MFSynDCP: multi-source feature collaborative interactive learning for drug combination synergy prediction.

Drug combination therapy is generally more effective than monotherapy in the field of cancer treatment. However, screening for effective synergistic combinations from a wide range of drug combinations is particularly important given the increase in the number of available drug classes and potential drug-drug interactions. Existing methods for predicting the synergistic effects of drug combinations primarily focus on extracting structural features of drug molecules and cell lines, but neglect the interaction mechanisms between cell lines and drug combinations. Consequently, there is a deficiency in comprehensive understanding of the synergistic effects of drug combinations. To address this issue, we propose a drug combination synergy prediction model based on multi-source feature interaction learning, named MFSynDCP, aiming to predict the synergistic effects of anti-tumor drug combinations. This model includes a graph aggregation module with an adaptive attention mechanism for learning drug interactions and a multi-source feature interaction learning controller for managing information transfer between different data sources, accommodating both drug and cell line features. Comparative studies with benchmark datasets demonstrate MFSynDCP's superiority over existing methods. Additionally, its adaptive attention mechanism graph aggregation module identifies drug chemical substructures crucial to the synergy mechanism. Overall, MFSynDCP is a robust tool for predicting synergistic drug combinations. The source code is available from GitHub at https://github.com/kkioplkg/MFSynDCP .

Investigating the influence of the gut microbiome on cholelithiasis: unveiling insights through sequencing and predictive modeling.

BACKGROUND: Cholelithiasis is one of the most common disorders of hepatobiliary system. Gut bacteria may be involved in the process of gallstone formation and are, therefore considered as potential targets for cholelithiasis prediction. OBJECTIVE: To reveal the correlation between cholelithiasis and gut bacteria. METHODS: Stool samples were collected from 100 cholelithiasis and 250 healthy individuals from Huzhou Central Hospital; The 16S rRNA of gut bacteria in the stool samples was sequenced using the third-generation Pacbio sequencing platform; Mothur v.1.21.1 was used to analyze the diversity of gut bacteria; Wilcoxon rank-sum test and linear discriminant analysis of effect sizes (LEfSe) were used to analyze differences in gut bacteria between patients suffering from cholelithiasis and healthy individuals; Chord diagram and Plot-related heat maps were used to analyze the correlation between cholelithiasis and gut bacteria; six machine algorithms were used to construct models to predict cholelithiasis. RESULTS: There were differences in the abundance of gut bacteria between cholelithiasis and healthy individuals, but there were no differences in their community diversity. Increased abundance of Costridia, Escherichia flexneri, and Klebsiella pneumonae were found in cholelithiasis, while Bacteroidia, Phocaeicola, and Phocaeicola vulgatus were more abundant in healthy individuals. The top four bacteria that were most closely associated with cholelithiasis were Escherichia flexneri, Escherichia dysenteriae, Streptococcus salivarius, and Phocaeicola vulgatus. The cholelithiasis model based on CatBoost algorithm had the best prediction effect (sensitivity: 90.48%, specificity: 88.32%, and AUC: 0.962). CONCLUSION: The identification of characteristic gut bacteria may provide new predictive targets for gallstone screening. As being screened by the predictive model, people at high risk of cholelithiasis can determine the need for further testing, thus enabling early warning of cholelithiasis.

COVID-19 Infection Percentage Estimation from Computed Tomography Scans: Results and Insights from the International Per-COVID-19 Challenge.

COVID-19 analysis from medical imaging is an important task that has been intensively studied in the last years due to the spread of the COVID-19 pandemic. In fact, medical imaging has often been used as a complementary or main tool to recognize the infected persons. On the other hand, medical imaging has the ability to provide more details about COVID-19 infection, including its severity and spread, which makes it possible to evaluate the infection and follow-up the patient's state. CT scans are the most informative tool for COVID-19 infection, where the evaluation of COVID-19 infection is usually performed through infection segmentation. However, segmentation is a tedious task that requires much effort and time from expert radiologists. To deal with this limitation, an efficient framework for estimating COVID-19 infection as a regression task is proposed. The goal of the Per-COVID-19 challenge is to test the efficiency of modern deep learning methods on COVID-19 infection percentage estimation (CIPE) from CT scans. Participants had to develop an efficient deep learning approach that can learn from noisy data. In addition, participants had to cope with many challenges, including those related to COVID-19 infection complexity and crossdataset scenarios. This paper provides an overview of the COVID-19 infection percentage estimation challenge (Per-COVID-19) held at MIA-COVID-2022. Details of the competition data, challenges, and evaluation metrics are presented. The best performing approaches and their results are described and discussed.

Cholesterol induction in CD8+ T cell exhaustion in colorectal cancer via the regulation of endoplasmic reticulum-mitochondria contact sites.

BACKGROUND: Hypercholesterolemia is one of the risk factors for colorectal cancer (CRC). Cholesterol can participate in the regulation of human T cell function and affect the occurrence and development of CRC. OBJECTIVE: To elucidate the pathogenesis of CRC immune escape mediated by CD8+ T cell exhaustion induced by cholesterol. METHODS: CRC samples (n = 217) and healthy individuals (n = 98) were recruited to analyze the relationship between peripheral blood cholesterol levels and the clinical features of CRC. An animal model of CRC with hypercholesterolemia was established. Intraperitoneal intervention with endoplasmic reticulum stress (ERS) inhibitors in hypercholesterolemic CRC mice was performed. CD69, PD1, TIM-3, and CTLA-4 on CD8+ T cells of spleens from C57BL/6 J mice were detected by flow cytometry. CD8+ T cells were cocultured with MC38 cells (mouse colon cancer cell line). The proliferation, apoptosis, migration and invasive ability of MC38 cells were detected by CCK-8 assay, Annexin-V APC/7-AAD double staining, scratch assay and transwell assay, respectively. Transmission electron microscopy was used to observe the ER structure of CD8+ T cells. Western blotting was used to detect the expression of ERS and mitophagy-related proteins. Mitochondrial function and energy metabolism were measured. Immunoprecipitation was used to detect the interaction of endoplasmic reticulum-mitochondria contact site (ERMC) proteins. Immunofluorescence colocalization was used to detect the expression and intracellular localization of ERMC-related molecules. RESULTS: Peripheral blood cholesterol-related indices, including Tc, low density lipoproteins (LDL) and Apo(a), were all increased, and high density lipoprotein (HDL) was decreased in CRCs. The proliferation, migration and invasion abilities of MC38 cells were enhanced, and the proportion of tumor cell apoptosis was decreased in the high cholesterol group. The expression of IL-2 and TNF-α was decreased, while IFN-γ was increased in the high cholesterol group. It indicated high cholesterol could induce exhaustion of CD8+ T cells, leading to CRC immune escape. Hypercholesterolemia damaged the ER structure of CD8+ T cells and increased the expression of ER stress molecules (CHOP and GRP78), lead to CD8+ T cell exhaustion. The expression of mitophagy-related proteins (BNIP3, PINK and Parkin) in exhausted CD8+ T cells increased at high cholesterol levels, causing mitochondrial energy disturbance. High cholesterol enhanced the colocalization of Fis1/Bap31, MFN2/cox4/HSP90B1, VAPB/PTPIP51, VDAC1/IPR3/GRP75 in ERMCs, indicated that high cholesterol promoted the intermolecular interaction between ER and mitochondrial membranes in CD8+ T cells. CONCLUSION: High cholesterol regulated the ERS-ERMC-mitophagy axis to induce the exhaustion of CD8+ T cells in CRC.

Graph Neural Network for representation learning of lung cancer.

The emergence of image-based systems to improve diagnostic pathology precision, involving the intent to label sets or bags of instances, greatly hinges on Multiple Instance Learning for Whole Slide Images(WSIs). Contemporary works have shown excellent performance for a neural network in MIL settings. Here, we examine a graph-based model to facilitate end-to-end learning and sample suitable patches using a tile-based approach. We propose MIL-GNN to employ a graph-based Variational Auto-encoder with a Gaussian mixture model to discover relations between sample patches for the purposes to aggregate patch details into an individual vector representation. Using the classical MIL dataset MUSK and distinguishing two lung cancer sub-types, lung cancer called adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC), we exhibit the efficacy of our technique. We achieved a 97.42% accuracy on the MUSK dataset and a 94.3% AUC on the classification of lung cancer sub-types utilizing features.

The effect of surface gloss on categorical color constancy in real scenes.

Categorical color constancy has been widely investigated and found to be very robust. As one of object material properties, the surface gloss was found to barely contribute to color constancy in a natural viewing condition. In this study, the effect of surface gloss on categorical color constancy was investigated by asking eight observers to categorize 208 Munsell matte surfaces and 260 Munsell glossy surfaces under D65, F, and TL84 illuminants in a viewing chamber with a uniform gray background. A color constancy index based on the centroid shift of the color category was used to evaluate color constancy degree of each color category across illumination changes from D65 to F or TL84 illuminant. The result showed that both matte and glossy surfaces showed almost perfect color constancy on all color categories under F and TL84 illuminants, and there was no significant difference between them. This result suggests that surface gloss has little effect on categorical color constancy in a uniform gray background where the local surround cue was present, which is consistent with the previous findings.

Robust categorical color constancy along daylight locus in red-green color deficiency.

Categorical color constancy in normal trichromats has been found to be very robust in real scenes. In this study, we investigated categorical color constancy in red-green dichromats and anomalous trichromats. Eight dichromats (two protanopes and six deuteranopes), eight anomalous trichromats (four protanomalous and four deuteranomalous trichromats), and eight normal trichromats sorted 208 Munsell matte surfaces into Berlin and Kay's basic color categories under D65 illuminant, F illuminant with correlated color temperature 4200 K, and TL84 illuminant with correlated color temperature 2700 K. Color constancy was quantified by a color constancy index. The results showed that the constancy index of dichromats (0.79) was considerable and significantly lower than that of normal trichromats (0.87) while that of anomalous trichromats (0.84) was not. The impairment of color constancy performance in dichromats was expected to be caused by their large intra-subject variabilities in color naming. The results indicate robust categorical color constancy along daylight locus in red-green dichromats and anomalous trichromats, which might be contributed by cone adaptation mechanism and be independent of color discrimination mechanism. It suggests that the color categorization by color vision deficient subjects can be reasonable without any assistants of artificial equipment in daily life under sunlight and common illuminations.

The estimated effect of graphic warning labels on smoker's intention to quit in Shanghai, China: a cross-sectional study.

BACKGROUND: Tobacco consumption is the leading cause of death worldwide. Overwhelming studies demonstrate graphic warning labels (GWLs) on cigarette packs are effective in eliciting negative response to tobacco smoking, modifying beliefs about tobacco dangers, and increasing reported intention to quit, but the estimated effect of GWLs on smoking cessation intention among smokers is still limited in China. In this study, we aim to understand the smoking intensity, smoking duration and smoking cessation intention among current smokers, and to explore how their smoking cessation intention would be influenced by the GWLs in Shanghai. METHODS: From January to June 2021, we totally recruited 1104 current smokers in Songjiang district and Fengxian district of Shanghai by multistage sampling design. We used Android pad assisted electronic questionnaire for data collection, and then implemented logistic regression for odds ratio (OR) and 95% confidence interval (CI) calculation to explore how smoking cessation intention would be influenced by the GWLs among current smokers. RESULTS: One thousand one hundred four current smokers included 914 males (82.79%), with an average age of 43.61 years. 58.06% of current smokers reported smoking cessation intention due to GWLs. Logistic regression indicated a higher percentage of smoking cessation intention due to GWLs was among female smokers [OR = 2.41, 95% CI (1.61-3.59)], smokers with smoking intensity < 20 cigarette/day [OR = 1.92, 95% CI (1.44-2.55)], smokers with tobacco burden < 20% [OR = 1.94, 95% CI (1.35-2.79)], and among smokers had plan to quit in a year [OR = 6.58, 95% CI (4.71-9.18). Smokers with higher individual monthly income had lower percentage of smoking cessation intention (OR were 0.35, 0.46 and 0.41). Meanwhile, among 642 current smokers without plan to quit in a year, approximately 40% of them reported smoking cessation intention due to GWLs. CONCLUSIONS: Smoking cessation intention due to the assumed GWLs on cigarette packs is high among current smokers in Shanghai, especially in female smokers, smokers with light tobacco burden and mild nicotine dependence. Incorporating smoking intensity as well as smoking burden into the implementation of GWLs as tobacco control measures would discourage smoking in China.

MLW-gcForest: a multi-weighted gcForest model towards the staging of lung adenocarcinoma based on multi-modal genetic data.

BACKGROUND: Lung cancer is one of the most common types of cancer, among which lung adenocarcinoma accounts for the largest proportion. Currently, accurate staging is a prerequisite for effective diagnosis and treatment of lung adenocarcinoma. Previous research has used mainly single-modal data, such as gene expression data, for classification and prediction. Integrating multi-modal genetic data (gene expression RNA-seq, methylation data and copy number variation) from the same patient provides the possibility of using multi-modal genetic data for cancer prediction. A new machine learning method called gcForest has recently been proposed. This method has been proven to be suitable for classification in some fields. However, the model may face challenges when applied to small samples and high-dimensional genetic data. RESULTS: In this paper, we propose a multi-weighted gcForest algorithm (MLW-gcForest) to construct a lung adenocarcinoma staging model using multi-modal genetic data. The new algorithm is based on the standard gcForest algorithm. First, different weights are assigned to different random forests according to the classification performance of these forests in the standard gcForest model. Second, because the feature vectors generated under different scanning granularities have a diverse influence on the final classification result, the feature vectors are given weights according to the proposed sorting optimization algorithm. Then, we train three MLW-gcForest models based on three single-modal datasets (gene expression RNA-seq, methylation data, and copy number variation) and then perform decision fusion to stage lung adenocarcinoma. Experimental results suggest that the MLW-gcForest model is superior to the standard gcForest model in constructing a staging model of lung adenocarcinoma and is better than the traditional classification methods. The accuracy, precision, recall, and AUC reached 0.908, 0.896, 0.882, and 0.96, respectively. CONCLUSIONS: The MLW-gcForest model has great potential in lung adenocarcinoma staging, which is helpful for the diagnosis and personalized treatment of lung adenocarcinoma. The results suggest that the MLW-gcForest algorithm is effective on multi-modal genetic data, which consist of small samples and are high dimensional.

Heat Transfer and Flow Structures of Laminar Confined Slot Impingement Jet with Power-Law Non-Newtonian Fluid.

Heat transfer performances and flow structures of laminar impinging slot jets with power-law non-Newtonian fluids and corresponding typical industrial fluids (Carboxyl Methyl Cellulose (CMC) solutions and Xanthangum (XG) solutions) have been studied in this work. Investigations are performed for Reynolds number Re less than 200, power-law index n ranging from 0.5 to 1.5 and consistency index K varying from 0.001 to 0.5 to explore heat transfer and flow structure of shear-thinning fluid and shear-thickening fluid. Results indicate that with the increase of n, K for a given Re, wall Nusselt number increases mainly attributing to the increase of inlet velocity U. For a given inlet velocity, wall Nusselt number decreases with the increase of n and K, which mainly attributes to the increase of apparent viscosity and the reduction of momentum diffusion. For the same Re, U and Pr, wall Nusselt number decreases with the increase of n. Among the study of industrial power-law shear-thinning fluid, CMC solution with 100 ppm shows the best heat transfer performance at a given velocity. Moreover, new correlation of Nusselt number about industrial fluid is proposed. In general, for the heat transfer of laminar confined impinging jet, it is best to use the working fluid with low viscosity.

Individualized perfusion of sargent gloryvine decoction for clinical severe acute pancreatitis treatment.

BACKGROUND/AIMS: To observe the effects of Sargent gloryvine decoction (SGD) on severe acute pancreatitis (SAP) treatment and to evaluate its clinical value. METHODOLOGY: 112 patients of SAP in our hospital from January, 2005 to December, 2012 were recruited for retrospective analysis. They were divided into two groups, SGD group (62 patients) and control group without treated with SGD (50 patients). Inflammation factor, CT grade and Ranson grade were used to estimate the severity of SAP, and were compared in these two groups. In addition, peripancreatic infection, incidence of pseudo pancreatic cyst, time of anal exsufflation and duration of fever were used to evaluate the effect of SGD treatment. After perfusion of SGD for different time, hospitalization days and cost were recorded to evaluate clinical value of SGD. RESULTS: After perfusion, many indexes in SGD were remarkably superior to those of control group, such as duration of fever, incidence of pseudo pancreatic cyst, peripancreatic infection and Ranson grade. Meanwhile, SGD can sharply down-regulate inflammation reaction levels of SAP patients, so that the hospitalization days and costs can be obviously saved. CONCLUSION: According to comparison, perfusion of SGD is a potential candidate for SAP treatment and is valuable in clinical application.

[Association between occupational psychological stress and metabolic syndrome].

OBJECTIVE: To investigate the association between occupational psychological stress and metabolic syndrome (MS) in Hui and Han populations in Ningxia, China. METHODS: A 1:1 matched case-control study was performed. A total of 600 unrelated patients aged from 20 to 60 years who were clearly diagnosed with MS in General Hospital of Ningxia Medical University and Wuzhong People's Hospital from October 2011 to October 2012 were collected as the case group (MS group). A total of 600 healthy people who underwent a regular health examination in the same hospital during the same period were selected as the control group with matched gender, nationality, and age (≤ ± 3 years). The self-designed questionnaire was used to investigate the general situations and do the physical examination, and the fasting venous blood samples were collected for laboratory biochemical blood tests. The Occupational Stress Inventory (OSI) was used to investigate the subjects' occupational stress factors and stress levels. RESULTS: With the increase in stress levels, the levels of WC, FPG, TG, AST, and UA were increased, WHR, SBP, and DBP first increased and then decreased, and the level of HDL-C increased. There were statistically significant differences in these parameters between the two groups (P < 0.05 or 0.01). The occupational psychological stress test results showed that the total score of stress factors (t = 6.676, P < 0.05), workload (t = 10.269, P < 0.05), interpersonal relationship (t = 6.569, P < 0.05), family/work balance (t = 2.028, P < 0.05), cognitive load (t = 8.714, P < 0.05), and other scores (t = 2.838, P < 0.05) in the MS group were all significantly higher than those in the control group, but there were no significant differences in the scores of management role, work responsibilities, and organizational climate between the MS group and the control group (P>0.05). There were no significant differences in the total score of stress factors and the score of each factor between Hui and Han groups (P>0.05). The relative risks of MS in the people with moderate stress exposure were 2.325 and 2.331 times those in the people with mild stress exposure before and after adjustment for age, gender, education level, marriage status, smoking, and drinking, and the relative risks for MS in the people with severe stress exposure were 3.000 and 3.126 times those in the people with mild stress exposure. There were significant differences in the detection rates of abdominal obesity, high TG, low HDL-C, hypertension, hyperglycemia, and diabetes between the sub-groups with different stress levels in the MS group (χ² = 17.636, 8.514, 14.640, 14.280, and 33.323, P < 0.01). The results of multivariate conditional logistic regression analysis showed that the risk factors for MS were SBP, TG, LDL-C, UA, BMI, fasting blood glucose, family history of hypertension, family history of diabetes, and the level of psychological stress in Ningxia, and the protective factor for MS was HDL-C. CONCLUSION: The occupational psychological stress is closely associated with MS, and it is an environmental risk factor for MS. With the increase in the stress level, the detection rates of MS components and the relative risk for MS are significantly increased. And there is no significant difference in the level of occupational psychological stress between the Hui and Han nationality groups.

Management of ovarian cysts in infants.

BACKGROUND: To discuss the experience of diagnosis and treatment of ovarian cyst in infants. MATERIALS AND METHODS: A retrospective review was conducted on 20 infants who suffered from ovarian cyst. RESULTS: There were no dysplasia ovarian was found in children which were preoperatively diagnosed simplex cyst. Within thirteen children preoperatively detected mixed cystic-solid lesion, six cases ovarian cysts disappeared and two cases underwent poor blood supply in the following time. CONCLUSION: Adverse effects for ovarian cyst in infants can be prevented by agressive surgical intervention. Harmful effects of ovarian cyst can be prevented by positive surgical intervention despite the diagnostic difficulties in children with clinical symptoms of this condition.

Barriers and Stimulus in Shared Decision Making Among Aesthetic Dermatologists in China: Findings from a Cross-Sectional Study.

Introduction: Shared decision making (SDM) is a collaborative process involving both healthcare providers and patients in making medical decisions, which gains increasing prominence in healthcare practice. But evidence on the level of SDM in medical practice and barriers as well as stimulus during the SDM implementation among aesthetic dermatologists is limited in China. Methods: From July to August 2023, 1938 dermatologists were recruited online in China. Data were collected through an electronic questionnaire covering: (1) demographic features; (2) SDM questionnaire physician version (SDM-Q-Doc); and (3) stimulus and barriers in SDM implementation. Logistic regression was applied to explore factors associated with SDM practice, barriers, and stimulus of SDM implementation, respectively. Results: The 1938 dermatologists included 1329 females (68.6%), with an average age of 35 years. The total SDM score ranged from 0 to 45, with a median value of 40 (IQR: 35-44), and the median stimulus score and barriers scores were 28 (IQR: 24-32) and 19 (IQR: 13-26), respectively. The prevalence of good SDM was 27.2%, logistic regression indicated that female dermatologists (odds ratio, OR=1.21, 95% confidence interval, CI: 0.96-1.51), and dermatologists with more years of aesthetic practice had a higher proportion of good SDM practice (OR was 1.44 for 5-9 years, 1.58 for 10-15 years and 1.77 for over 15 years). Moreover, female dermatologists and dermatologists with higher education level and serviced in private settings had lower barrier scores; female dermatologists and dermatologists with more years of aesthetic practice had higher stimulus scores. Conclusion: Chinese aesthetic dermatologists appear to implement SDM at an active level, with more stimulus and less barriers in SDM implementation. The integration of SDM into clinical practice among dermatologists is beneficial both for patients and dermatologists. Moreover, SDM practice should be strongly promoted and enhanced during medical aesthetics, especially among male dermatologists, dermatologists with less working experience, and those who work at public institutions.

Shared decision-making implementation status among dermatologists engaging in medical esthetics: a cross-sectional study in China.

Objective: Shared decision-making (SDM) is a collaborative process in which patients and healthcare providers jointly make a medical decision. This cross-sectional study aimed to identify the implementation status of shared decision-making among dermatologists engaging in medical esthetics in China and to identify factors associated with the good practice of SDM among them. Methods: From January to June 2023, a total of 1,287 dermatologists engaging in medical esthetics in China were recruited and completed the online interviews about their implementation of SDM based on the Shared Decision-Making Questionnaire for Doctors (SDM-Q-Doc). Logistic regression was used to calculate the odds ratio (OR) and 95% confidence interval (CI) to explore factors associated with the higher SDM score achievement among dermatologists with medical esthetic practice. Results: The median value of the total SDM score was 39, and 48% (621/1278) of dermatologists with medical esthetic practice achieved at least 40 out of 45 scores. Logistic regression indicated that dermatologists aged 40-49 or ≥ 50 years and those engaging in medical esthetic practice for ≥5 years were more likely to achieve at least 40 out of 45 scores compared to dermatologists aged <30 years with less than 5 years of medical esthetic practice. The ORs were 1.82 (95% CI: 1.13-3.12), 1.94 (95% CI: 1.13-3.61), and 1.76 (95% CI: 1.34-2.31), respectively. Conclusion: The SDM implementation level among Chinese dermatologists engaging in medical esthetics is high, especially for those who are older age and have more years of practice. Hence, it is highly recommended to promote and enhance SDM practice among younger dermatologists engaging in medical esthetics with less working experience.

CHNet: A multi-task global-local Collaborative Hybrid Network for KRAS mutation status prediction in colorectal cancer.

Accurate prediction of Kirsten rat sarcoma (KRAS) mutation status is crucial for personalized treatment of advanced colorectal cancer patients. However, despite the excellent performance of deep learning models in certain aspects, they often overlook the synergistic promotion among multiple tasks and the consideration of both global and local information, which can significantly reduce prediction accuracy. To address these issues, this paper proposes an innovative method called the Multi-task Global-Local Collaborative Hybrid Network (CHNet) aimed at more accurately predicting patients' KRAS mutation status. CHNet consists of two branches that can extract global and local features from segmentation and classification tasks, respectively, and exchange complementary information to collaborate in executing these tasks. Within the two branches, we have designed a Channel-wise Hybrid Transformer (CHT) and a Spatial-wise Hybrid Transformer (SHT). These transformers integrate the advantages of both Transformer and CNN, employing cascaded hybrid attention and convolution to capture global and local information from the two tasks. Additionally, we have created an Adaptive Collaborative Attention (ACA) module to facilitate the collaborative fusion of segmentation and classification features through guidance. Furthermore, we introduce a novel Class Activation Map (CAM) loss to encourage CHNet to learn complementary information between the two tasks. We evaluate CHNet on the T2-weighted MRI dataset, and achieve an accuracy of 88.93% in KRAS mutation status prediction, which outperforms the performance of representative KRAS mutation status prediction methods. The results suggest that our CHNet can more accurately predict KRAS mutation status in patients via a multi-task collaborative facilitation and considering global-local information way, which can assist doctors in formulating more personalized treatment strategies for patients.

Two-stream vision transformer based multi-label recognition for TCM prescriptions construction.

Traditional Chinese medicine (TCM) observation diagnosis images (including facial and tongue images) provide essential human body information, holding significant importance in clinical medicine for diagnosis and treatment. TCM prescriptions, known for their simplicity, non-invasiveness, and low side effects, have been widely applied worldwide. Exploring automated herbal prescription construction based on visual diagnosis holds vital value in delving into the correlation between external features and herbal prescriptions and offering medical services in mobile healthcare systems. To effectively integrate multi-perspective visual diagnosis images and automate prescription construction, this study proposes a multi-herb recommendation framework based on Visual Transformer and multi-label classification. The framework comprises three key components: image encoder, label embedding module, and cross-modal fusion classification module. The image encoder employs a dual-stream Visual Transformer to learn dependencies between different regions of input images, capturing both local and global features. The label embedding module utilizes Graph Convolutional Networks to capture associations between diverse herbal labels. Finally, two Multi-Modal Factorized Bilinear modules are introduced as effective components to fuse cross-modal vectors, creating an end-to-end multi-label image-herb recommendation model. Through experimentation with real facial and tongue images and generating prescription data closely resembling real samples. The precision is 50.06 %, the recall rate is 48.33 %, and the F1-score is 49.18 %. This study validates the feasibility of automated herbal prescription construction from the perspective of visual diagnosis. Simultaneously, it provides valuable insights for constructing herbal prescriptions automatically from more physical information.

Ultra-short-term forecasting model of power load based on fusion of power spectral density and Morlet wavelet.

An accurate ultra-short-term time series prediction of a power load is an important guarantee for power dispatching and the safe operation of power systems. Problems of the current ultra-short-term time series prediction algorithms include low prediction accuracy, difficulty capturing the local mutation features, poor stability, and others. From the perspective of series decomposition, a multi-scale sequence decomposition model (TFDNet) based on power spectral density and the Morlet wavelet transform is proposed that combines the multidimensional correlation feature fusion strategy in the time and frequency domains. By introducing the time-frequency energy selection module, the "prior knowledge" guidance module, and the sequence denoising decomposition module, the model not only effectively delineates the global trend and local seasonal features, completes the in-depth information mining of the smooth trend and fluctuating seasonal features, but more importantly, realizes the accurate capture of the local mutation seasonal features. Finally, on the premise of improving the forecasting accuracy, single-point load forecasting and quantile probabilistic load forecasting for ultra-short-term load forecasting are realized. Through the experiments conducted on three public datasets and one private dataset, the TFDNet model reduces the mean square error (MSE) and mean absolute error (MAE) by 19.80 and 11.20% on average, respectively, as compared with the benchmark method. These results indicate the potential applications of the TFDNet model.

SG-Transunet: A segmentation-guided Transformer U-Net model for KRAS gene mutation status identification in colorectal cancer.

Accurately identifying the Kirsten rat sarcoma virus (KRAS) gene mutation status in colorectal cancer (CRC) patients can assist doctors in deciding whether to use specific targeted drugs for treatment. Although deep learning methods are popular, they are often affected by redundant features from non-lesion areas. Moreover, existing methods commonly extract spatial features from imaging data, which neglect important frequency domain features and may degrade the performance of KRAS gene mutation status identification. To address this deficiency, we propose a segmentation-guided Transformer U-Net (SG-Transunet) model for KRAS gene mutation status identification in CRC. Integrating the strength of convolutional neural networks (CNNs) and Transformers, SG-Transunet offers a unique approach for both lesion segmentation and KRAS mutation status identification. Specifically, for precise lesion localization, we employ an encoder-decoder to obtain segmentation results and guide the KRAS gene mutation status identification task. Subsequently, a frequency domain supplement block is designed to capture frequency domain features, integrating it with high-level spatial features extracted in the encoding path to derive advanced spatial-frequency domain features. Furthermore, we introduce a pre-trained Xception block to mitigate the risk of overfitting associated with small-scale datasets. Following this, an aggregate attention module is devised to consolidate spatial-frequency domain features with global information extracted by the Transformer at shallow and deep levels, thereby enhancing feature discriminability. Finally, we propose a mutual-constrained loss function that simultaneously constrains the segmentation mask acquisition and gene status identification process. Experimental results demonstrate the superior performance of SG-Transunet over state-of-the-art methods in discriminating KRAS gene mutation status.

Detection of Mild Cognitive Impairment Through Hand Motor Function Under Digital Cognitive Test: Mixed Methods Study.

BACKGROUND: Early detection of cognitive impairment or dementia is essential to reduce the incidence of severe neurodegenerative diseases. However, currently available diagnostic tools for detecting mild cognitive impairment (MCI) or dementia are time-consuming, expensive, or not widely accessible. Hence, exploring more effective methods to assist clinicians in detecting MCI is necessary. OBJECTIVE: In this study, we aimed to explore the feasibility and efficiency of assessing MCI through movement kinetics under tablet-based "drawing and dragging" tasks. METHODS: We iteratively designed "drawing and dragging" tasks by conducting symposiums, programming, and interviews with stakeholders (neurologists, nurses, engineers, patients with MCI, healthy older adults, and caregivers). Subsequently, stroke patterns and movement kinetics were evaluated in healthy control and MCI groups by comparing 5 categories of features related to hand motor function (ie, time, stroke, frequency, score, and sequence). Finally, user experience with the overall cognitive screening system was investigated using structured questionnaires and unstructured interviews, and their suggestions were recorded. RESULTS: The "drawing and dragging" tasks can detect MCI effectively, with an average accuracy of 85% (SD 2%). Using statistical comparison of movement kinetics, we discovered that the time- and score-based features are the most effective among all the features. Specifically, compared with the healthy control group, the MCI group showed a significant increase in the time they took for the hand to switch from one stroke to the next, with longer drawing times, slow dragging, and lower scores. In addition, patients with MCI had poorer decision-making strategies and visual perception of drawing sequence features, as evidenced by adding auxiliary information and losing more local details in the drawing. Feedback from user experience indicates that our system is user-friendly and facilitates screening for deficits in self-perception. CONCLUSIONS: The tablet-based MCI detection system quantitatively assesses hand motor function in older adults and further elucidates the cognitive and behavioral decline phenomenon in patients with MCI. This innovative approach serves to identify and measure digital biomarkers associated with MCI or Alzheimer dementia, enabling the monitoring of changes in patients' executive function and visual perceptual abilities as the disease advances.