of the four stomachs of yak is closely related to its health and performance, however the underlying molecular mechanisms are largely unknown. Here, we systematically analyzed mRNAs of four stomachs in five growth time points [0 day, 20 days, 60 days, 15 months and 3 years (adult)] of yaks. Overall, the expression patterns of DEmRNAs were unique at 0 d, similar at 20 d and 60 d, and similar at 15 m and adult in four stomachs. The expression pattern in abomasum was markedly different from that in rumen, reticulum and omasum. Short Time-series Expression Miner (STEM) analysis demonstrated that multi-model spectra are drastically enriched over time in four stomachs. All the identified mRNAs in rumen, reticulum, omasum and abomasum were classified into 6, 4, 7, and 5 cluster profiles, respectively. Modules 9, 38, and 41 were the most significant three colored modules. By weighted gene co-expression network analysis (WGCNA), a total of 5,486 genes were categorized into 10 modules. CCKBR, KCNQ1, FER1L6, and A4GNT were the hub genes of the turquoise module, and PAK6, TRIM29, ADGRF4, TGM1, and TMEM79 were the hub genes of the blue module. Furthermore, functional KEGG enrichment analysis suggested that the turquoise module was involved in gastric acid secretion, sphingolipid metabolism, ether lipid metabolism, etc., and the blue module was enriched in pancreatic secretion, pantothenate and CoA biosynthesis, and starch and sucrose metabolism, etc. Our study aims to lay a molecular basis for the study of the physiological functions of rumen, reticulum, omasum and abomasum in yaks. It can further elucidate the important roles of these mRNAs in regulation of growth, development and metabolism in yaks, and to provide a theoretical basis for age-appropriate weaning and supplementary feeding in yaks.

SSA Net: Small Scale-Aware Enhancement Network for Human Pose Estimation.

In the field of human pose estimation, heatmap-based methods have emerged as the dominant approach, and numerous studies have achieved remarkable performance based on this technique. However, the inherent drawbacks of heatmaps lead to serious performance degradation in methods based on heatmaps for smaller-scale persons. While some researchers have attempted to tackle this issue by improving the performance of small-scale persons, their efforts have been hampered by the continued reliance on heatmap-based methods. To address this issue, this paper proposes the SSA Net, which aims to enhance the detection accuracy of small-scale persons as much as possible while maintaining a balanced perception of persons at other scales. SSA Net utilizes HRNetW48 as a feature extractor and leverages the TDAA module to enhance small-scale perception. Furthermore, it abandons heatmap-based methods and instead adopts coordinate vector regression to represent keypoints. Notably, SSA Net achieved an AP of 77.4% on the COCO Validation dataset, which is superior to other heatmap-based methods. Additionally, it achieved highly competitive results on the Tiny Validation and MPII datasets as well.

Genome Variation Map of Domestic Qinghai-Tibet Plateau Yaks by SLAF-Seq Reveals Genetic Footprint during Artificial Selection.

The yak (Bos grunniens) was domesticated in the high-altitude QTP. Research about their genetic diversity and population structure is limited. In this study, we resequenced the genome of 494 domestic yaks using Specific-Locus Amplified Fragment Sequencing (SLAF-seq). The survey was conducted on six populations sampled from isolated locations in China in order to analyze their structure and genetic diversity. These six domestic populations were clearly grouped into two independent clusters, with Jinchuan, Changtai, and Jiulong showing a tight genetic relationship with the wild yak. Nerve development pathways were enriched with GO enrichment analysis of 334 domesticated genes. Major genomic regions associated with the differentiation of domestic yaks were detected. These findings provide preliminary information on the yak genome variability, useful to understand the genomic characteristics of different populations in QTP.

An efficient ECG denoising method by fusing ECA-Net and CycleGAN.

For wearable electrocardiogram (ECG) acquisition, it was easy to infer motion artifices and other noises. In this paper, a novel end-to-end ECG denoising method was proposed, which was implemented by fusing the Efficient Channel Attention (ECA-Net) and the cycle consistent generative adversarial network (CycleGAN) method. The proposed denoising model was optimized by using the ECA-Net method to highlight the key features and introducing a new loss function to further extract the global and local ECG features. The original ECG signal came from the MIT-BIH Arrhythmia Database. Additionally, the noise signals used in this method consist of a combination of Gaussian white noise and noises sourced from the MIT-BIH Noise Stress Test Database, including EM (Electrode Motion Artifact), BW (Baseline Wander) and MA (Muscle Artifact), as well as mixed noises composed of EM+BW, EM+MA, BW+MA and EM+BW+MA. Moreover, corrupted ECG signals were generated by adding different levels of single and mixed noises to clean ECG signals. The experimental results show that the proposed method has better denoising performance and generalization ability with higher signal-to-noise ratio improvement (SNRimp), as well as lower root-mean-square error (RMSE) and percentage-root-mean-square difference (PRD).

Dynamic Profile of the Yak Mammary Transcriptome during the Lactation Cycle.

The objective of this study was to assess the transcriptome of the mammary tissue of four yaks during the whole lactation cycle. For this purpose, biopsies of the mammary gland were performed at -30, -15, 1, 15, 30, 60, 120, 180, and 240 days relative to parturition (d). The transcriptome analysis was performed using a commercial bovine microarray platform and the results were analyzed using several bioinformatic tools. The statistical analysis using an overall false discovery rate ≤ 0.05 for the effect of whole lactation and p < 0.05 for each comparison identified >6000 differentially expressed genes (DEGs) throughout lactation, with a large number of DEGs observed at the onset (1 d vs. -15 d) and at the end of lactation (240 d vs. 180 d). Bioinformatics analysis revealed a major role of genes associated with BTA3, BTA4, BTA6, BTA9, BTA14, and BTA28 in lactation. Functional analysis of DEG underlined an overall induction of lipid metabolism, suggesting an increase in triglycerides synthesis, likely regulated by PPAR signaling. The same analysis revealed an induction of amino acid metabolism and secretion of protein, with a concomitant decrease in proteasome, indicating a major role of amino acid handling and reduced protein degradation in the synthesis and secretion of milk proteins. Glycan biosynthesis was induced for both N-glycan and O-glycan, suggesting increased glycan content in the milk. The cell cycle and immune response, especially antigen processing and presentation, were strongly inhibited during lactation, suggesting that morphological changes are minimized during lactation, while the mammary gland prevents immune hyper-response. Transcripts associated with response to radiation and low oxygen were enriched in the down-regulated DEG affected by the stage of lactation. Except for this last finding, the functions affected by the transcriptomic adaptation to lactation in mammary tissue of yak are very similar to those observed in dairy cows.

Erratum: [Corrigendum] Interpretation of HbA1c lies at the intersection of analytical methodology, clinical biochemistry and hematology (Review).

[This corrects the article DOI: 10.3892/etm.2022.11643.].

HFIST-Net: High-throughput fast iterative shrinkage thresholding network for accelerating MR image reconstruction.

BACKGROUND AND OBJECTIVES: Compressed sensing (CS) is often used to accelerate magnetic resonance image (MRI) reconstruction from undersampled k-space data. A novelty deeply unfolded networks (DUNs) based method, designed by unfolding a traditional CS-MRI optimization algorithm into deep networks, can provide significantly faster reconstruction speeds than traditional CS-MRI methods while improving image quality. METHODS: In this paper, we propose a High-Throughput Fast Iterative Shrinkage Thresholding Network (HFIST-Net) for reconstructing MR images from sparse measurements by combining traditional model-based CS techniques and data-driven deep learning methods. Specifically, the conventional Fast Iterative Shrinkage Thresholding Algorithm (FISTA) method is expanded as a deep network. To break the bottleneck of information transmission, a multi-channel fusion mechanism is proposed to improve the efficiency of information transmission between adjacent network stages. Moreover, a simple yet efficient channel attention block, called Gaussian context transformer (GCT), is proposed to improve the characterization capabilities of deep Convolutional Neural Network (CNN,) which utilizes Gaussian functions that satisfy preset relationships to achieve context feature excitation. RESULTS: T1 and T2 brain MR images from the FastMRI dataset are used to validate the performance of the proposed HFIST-Net. The qualitative and quantitative results showed that our method is superior to those compared state-of-the-art unfolded deep learning networks. CONCLUSIONS: The proposed HFIST-Net is capable of reconstructing more accurate MR image details from highly undersampled k-space data while maintaining fast computational speed.

Transcriptome-Based Evaluation of Optimal Reference Genes for Quantitative Real-Time PCR in Yak Stomach throughout the Growth Cycle.

Efficient nutritional assimilation and energy metabolism in the stomachs of yaks contribute to their adaption to harsh environments. Accurate gene expression profile analysis will help further reveal the molecular mechanism of nutrient and energy metabolism in the yak stomach. RT-qPCR is regarded as an accurate and dependable method for analyzing gene expression. The selection of reference genes is essential to obtain meaningful RT-qPCR results, especially in longitudinal gene expression studies of tissues and organs. Our objective was to select and validate optimal reference genes from across the transcriptome as internal controls for longitudinal gene expression studies in the yak stomach. In this study, 15 candidate reference genes (CRGs) were determined according to transcriptome sequencing (RNA-seq) results and the previous literature. The expression levels of these 15 CRGs were quantified using RT-qPCR in the yak stomach, including the rumen, reticulum, omasum and abomasum at five stages: 0 days, 20 days, 60 days, 15 months and three years old (adult). Subsequently, the expression stabilities of these 15 CRGs were evaluated via four algorithms: geNorm, NormFinder, BestKeeper and the comparative CT method. Furthermore, RefFinder was employed to obtain a comprehensive ranking of the stability of CRGs. The analysis results indicate that RPS15, MRPL39 and RPS23 are the most stable genes in the yak stomach throughout the growth cycle. In addition, to verify the reliability of the selected CRGs, the relative expression levels of HMGCS2 were quantified via RT-qPCR using the three most stable or the three least stable CRGs. Overall, we recommend combining RPS15, MRPL39 and RPS23 as reference genes for the normalization of RT-qPCR data in the yak stomach throughout the growth cycle.

Partitioned estimation methodology of biological neuronal networks with topology-based module detection.

Parameter estimation of neuronal networks is closely related with information processing mechanisms in neural systems. Estimation of synaptic parameters for neuronal networks was an time consuming task. Due to complex interactions between neurons, computational efficiency and accuracy of estimation methods is relatively low. Meanwhile, inherent topological properties such as core-periphery and modular structures are not fully considered in estimation. In order to improve the efficiency and accuracy of estimation, this study proposes a two-stage PartitionMLE method which introduces detected neuronal modules as topological constraints in estimation. The proposed PartitionMLE method firstly decomposes the system into multiple non-overlapping neuronal modules, by performing topology-based module detection. Dynamic parameters including intra-modular and inter-modular parameters are estimated in two stages, using detected hubs to connect non-overlapping neuronal modules. The contributions of PartitionMLE method are two-folds: reducing estimation errors and improving the model interpretability. Experiments about neuronal networks consisting of Hodgkin-Huxley (HH) and leaky integrate-and-firing (LIF) neurons validated the effectiveness of the PartitionMLE method, with comparison to the single-stage MLE method.

Deep arrhythmia classification based on SENet and lightweight context transform.

Arrhythmia is one of the common cardiovascular diseases. Nowadays, many methods identify arrhythmias from electrocardiograms (ECGs) by computer-aided systems. However, computer-aided systems could not identify arrhythmias effectively due to various the morphological change of abnormal ECG data. This paper proposes a deep method to classify ECG samples. Firstly, ECG features are extracted through continuous wavelet transform. Then, our method realizes the arrhythmia classification based on the new lightweight context transform blocks. The block is proposed by improving the linear content transform block by squeeze-and-excitation network and linear transformation. Finally, the proposed method is validated on the MIT-BIH arrhythmia database. The experimental results show that the proposed method can achieve a high accuracy on arrhythmia classification.

Interpretation of HbA1c lies at the intersection of analytical methodology, clinical biochemistry and hematology (Review).

Over past few decades, diabetes has become widespread on a global scale. Hemoglobin A1c (HbA1c) assessment is crucial for diabetes care, since it allows for the monitoring of an individual's level of glycemic control over the course of 2 to 3 months and risk assessment to determine any possible complications. Numerous methods, including cation-exchange chromatography, electrophoresis, immunoassays and affinity chromatography, can be used to determine the HbA1c level. Each method has its limitations, however. The amount of HbA1c in patient samples is not only dependent on blood glucose levels, but is also strongly influenced by changes in red blood cell lifespan and globin chain structure. Consequently, hematological, clinical biochemistry and analytical methods all intertwine when interpreting HbA1c. There are numerous reports on the interactions of HbA1c with inherited and acquired diseases. Some of these impacts are inconsistent and difficult to explain. The present review article aimed to summarize and classify these effects and evaluate their clinical relevance. The findings discussed herein may serve as a reminder that clinical HbA1c values need to be analyzed with caution.

Identification of a rare compound heterozygous hemoglobin variant ß0-thal [ß17(A14) Lys>Stop, HBB: c.52A>T] and Hb J-Lome [ß59(E3) Lys>Asn, HBB: c.180G>C].

BACKGROUND: HbA1c is the validated biomarker for glycemic management in diabetic individuals. Here, we report a compound heterozygote for ß0-thal and Hb J-Lomeand evaluate its effect on HbA1c measurements. METHODS: A 51-year-old female was suspected of harboring a hemoglobin variant following no value of HbA1c levelby Arkray HA-8180 V (48s HbA1c mode), abnormal hematological data, and abnormalhemoglobin analysison capillary electrophoresis (Capillarys 2 Flex Piercing, Hb program). Sanger sequencing of the α and ß genes was subsequently performed on the proband.HbA1c was reanalyzed using D10 (Bio-Rad), Capillarys 2 Flex Piercing (Sebia), and Roche Cobas c501 (Roche Diagnostics). RESULTS: Sanger sequencing identified a compound heterozygote for ß0-thal [ß17(A14) Lys > Stop, HBB: c.52A > T] and Hb J-Lome [ß59(E3) Lys > Asn, HBB: c.180G > C].HbA1c values ⁣⁣determinedby D10, Capillarys 2 Flex Piercing (HbA1c program), and Roche Cobas c501were 2.3%, no HbA1c value, and 5.1 (32 mmol/mol), respectively. During pedigree analysis, the son of the proband was found to have normal blood glucose (5.55 mmol/L), decreased HbA1c (3.6%, 16 mmol/mol)by Arkray HA-8180 V (48s HbA1c mode), an abnormal band on the electrophoretogram of Capillarys2 (Hb program), and the Hb J-Lome mutation in the ß globin gene.Subsequently, HbA1c values ⁣⁣determinedby D10, Capillarys 2 Flex Piercing (HbA1c program), and Roche Cobas c501 were4.0% (20 mmol/mol), no HbA1c value, and 5.0 (31 mmol/mol), respectively. CONCLUSION: Atypically low HbA1c levels or a discrepancy between blood glucose and HbA1c levels should raise concerns about hemoglobin variations.

Quantifying the impact of Pyramid Squeeze Attention mechanism and filtering approaches on Alzheimer's disease classification.

Brain medical imaging and deep learning are important foundations for diagnosing and predicting Alzheimer's disease. In this study, we explored the impact of different image filtering approaches and Pyramid Squeeze Attention (PSA) mechanism on the image classification of Alzheimer's disease. First, during the image preprocessing, we register MRI images and remove skulls, then apply median filtering, Gaussian blur filtering, and anisotropic diffusion filtering to obtain different experimental images. After that, we add the Squeeze and Excitation (SE) mechanism and Pyramid Squeeze Attention (PSA) mechanism to the Fully Convolutional Network (FCN) model respectively, to obtain each MRI image's corresponding feature information of disease probability map. Besides, we also construct Multi-Layer Perceptron (MLP) model's framework, combining feature information of disease probability map with age, gender, and Mini-Mental State Examination (MMSE) of each sample, to get the final classification performance of model. Among them, the accuracy of the MLP-C model combining anisotropic diffusion filtering with the Pyramid Squeeze Attention mechanism can reach 98.85%. The corresponding quantitative experimental results show that different image filtering approaches and attention mechanisms provide effective assistance for the diagnosis and classification of Alzheimer's disease.

Prediction of fluid intelligence from T1-w MRI images: A precise two-step deep learning framework.

The Adolescent Brain Cognitive Development (ABCD) Neurocognitive Prediction Challenge (ABCD-NP-Challenge) is a community-driven competition that challenges competitors to develop algorithms to predict fluid intelligence scores from T1-w MRI images. In this work, a two-step deep learning pipeline is proposed to improve the prediction accuracy of fluid intelligence scores. In terms of the first step, the main contributions of this study include the following: (1) the concepts of the residual network (ResNet) and the squeeze-and-excitation network (SENet) are utilized to improve the original 3D U-Net; (2) in the segmentation process, the pixels in symmetrical brain regions are assigned the same label; (3) to remove redundant background information from the segmented regions of interest (ROIs), a minimum bounding cube (MBC) is used to enclose the ROIs. This new segmentation structure can greatly improve the segmentation performance of the ROIs in the brain as compared with the classical convolutional neural network (CNN), which yields a Dice coefficient of 0.8920. In the second stage, MBCs are used to train neural network regression models for enhanced nonlinearity. The fluid intelligence score prediction results of the proposed method are found to be superior to those of current state-of-the-art approaches, and the proposed method achieves a mean square error (MSE) of 82.56 on a test data set, which reflects a very competitive performance.