The medical information engineering major has been set up for 20 years in China. This study aimed to outline the current status of medical information engineering education, analyze the curriculums in 11 representative universities, and suggest reasonable modules for the professional introduction course. Universities setting up the medical information engineering major were systematically surveyed through the official website of the Ministry of Education over a period from 2000 to 2025, and scales of enrollment in 2024 and quota in 2025 were collected from undergraduate admission websites. Talent training programs and curriculums of 11 representative universities were collected and analyzed to figure out the distributions of credits and numbers of professional courses, and word clouds were generated to visualize the high-frequency professional courses. In addition, a professional introduction course was designed according to the curriculums, and its teaching effects were evaluated by questionnaire investigation in years of 2023 and 2024. A total of 67 universities from different regions were approved to set up the major in different years, and 56 universities of them enrolled more than 4300 undergraduates in 2025. For the 11 representative universities, credits of all courses and professional courses were about 174.41 and 88.64, and the credit ratio of basic, core and direction courses was about 10:7:4. Five highlighted courses were respectively revealed by word clouds for all courses, basic courses, core courses and direction courses. Results of questionnaires demonstrated that the professional introduction course can promote the undergraduates in cognition, knowledge and abilities. The past and present of medical information engineering education in China are systematically outlined, and a well-designed professional introduction course is provided with good teaching effects. The methodology of this study can be extended to improve the curriculums and applied to other specialized majors for diverse universities.

Objective.To develop a deep learning framework for the efficient and accurate quantification of hepatic steatosis in whole-slide images (WSIs).Approach.Two retrospective histopathological WSI datasets were collected and graded. A deep learning framework was proposed by incorporating tile scoring and screening, fat droplet segmentation, and WSI-level fat fraction (FF) calculation. Specifically, tile score was calculated based on morphological and color features, and used to screen tiles for subsequent segmentation. Six segmentation models were trained and compared, and the model with the highest Dice coefficient was used to automatically segment fat droplets within tiles. Tile-level FFs were yielded and aggregated through weighted averaging to obtain WSI-level FFs. Performances of the tile scoring and screening strategy, segmentation models and the tile matrix size were evaluated, and agreement with expert-assessed steatosis grades.Main results.Optimal tile scoring parameters were determined as a weight of 0.54 and a decision threshold of 0.23, achieving an F1-score of 0.97 and an area under the curve of 0.94, effectively distinguishing tiles with and without fat droplets. UNet++ exhibited the best segmentation performance (Dice = 0.97). With an optimal tile matrix size of 2048 × 2048 pixels, the integration of the scoring and screening strategy improved processing efficiency 5.6-fold, reducing the mean WSI processing time from 141.5 s to 25.2 s, while WSI-level FFs agreed with expert-assessed steatosis grades with a mean absolute error of 0.44%.Significance.The proposed framework enables accurate and efficient hepatic steatosis quantification, demonstrating high concordance with expert assessment and offering a promising clinical solution.

PURPOSE:To calibrate the effect of hepatic steatosis on diffusion parameters, including ADC and pure diffusion coefficient ( D ), by Monte Carlo simulation and phantom studies. METHODS:In the simulation study, four types of virtual liver models were constructed by incorporating size, nearest neighbor distance and regional anisotropy of fat droplets. By simulating magnetic field, proton movement and phase accrual, DWI signals were synthesized and analyzed to predict ADC and D . Correlation analyses were performed to investigate relationships between fat fraction (FF) and diffusion parameters predictions. Bland Altman analysis was conducted to evaluate the effects of virtual liver model, fat susceptibility and proton diffusivity on these predictions. In the phantom study, a fat-water phantom was scanned to investigate the relationships between proton density fat fraction (PDFF) and diffusion parameters. RESULTS:The simulation study demonstrated that ADC and D predictions were linearly and negatively correlated with FFs ( r ≤ - 0.924 ), unaffected by virtual liver model and fat susceptibility ( p ≥ 0.756 ). Additionally, proton diffusivity posed positive effects on these predictions. Compared with in vivo calibration, simulations with the virtual liver model incorporating all criteria, fat susceptibility of 0.2 ppm and proton diffusivity of 0.96 μm2/ms can accurately predict the diffusion parameters. The phantom study indicated that ADC and D measurements were linearly and negatively correlated with PDFF ( r ≤ - 0.895 ). CONCLUSION:Hepatic steatosis decreases both diffusion parameters and should be calibrated in the DWI-based assessment of liver fibrosis.

In the field of spinal pathology, sagittal balance of the spine is usually judged by the spatial structure and morphology of pelvis, which can be represented by pelvic parameters. Pelvic parameters, including pelvic incidence, pelvic tilt and sacral slope, are therefore essential for the diagnosis and treatment of spinal disorders, however, it is a time-consuming and laborious procedure to measure these parameters by traditional methods. In this paper, an automatic measurement framework for pelvic CT images was proposed to calculate three-dimensional (3D) pelvic parameters with the support of deep learning technology. Pelvic images were first preprocessed, and 3D reconstruction was then performed to obtain 3D pelvic model by the Visualization Toolkit. DRINet was trained to segment the femoral head region in the pelvic images, and 3D sphere fitting was performed to locate the femoral heads. In addition, VGG16 was adopted to recognize images containing superior sacral endplate, and the plane growth algorithm was used to fit the plane so that the midpoint and normal vector of the superior sacral endplate could be obtained. Finally, 3D pelvic parameters were automatically calculated, and compared with manual measurements for 15 patients. The proposed framework automatically generated 3D pelvic models, and calculated two-dimensional (2D) and 3D pelvic parameters from continuous CT images. Experiments demonstrated that the framework can greatly speed up the calculation of pelvic parameters, and these parameters are accurate when compared with the manual measurements. In conclusion, the proposed framework demonstrates good performance on automatic pelvimetry measurement by incorporating deep learning technology, and can well replace the traditional methods for pelvic parameter measurement.

To develop Monte Carlo simulations to predict the relationship of R 2 * $$ {\mathrm{R}}_2^{\ast } $$ with liver fat content at 1.5 T and 3.0 T. For various fat fractions (FFs) from 1% to 25%, four types of virtual liver models were developed by incorporating the size and spatial distribution of fat droplets. Magnetic fields were then generated under different fat susceptibilities at 1.5 T and 3.0 T, and proton movement was simulated for phase accrual and MRI signal synthesis. The synthesized signal was fit to single-peak and multi-peak fat signal models for R 2 * $$ {\mathrm{R}}_2^{\ast } $$ and proton density fat fraction (PDFF) predictions. In addition, the relationships between R 2 * $$ {\mathrm{R}}_2^{\ast } $$ and PDFF predictions were compared with in vivo calibrations and Bland-Altman analysis was performed to quantitatively evaluate the effects of these components (type of virtual liver model, fat susceptibility, and fat signal model) on R 2 * $$ {\mathrm{R}}_2^{\ast } $$ predictions. A virtual liver model with realistic morphology of fat droplets was demonstrated, and R 2 * $$ {\mathrm{R}}_2^{\ast } $$ and PDFF values were predicted by Monte Carlo simulations at 1.5 T and 3.0 T. R 2 * $$ {\mathrm{R}}_2^{\ast } $$ predictions were linearly correlated with PDFF, while the slope was unaffected by the type of virtual liver model and increased as fat susceptibility increased. Compared with in vivo calibrations, the multi-peak fat signal model showed superior performance to the single-peak fat signal model, which yielded an underestimation of liver fat. The R 2 * $$ {\mathrm{R}}_2^{\ast } $$ -PDFF relationships by simulations with fat susceptibility of 0.6 ppm and the multi-peak fat signal model were R 2 * = 0.490 × PDFF + 28.0 $$ {\mathrm{R}}_2^{\ast }=0.490\times \mathrm{PDFF}+28.0 $$ ( R 2 = 0.967 $$ {R}^2=0.967 $$ , p < 0.01 $$ p<0.01 $$ ) at 1.5 T and R 2 * = 0.928 × PDFF + 39.4 $$ {\mathrm{R}}_2^{\ast }=0.928\times \mathrm{PDFF}+39.4 $$ ( R 2 = 0.972 $$ {R}^2=0.972 $$ , p < 0.01 $$ p<0.01 $$ ) at 3.0 T. Monte Carlo simulations provide a new means for R 2 * $$ {\mathrm{R}}_2^{\ast } $$ -PDFF prediction, which is primarily determined by fat susceptibility, fat signal model, and magnetic field strength. Accurate R 2 * $$ {\mathrm{R}}_2^{\ast } $$ -PDFF calibration has the potential to correct the effect of fat on R 2 * $$ {\mathrm{R}}_2^{\ast } $$ quantification, and may be helpful for accurate R 2 * $$ {\mathrm{R}}_2^{\ast } $$ measurements in liver iron overload.

To develop Monte Carlo simulations to predict the relationship of R-2(* ) with liver fat content at 1.5 T and 3.0 T. For various fat fractions (FFs) from 1% to 25%, four types of virtual liver models were developed by incorporating the size and spatial distribution of fat droplets. Magnetic fields were then generated under different fat susceptibilities at 1.5 T and 3.0 T, and proton movement was simulated for phase accrual and MRI signal synthesis. The synthesized signal was fit to single-peak and multi-peak fat signal models for R-2(* ) and proton density fat fraction (PDFF) predictions. In addition, the relationships between R-2(* ) and PDFF predictions were compared with in vivo calibrations and Bland-Altman analysis was performed to quantitatively evaluate the effects of these components (type of virtual liver model, fat susceptibility, and fat signal model) on R-2(* ) predictions. A virtual liver model with realistic morphology of fat droplets was demonstrated, and R-2(* ) and PDFF values were predicted by Monte Carlo simulations at 1.5 T and 3.0 T. R-2(* ) predictions were linearly correlated with PDFF, while the slope was unaffected by the type of virtual liver model and increased as fat susceptibility increased. Compared with in vivo calibrations, the multi-peak fat signal model showed superior performance to the single-peak fat signal model, which yielded an underestimation of liver fat. The R-2(* ) -PDFF relationships by simulations with fat susceptibility of 0.6 ppm and the multi-peak fat signal model were R-2(* ) =0.490xPDFF+28.0 (R-2=0.967, p<0.01) at 1.5 T and R-2(* ) =0.928xPDFF+39.4 (R-2=0.972, p<0.01) at 3.0 T. Monte Carlo simulations provide a new means for R2*-PDFF prediction, which is primarily determined by fat susceptibility, fat signal model, and magnetic field strength. Accurate R-2(* ) -PDFF calibration has the potential to correct the effect of fat on R-2(* ) quantification, and may be helpful for accurate R-2(* ) measurements in liver iron overload. In this study, a Monte Carlo simulation of hepatic steatosis was developed to predict the relationship between R-2(* ) and PDFF. Furthermore, the effects of fat droplet morphology, fat susceptibility, fat signal model, and magnetic field strength were evaluated for the R-2(* ) -PDFF calibration. Our results suggest that Monte Carlo simulations provide a new means for R-2(* ) -PDFF prediction and this means can be easily generated for various regimes, such as simulations with higher fields and different echo times, as well as correction of magnetic susceptibility measurements for liver iron quantification. Abbreviations

Accurately detecting spine vertebrae plays a crucial role in successful orthopedic surgery. However, identifying and classifying lumbar vertebrae from arbitrary spine X-ray images remains challenging due to their similar appearance and varying sizes among individuals. In this paper, we propose a novel approach to enhance vertebrae detection accuracy by leveraging both global and local spatial relationships between neighboring vertebrae. Our method incorporates a two-stage detector architecture that captures global contextual information using an intermediate heatmap from the first stage. Additionally, we introduce a detection head in the second stage to capture local spatial information, enabling each vertebra to learn neighboring spatial details, visibility, and relative offset. During inference, we employ a fusion strategy that combines spatial offsets of neighboring vertebrae and heatmap from a conventional detection head. This enables the model to better understand relationships and dependencies between neighboring vertebrae. Furthermore, we introduce a new representation of object centers that emphasizes critical regions and strengthens the spatial priors of human spine vertebrae, resulting in an improved detection accuracy. We evaluate our method using two lumbar spine image datasets and achieve promising detection performance. Compared to the baseline, our algorithm achieves a significant improvement of 13.6% AP in the CM dataset and surpasses 6.5% and 4.8% AP in the anterior and lateral views of the BUU dataset, respectively.

PURPOSE:To investigate the effect of particle size on liver R 2 * $$ {\mathrm{R}}_2^{\ast } $$ by Monte Carlo simulation and phantom studies at both 1.5 T and 3.0 T. METHODS:Two kinds of particles (i.e., iron sphere and fat droplet) with varying sizes were considered separately in simulation and phantom studies. MRI signals were synthesized and analyzed for predicting R 2 * $$ {\mathrm{R}}_2^{\ast } $$ , based on simulations by incorporating virtual liver model, particle distribution, magnetic field generation, and proton movement into phase accrual. In the phantom study, iron-water and fat-water phantoms were constructed, and each phantom contained 15 separate vials with combinations of five particle concentrations and three particle sizes. R 2 * $$ {\mathrm{R}}_2^{\ast } $$ measurements in the phantom were made at both 1.5 T and 3.0 T. Finally, differences in R 2 * $$ {\mathrm{R}}_2^{\ast } $$ predictions or measurements were evaluated across varying particle sizes. RESULTS:In the simulation study, strong linear and positively correlated relationships were observed between R 2 * $$ {\mathrm{R}}_2^{\ast } $$ predictions and particle concentrations across varying particle sizes and magnetic field strengths ( r ≥ 0.988 $$ r\ge 0.988 $$ ). The relationships were affected by iron sphere size ( p < 0.001 $$ p<0.001 $$ ), where smaller iron sphere size yielded higher predicted R 2 * $$ {\mathrm{R}}_2^{\ast } $$ , whereas fat droplet size had no effect on R 2 * $$ {\mathrm{R}}_2^{\ast } $$ predictions ( p ≥ 0.617 $$ p\ge 0.617 $$ ) for constant total fat concentration. Similarly, the phantom study showed that R 2 * $$ {\mathrm{R}}_2^{\ast } $$ measurements were relatively sensitive to iron sphere size ( p ≤ 0.004 $$ p\le 0.004 $$ ) unlike fat droplet size ( p ≥ 0.223 $$ p\ge 0.223 $$ ). CONCLUSION:Liver R 2 * $$ {\mathrm{R}}_2^{\ast } $$ is affected by iron sphere size, but is relatively unaffected by fat droplet size. These findings may lead to an improved understanding of the underlying mechanisms of R 2 * $$ {\mathrm{R}}_2^{\ast } $$ relaxometry in vivo, and enable improved quantitative MRI phantom design.

PURPOSE:Through three neurocritical care unit (NCCU) surveys in China, we tried to understand the development status of neurocritical care and clarify its future development. METHODS:Using a cross-sectional survey method and self-report questionnaires, the number and quality of NCCUs were investigated through three steps: administering the questionnaire, sorting the survey data, and analyzing the survey data. RESULTS:At the second and third surveys, the number of NCCUs (76/112/206) increased by 47% and 84%, respectively. The NCCUs were located in tertiary grade A hospitals or teaching hospitals (65/100/181) in most provinces (24/28/29). The numbers of full-time doctors (359/668/1337) and full-time nurses (904/1623/207) in the NCCUs increased, but the doctor-bed ratio and nurse-bed ratio were still insufficient (0.4:1 and 1.3:1). CONCLUSION:In the past 20 years, the growth rate of NCCUs in China has accelerated, while the allocation of medical staff has been insufficient. Although most NCCU hospital bed facilities and instruments and equipment tend to be adequate, there are obvious defects in some aspects of NCCUs.

Evaluation of vitiligo relies on accurate segmentation of lesions, and traditional segmentation methods mainly focus on near-field images. This study proposes a deep learning-based model for accurately segmenting lesions in wide-field vitiligo images. In this study, a dataset of 1267 wide-field vitiligo images was established to train and evaluate segmentation models. A Swin R-CNN model, which combined a Swin Transformer tiny network with a watershed algorithm, was proposed for segmenting lesions. The performances of the Swin R-CNN model and five other models were evaluated and compared through visual and quantitative perspectives. Additionally, the Spearman rank correlation test was performed to analyze result consistency between the Swin R-CNN model and dermatologists in measuring lesion area. The Swin R-CNN model accurately segmented lesions in the wide-field vitiligo images, surpassing other models in both visual and quantitative performance, with an average precision of 84.72% and an average recall of 77.81%. The correlation coefficients between the evaluation results of the Swin R-CNN model and three dermatologists were 0.88, 0.94, and 0.91, respectively. The Swin R-CNN model accurately segments lesions in the wide-field vitiligo images and quantifies lesion area at the dermatologist level. The Swin R-CNN model can provide reliable analytical results for the vitiligo evaluation.

There is a lack of objective, accurate, and convenient methods for classification diagnostic hypopigmented dermatoses (HD) and severity evaluation of vitiligo. To achieve an accurate and intelligent classification diagnostic model of HD and severity evaluation model of vitiligo using a deep learning-based method. A total of 11,483 images from 4744 patients with HD were included in this study. An optimal diagnostic model was constructed by merging the squeeze-and-excitation (SE) module with the candidate model, its diagnostic efficiency was compared with that of 98 dermatologists. An objective severity evaluation indicator was proposed through weighting method and combined with a segmentation model to form a severity evaluation model, which was then compared with the assessments conducted by three experienced dermatologists using the naked eye. The improved diagnosis model SE_ResNet-18 outperformed the other 11 classic models with an accuracy of 0.9389, macro-specificity of 0.9878, and macro-f1 score of 0.9395, and outperformed the different categories of 98 dermatologists (P < 0.001). The weighted Kappa test indicated medium consistency between the Indicatorv and the VASIchange (K = 0.567, P < 0.05). The optimal segmented model, HR-Net, had 0.8421 mIOU. The model-based severity evaluation results were not significantly different among the three experienced dermatologists. This study proposes an objective, accurate, and convenient hybrid model for diagnosing HD and evaluating the severity of vitiligo, providing a method for dermatologists especially in grassroots hospitals, and provides a foundation for telemedicine.

由于不健康的饮食和缺乏运动,脂肪含量异常已逐渐成为危害人体健康的重要因素之一.在组织器官周围过度聚集的脂肪会破坏人体内信息传导系统,使得组织器官向身体传递错误信号,导致内分泌系统紊乱,这通常与脂肪肝、2型糖尿病、高血压和骨质疏松等慢性疾病密切相关.因此,脂肪精准量化对上述疾病的预防、病情评估和诊断治疗有着极其重要的临床意义.活检是目前脂肪量化的金标准,但其具有一定的有创性和抽样误差.近年来,磁共振成像技术越来越广泛地应用于脂肪量化研究.本文就多种磁共振成像技术,包括磁共振波谱、脂肪抑制成像、水脂分离和质子密度脂肪分数,量化人体不同部位(肝脏、胰腺、椎体骨髓和肌肉)脂肪含量的研究进展作一综述,旨在为脂肪量化提供更准确的生物标志物,以达到临床精准诊治的目的.

Monte Carlo modeling enables characterization of MR signals in various tissues, and has been applied to liver MR in the presence of fat. However, Monte Carlo modeling requires accurate information about the underlying tissue properties. In this work, we investigate the size and clustering of fat droplets in the liver using stereology and spatial statistics for three human liver biopsy samples with steatosis. Results show that the generalized gamma distribution function can accurately determine the size and location distributions of fat droplets. This may enable analysis of the underlying biophysical mechanisms between fat fraction and R2* from microscopic magnetic sensitivity.

ObjectiveTo understand the varieties, evaluation, treatment, and prognosis of severe neurological diseases using the third NCU survey in China.DesignA cross-sectional questionnaire study. The study was completed in three main steps: filling in the questionnaire, sorting out the survey data, and analyzing the survey data.ResultsOf 206 NCUs, 165 (80%) provided relatively complete information. It was estimated that 96,201 patients with severe neurological diseases were diagnosed and treated throughout the year, with an average fatality rate of 4.1%. The most prevalent severe neurological disease was cerebrovascular disease (55.2%). The most prevalent comorbidity was hypertension (56.7%). The most prevalent complication was hypoproteinemia (24.2%). The most common nosocomial infection was hospital-acquired pneumonia (10.6%). The GCS, APACHE II, EEG, and TCD were the most commonly used (62.4–95.2%). The implementation rate of the five nursing evaluation techniques reached 55.8–90.9%. Routinely raising the head of the bed by 30°, endotracheal intubation and central venous catheterization were the mostprevalent treatment strategies (97.6, 94.5, and 90.3%, respectively). Traditional tracheotomy, invasive mechanical ventilation and nasogastric tube feeding (75.8, 95.8, and 95.8%, respectively) were more common than percutaneous tracheotomy, non-invasive mechanical ventilation and nasogastric tube insertion (57.6, 57.6, and 66.7%, respectively). Body surface hypothermia brain protection technology was more commonly used than intravascular hypothermia technology (67.3 > 6.1%). The rates of minimally invasive hematoma removal and ventricular puncture were only 40.0 and 45.5%, respectively.ConclusionIn addition to traditional recognized basic life assessment and support technology, it is necessary to the use of promote specialized technology for neurological diseases, according to the characteristics of critical neurological diseases.

BACKGROUND AND OBJECTIVE:To model hepatic steatosis in adult humans with non-alcoholic fatty liver disease based on stereology and spatial distribution of fat droplets from liver biopsy specimens. METHODS:Histological analysis was performed on 30 adult human liver biopsy specimens with varying degrees of steatosis. Morphological features of fat droplets were characterized by gamma distribution function (GDF) in both two-dimensional (2D) and three-dimensional (3D) spaces from three aspects: 1) size distribution indicating non-uniformity of fat droplets in radius; 2) nearest neighbor distance distribution indicating heterogeneous accumulation (i.e., clustering) of fat droplets; 3) regional anisotropy indicating inter-regional variability in fat fraction (FF). To generalize the morphological description of hepatic steatosis to different FFs, correlation analysis was performed among the estimated GDF parameters and FFs for all specimens. Finally, Monte Carlo modeling of hepatic steatosis was developed to simulate fat droplet distribution in tissue. RESULTS:Morphological features, including size and nearest neighbor distance in 2D and 3D spaces as well as regional anisotropy, statistically captured the distribution of fat droplets by the GDF fit (R2 > 0.54). The estimated GDF parameters (i.e., scale and shape parameters) and FFs were well correlated, with R2 > 0.55. In addition, simulated 3D liver morphological models demonstrated similar sections to real histological samples both visually and quantitatively. CONCLUSIONS:The morphology of hepatic steatosis is well characterized by stereology and spatial distribution of fat droplets. Simulated models demonstrate similar appearances to real histological samples. Furthermore, the model may help understand MRI signal behavior in the presence of liver steatosis.

The accurate and fast segmentation method of tumor regions in brain Magnetic Resonance Imaging (MRI) is significant for clinical diagnosis, treatment and monitoring, given the aggressive and high mortality rate of brain tumors. However, due to the limitation of computational complexity, convolutional neural networks (CNNs) face challenges in being efficiently deployed on resource-limited devices, which restricts their popularity in practical medical applications. To address this issue, we propose a lightweight and efficient 3D convolutional neural network SDS-Net for multimodal brain tumor MRI image segmentation. SDS-Net combines depthwise separable convolution and traditional convolution to construct the 3D lightweight backbone blocks, lightweight feature extraction (LFE) and lightweight feature fusion (LFF) modules, which effectively utilizes the rich local features in multimodal images and enhances the segmentation performance of sub-tumor regions. In addition, 3D shuffle attention (SA) and 3D self-ensemble (SE) modules are incorporated into the encoder and decoder of the network. The SA helps to capture high-quality spatial and channel features from the modalities, and the SE acquires more refined edge features by gathering information from each layer. The proposed SDS-Net was validated on the BRATS datasets. The Dice coefficients were achieved 92.7, 80.0 and 88.9% for whole tumor (WT), enhancing tumor (ET) and tumor core (TC), respectively, on the BRTAS 2020 dataset. On the BRTAS 2021 dataset, the Dice coefficients were 91.8, 82.5 and 86.8% for WT, ET and TC, respectively. Compared with other state-of-the-art methods, SDS-Net achieved superior segmentation performance with fewer parameters and less computational cost, under the condition of 2.52 M counts and 68.18 G FLOPs.

Executive function is vital for normal social, cognitive, and motor functions. Executive function decline due to aging increases the risk of disability and falls in older adults, which has become an urgent public health issue. Fast and convenient neuropsychological tools are thus needed to identify high-risk groups as early as possible to conduct a timely intervention. Card sorting tasks, such as Wisconsin Card Sorting Task (WCST) and its variants, are popular tools for measuring executive function. This study investigated the reliability of an open-source, self-administered, online, short-version card sorting task with a sample of young ( n = 107, 65 females, age: M = 30.1 years, SD = 5.5 years) and elderly Chinese ( n = 113, 53 females, age: M = 64.0 years, SD = 6.7 years). We developed an automated scoring and visualization procedure following the recent recommendations on scoring perseverative responses to make the results comparable to the standardized WCST. Reliability estimates of commonly used measures were calculated using the split-half method. All task indices' reliabilities were reasonably good in both old and young groups except for "failure-to-maintain-set." Elderly Chinese adults showed compromised task performance on all measures compared with the young Chinese adults at the group level. The R script of automated scoring and estimation of reliability is publicly available.

In the diagnosis of cardiovascular diseases, the analysis of electrocardiogram (ECG) signals has always played a crucial role. At present, how to effectively identify abnormal heart beats by algorithms is still a difficult task in the field of ECG signal analysis. Based on this, a classification model that automatically identifies abnormal heartbeats based on deep residual network (ResNet) and self-attention mechanism was proposed. Firstly, this paper designed an 18-layer convolutional neural network (CNN) based on the residual structure, which helped model fully extract the local features. Then, the bi-directional gated recurrent unit (BiGRU) was used to explore the temporal correlation for further obtaining the temporal features. Finally, the self-attention mechanism was built to weight important information and enhance model's ability to extract important features, which helped model achieve higher classification accuracy. In addition, in order to mitigate the interference on classification performance due to data imbalance, the study utilized multiple approaches for data augmentation. The experimental data in this study came from the arrhythmia database constructed by MIT and Beth Israel Hospital (MIT-BIH), and the final results showed that the proposed model achieved an overall accuracy of 98.33% on the original dataset and 99.12% on the optimized dataset, which demonstrated that the proposed model can achieve good performance in ECG signal classification, and possessed potential value for application to portable ECG detection devices.

MRI of effective transverse relaxation rate (R2*) measurement is a reliable method for liver iron concentration quantification. However, R2* mapping can be degraded by noise, especially in the case of iron overload. This study aimed to develop a deep learning method for MRI R2* relaxometry of an iron-loaded liver using a two-stage cascaded neural network. The proposed method, named CadamNet, combines two convolutional neural networks separately designed for image denoising and parameter mapping into a cascade framework, and the physics-based R2* decay model was incorporated in training the mapping network to enforce data consistency further. CadamNet was trained using simulated liver data with Rician noise, which was constructed from clinical liver data. The performance of CadamNet was quantitatively evaluated on simulated data with varying noise levels as well as clinical liver data and compared with the single-stage parameter mapping network (MappingNet) and two conventional model-based R2* mapping methods. CadamNet consistently achieved high-quality R2* maps and outperformed MappingNet at varying noise levels. Compared with conventional R2* mapping methods, CadamNet yielded R2* maps with lower errors, higher quality, and substantially increased efficiency. In conclusion, the proposed CadamNet enables accurate and efficient iron-loaded liver R2* mapping, especially in the presence of severe noise.

Background Disorders of consciousness (DOC) are a spectrum of pathologies affecting one’s ability to interact with the external world. At present, High-Definition Transcranial Direct Current Stimulation (HD-tDCS) is used in many patients with DOC as a non-invasive treatment, but electrophysiological research on the effect of HD-tDCS on patients with DOC is limited. Objectives To explore how HD-tDCS affects the cerebral cortex and examine the possible electrophysiological mechanisms underlying the effects of HD-tDCS on the cerebral cortex. Methods A total of 19 DOC patients were assigned to HD-tDCS stimulation. Each of them underwent 10 anodal HD-tDCS sessions of the left dorsolateral prefrontal cortex (DLPFC) over 5 consecutive days. Coma Recovery Scale-Revision (CRS-R) scores were recorded to evaluate the consciousness level before and after HD-tDCS, while resting-state electroencephalography (EEG) recordings were obtained immediately before and after single and multiple HD-tDCS stimuli. Depending on whether the CRS-R score increased after stimulation, we classified the subjects into responsive (RE) and non-responsive (N-RE) groups and compared the differences in power spectral density (PSD) between the groups in different frequency bands and brain regions, and also examined the relationship between PSD values and CRS-R scores. Results For the RE group, the PSD value of the parieto-occipital region increased significantly in the 6–8 Hz frequency band after multiple stimulations by HD-tDCS. After a single stimulation, an increase in PSD was observed at 10–13 and 13–30 Hz. In addition, for all subjects, a positive correlation was observed between the change in PSD value in the parieto-occipital region at 10–13 and 6–8 Hz frequency band and the change in CRS-R score after a single stimulation. Conclusion Repeated anodal HD-tDCS of the left DLPFC can improve clinical outcomes in patients with DOC, and HD-tDCS-related increased levels of consciousness were associated with increased parieto-occipital PSD.