Machine learning is an effective method to predict the 3-month prognosis of patients with acute ischemic stroke.

Background and objectives: Upwards of 50% of acute ischemic stroke (AIS) survivors endure varying degrees of disability, with a recurrence rate of 17.7%. Thus, the prediction of outcomes in AIS may be useful for treatment decisions. This study aimed to determine the applicability of a machine learning approach for forecasting early outcomes in AIS patients. Methods: A total of 659 patients with new-onset AIS admitted to the Department of Neurology of both the First and Second Affiliated Hospitals of Bengbu Medical University from January 2020 to October 2022 included in the study. The patient' demographic information, medical history, Trial of Org 10,172 in Acute Stroke Treatment (TOAST), National Institute of Health Stroke Scale (NIHSS) and laboratory indicators at 24 h of admission data were collected. The Modified Rankine Scale (mRS) was used to assess the 3-mouth outcome of participants' prognosis. We constructed nine machine learning models based on 18 parameters and compared their accuracies for outcome variables. Results: Feature selection through the Least Absolute Shrinkage and Selection Operator cross-validation (Lasso CV) method identified the most critical predictors for early prognosis in AIS patients as white blood cell (WBC), homocysteine (HCY), D-Dimer, baseline NIHSS, fibrinogen degradation product (FDP), and glucose (GLU). Among the nine machine learning models evaluated, the Random Forest model exhibited superior performance in the test set, achieving an Area Under the Curve (AUC) of 0.852, an accuracy rate of 0.818, a sensitivity of 0.654, a specificity of 0.945, and a recall rate of 0.900. Conclusion: These findings indicate that RF models utilizing general clinical and laboratory data from the initial 24 h of admission can effectively predict the early prognosis of AIS patients.

A sensitive CRISPR/Cas12a-assisted fluorescent aptasensor for rapid detection of food allergens.

Food allergens elicit abnormal immune system responses among allergic individuals and sensitive detection for allergenic ingredient is greatly significant. To address this need, a novel fluorescent aptasensor, assisted by Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR), have been developed for food allergens. In this study, aptamer offers distinctive recognition capabilities in binding specific targets, while CRISPR-associated-12a protein (Cas12a) holds precise cis-cleavage for cutting fluorescent signal probes. Notably, the utilization of Cas12a cis-cleavage activity, rather than trans-cleavage, eliminates the necessity for additional fluorescent probes, thus reducing interference between substances and enhancing sensitivity. Throughout the process, complementary DNA (cDNA) plays a crucial dual role in target recognition conversion and signal presentation, representing a key challenge and innovative aspect of this study. To evaluate the performance of the aptasensor, lysozyme (LYS) is employed as a representative model target of food allergens. Under optimal conditions, the developed aptasensor could achieve an exceptional low limit of detection (LOD) of 6.10 pM with a dynamic detection range of 10 pM-320 pM. The aptasensor demonstrates high selectivity and great recovery rates. This strategy yields promising outcomes, holding the potential to serve as a valuable reference for various food allergens detection.

Highly Efficient Production of Cellulosic Ethanol from Poplar Using an Optimal C6/C5 Co-Fermentation Strain of Saccharomyces cerevisiae.

Cellulosic ethanol is the key technology to alleviate the pressure of energy supply and climate change. However, the ethanol production process, which is close to industrial production and has a high saccharification rate and ethanol yield, still needs to be developed. This study demonstrates the effective conversion of poplar wood waste into fuel-grade ethanol. By employing a two-step pretreatment using sodium chlorite (SC)-dilute sulfuric acid (DSA), the raw material achieved a sugar conversion rate exceeding 85% of the theoretical value. Under optimized conditions, brewing yeast co-utilizing C6/C5 enabled a yield of 35 g/L ethanol from 10% solid loading delignified poplar hydrolysate. We increased the solid loading to enhance the final ethanol concentration and optimized both the hydrolysis and fermentation stages. With 20% solid loading delignified poplar hydrolysate, the final ethanol concentration reached 60 g/L, a 71.4% increase from the 10% solid loading. Our work incorporates the pretreatment, enzymatic hydrolysis, and fermentation stages to establish a simple, crude poplar waste fuel ethanol process, expanding the range of feedstocks for second-generation fuel ethanol production.

Magnetic resonance imaging-based radiomics analysis for prediction of treatment response to neoadjuvant chemoradiotherapy and clinical outcome in patients with locally advanced rectal cancer: A large multicentric and validated study.

Our study investigated whether magnetic resonance imaging (MRI)-based radiomics features could predict good response (GR) to neoadjuvant chemoradiotherapy (nCRT) and clinical outcome in patients with locally advanced rectal cancer (LARC). Radiomics features were extracted from the T2 weighted (T2W) and Apparent diffusion coefficient (ADC) images of 1070 LARC patients retrospectively and prospectively recruited from three hospitals. To create radiomic models for GR prediction, three classifications were utilized. The radiomic model with the best performance was integrated with important clinical MRI features to create the combined model. Finally, two clinical MRI features and ten radiomic features were chosen for GR prediction. The combined model, constructed with the tumor size, MR-detected extramural venous invasion, and radiomic signature generated by Support Vector Machine (SVM), showed promising discrimination of GR, with area under the curves of 0.799 (95% CI, 0.760-0.838), 0.797 (95% CI, 0.733-0.860), 0.754 (95% CI, 0.678-0.829), and 0.727 (95% CI, 0.641-0.813) in the training and three validation datasets, respectively. Decision curve analysis verified the clinical usefulness. Furthermore, according to Kaplan-Meier curves, patients with a high likelihood of GR as determined by the combined model had better disease-free survival than those with a low probability. This radiomics model was developed based on large-sample size, multicenter datasets, and prospective validation with high radiomics quality score, and also had clinical utility.

Stacking-dependent interlayer magnetic interactions in CrSe2.

Recently, CrSe2, a new ferromagnetic van der Waals two-dimensional material, was discovered to be highly stable under ambient conditions, making it an attractive candidate for fundamental research and potential device applications. Here, we study the interlayer interactions of bilayer CrSe2using first-principles calculations. We demonstrate that the interlayer interaction depends on the stacking structure. The AA and AB stackings exhibit antiferromagnetic (AFM) interlayer interactions, while the AC stacking exhibits ferromagnetic (FM) interlayer interaction. Furthermore, the interlayer interaction can be further tuned by tensile strain and charge doping. Specifically, under large tensile strain, most stacking structures exhibit FM interlayer interactions. Conversely, under heavy electron doping, all stacking structures exhibit AFM interlayer interactions. These findings are useful for designing spintronic devices based on CrSe2.

Predictive nomogram for 28-day mortality risk in mitral valve disorder patients in the intensive care unit: A comprehensive assessment from the MIMIC-III database.

BACKGROUND: Mitral valve disorder (MVD) stands as the most prevalent valvular heart disease. Presently, a comprehensive clinical index to predict mortality in MVD remains elusive. The aim of our study is to construct and assess a nomogram for predicting the 28-day mortality risk of MVD patients. METHODS: Patients diagnosed with MVD were identified via ICD-9 code from the MIMIC-III database. Independent risk factors were identified utilizing the LASSO method and multivariate logistic regression to construct a nomogram model aimed at predicting the 28-day mortality risk. The nomogram's performance was assessed through various metrics including the area under the curve (AUC), calibration curves, Hosmer-Lemeshow test, integrated discriminant improvement (IDI), net reclassification improvement (NRI), and decision curve analysis (DCA). RESULTS: The study encompassed a total of 2771 patients diagnosed with MVD. Logistic regression analysis identified several independent risk factors: age, anion gap, creatinine, glucose, blood urea nitrogen level (BUN), urine output, systolic blood pressure (SBP), respiratory rate, saturation of peripheral oxygen (SpO2), Glasgow Coma Scale score (GCS), and metastatic cancer. These factors were found to independently influence the 28-day mortality risk among patients with MVD. The calibration curve demonstrated adequate calibration of the nomogram. Furthermore, the nomogram exhibited favorable discrimination in both the training and validation cohorts. The calculations of IDI, NRI, and DCA analyses demonstrate that the nomogram model provides a greater net benefit compared to the Simplified Acute Physiology Score II (SAPSII), Acute Physiology Score III (APSIII), and Sequential Organ Failure Assessment (SOFA) scoring systems. CONCLUSION: This study successfully identified independent risk factors for 28-day mortality in patients with MVD. Additionally, a nomogram model was developed to predict mortality, offering potential assistance in enhancing the prognosis for MVD patients. It's helpful in persuading patients to receive early interventional catheterization treatment, for example, transcatheter mitral valve replacement (TMVR), transcatheter mitral valve implantation (TMVI).

High-performance terahertz modulators induced by substrate field in Te-based all-2D heterojunctions.

High-performance active terahertz modulators as the indispensable core components are of great importance for the next generation communication technology. However, they currently suffer from the tradeoff between modulation depth and speed. Here, we introduce two-dimensional (2D) tellurium (Te) nanofilms with the unique structure as a new class of optically controlled terahertz modulators and demonstrate their integrated heterojunctions can successfully improve the device performances to the optimal and applicable levels among the existing all-2D broadband modulators. Further photoresponse measurements confirm the significant impact of the stacking order. We first clarify the direction of the substrate-induced electric field through first-principles calculations and uncover the unusual interaction mechanism in the photoexcited carrier dynamics associated with the charge transfer and interlayer exciton recombination. This advances the fundamental and applicative research of Te nanomaterials in high-performance terahertz optoelectronics.

Factors influencing public participation behavior relating to government microblogs on COVID-19 updates.

Introduction: During the global COVID-19 pandemic, densely populated megacities engaged in active international exchanges have faced the most severe impacts from both the disease and the associated infodemic. This study examines the factors influencing public participation behavior on government microblogs in these megacities during the pandemic. It guides megacities in disseminating epidemic information, promoting knowledge on epidemic prevention, managing public opinion, and addressing related matters. Methods: Utilizing the elaboration likelihood model's central and peripheral routes, drawing on an empirical analysis of 6,677 epidemic-related microblogs from seven Chinese megacities, this study analyses the influence mechanisms influencing public participation behavior and reveals the regulatory role of confirmed case numbers. Meanwhile,a qualitative comparative analysis examines and discusses diferent confgurations of ixn fuential factors. Results: The study reveals that microblog content richness demonstrates a U-shaped impact on public participation behavior. Conversely, content interaction, content length, and the number of fans positively impact participation, while update frequency has a negative impact. Additionally, the number of new confrmed cases positively regulates the impact of microblog content and publisher characteristics on public participation behavior. Public participation behavior also varies based on publishing time and content semantic features. This study further revealed the different confgurations of influential factors by QCA method. Conclusion: This study reveals the impact mechanism of the microblog content and publisher characteristics on public participation behavior. It also demonstrates the regulatory role of newly confrmed cases in the way content and publishers' characteristics influence public participation behavior. This study is of great significance for the operation of government microblogs, the release of emergency information, and the promotion of public participation.

Efficient Inverted Perovskite Photovoltaics Through Surface State Manipulation.

Inverted perovskite solar cells (PSCs) are considered as the most promising avenue for the commercialization of PSCs due to their potential inherent stability. However, suboptimal interface contacts between electron transport layer (ETL) (such as C60 ) and the perovskite absorbing layer within inverted PSCs always result in reduced efficiency and poor stability. Herein, a surface state manipulation strategy has been developed by employing a highly electronegative 4-fluorophenethylamine hydrochloride (p-F-PEACl) to effectively address the issue of poor interface contacts in the inverted PSCs. The p-F-PEACl demonstrates a robust interaction with perovskite film through bonding of amino group and Cl- with I- and Pb2+ ions in the perovskite, respectively. As such, the surface defects of perovskite film can be significantly reduced, leading to suppressed non-radiative recombination. Moreover, p-F-PEACl also plays a dual role in enhancing the surface potential and improving energy-level alignment at the interfaces between the perovskite and C60 carrier transport layer, which directly contributes to efficient charge extraction. Finally, the open-circuit voltage (Voc ) of devices increases from 1.104 V to 1.157 V, leading to an overall efficiency improvement from 22.34% to 24.78%. Furthermore, the p-F-PEACl-treated PSCs also display excellent stability.

Corynebacterium glutamicum cell factory design for the efficient production of cis, cis-muconic acid.

Cis, cis-muconic acid (MA) is widely used as a key starting material in the synthesis of diverse polymers. The growing demand in these industries has led to an increased need for MA. Here, we constructed recombinant Corynebacterium glutamicum by systems metabolic engineering, which exhibit high efficiency in the production of MA. Firstly, the three major degradation pathways were disrupted in the MA production process. Subsequently, metabolic optimization strategies were predicted by computational design and the shikimate pathway was reconstructed, significantly enhancing its metabolic flux. Finally, through optimization and integration of key genes involved in MA production, the recombinant strain produced 88.2 g/L of MA with the yield of 0.30 mol/mol glucose in the 5 L bioreactor. This titer represents the highest reported titer achieved using glucose as the carbon source in current studies, and the yield is the highest reported for MA production from glucose in Corynebacterium glutamicum. Furthermore, to enable the utilization of more cost-effective glucose derived from corn straw hydrolysate, we subjected the strain to adaptive laboratory evolution in corn straw hydrolysate. Ultimately, we successfully achieved MA production in a high solid loading of corn straw hydrolysate (with the glucose concentration of 83.56 g/L), resulting in a titer of 19.9 g/L for MA, which is 4.1 times higher than that of the original strain. Additionally, the glucose yield was improved to 0.33 mol/mol. These provide possibilities for a greener and more sustainable production of MA.

Insight into the Feasibility of Fatty Acyl Chlorides with 10-18 Carbons for the Ball-Milling Synthesis of Thermoplastic Cellulose Esters.

Fatty acid cellulose esters (FACE) are common cellulose-based thermoplastics, and their thermoplasticity is determined by both the contents and the lengths of the side chains. Herein, various FACE were synthesized by the ball-milling esterification of cellulose and fatty acyl chlorides containing 10-18 carbons, and their structures and thermoplasticity were thoroughly studied. The results showed that FACE with high degrees of substitution (DS) and low melting flow temperatures (Tf) were achieved as the chain lengths of the fatty acyl chlorides were reduced. In particular, a cellulose decanoate with a DS of 1.85 and a Tf of 186 °C was achieved by feeding 3 mol of decanoyl chloride per mole anhydroglucose units of cellulose. However, cellulose stearate (DS = 1.53) synthesized by the same protocols cannot melt even at 250 °C. More interestingly, the fatty acyl chlorides with 10 and 12 carbons resulted in FACE with superior toughness (elongation at break up to 94.4%). In contrast, due to their potential crystallization of the fatty acyl groups with 14-18 carbons, the corresponding FACE showed higher tensile strength and Young's modulus than the others. This study provides some theoretical basis for the mechanochemical synthesis of thermoplastic FACE with designated properties.

Enhancing brain metastasis prediction in non-small cell lung cancer: a deep learning-based segmentation and CT radiomics-based ensemble learning model.

BACKGROUND: Brain metastasis (BM) is most common in non-small cell lung cancer (NSCLC) patients. This study aims to enhance BM risk prediction within three years for advanced NSCLC patients by using a deep learning-based segmentation and computed tomography (CT) radiomics-based ensemble learning model. METHODS: This retrospective study included 602 stage IIIA-IVB NSCLC patients, 309 BM patients and 293 non-BM patients, from two centers. Patients were divided into a training cohort (N = 376), an internal validation cohort (N = 161) and an external validation cohort (N = 65). Lung tumors were first segmented by using a three-dimensional (3D) deep residual U-Net network. Then, a total of 1106 radiomics features were computed by using pretreatment lung CT images to decode the imaging phenotypes of primary lung cancer. To reduce the dimensionality of the radiomics features, recursive feature elimination configured with the least absolute shrinkage and selection operator (LASSO) regularization method was applied to select the optimal image features after removing the low-variance features. An ensemble learning algorithm of the extreme gradient boosting (XGBoost) classifier was used to train and build a prediction model by fusing radiomics features and clinical features. Finally, Kaplan‒Meier (KM) survival analysis was used to evaluate the prognostic value of the prediction score generated by the radiomics-clinical model. RESULTS: The fused model achieved area under the receiver operating characteristic curve values of 0.91 ± 0.01, 0.89 ± 0.02 and 0.85 ± 0.05 on the training and two validation cohorts, respectively. Through KM survival analysis, the risk score generated by our model achieved a significant prognostic value for BM-free survival (BMFS) and overall survival (OS) in the two cohorts (P < 0.05). CONCLUSIONS: Our results demonstrated that (1) the fusion of radiomics and clinical features can improve the prediction performance in predicting BM risk, (2) the radiomics model generates higher performance than the clinical model, and (3) the radiomics-clinical fusion model has prognostic value in predicting the BMFS and OS of NSCLC patients.

Modulation of Charge Transport from Two-Dimensional Perovskites to Industrial Charge Transport Layers by the Organic Spacer-Dependent Exciton-Phonon Interactions.

In the past decade, two-dimensional (2D) perovskite surface treatment has emerged as a promising strategy to improve the performance of three-dimensional (3D) perovskite solar cells (PSCs). However, systematic studies on the impact of organic spacers of 2D perovskites on charge transport in 2D/3D PSCs are still lacking. Here, using 2D perovskite film/C60 heterostructures with different organic spacers [butylamine (BA), phenylethylamine (PEA), and 3-fluorophenethylamine (m-F-PEA)], we systematically investigated the carrier diffusion and interfacial transfer process. Using a 2D perovskite film with a thickness of ∼7 nm, we observed subtle differences in electron transfer time between 2D perovskites and C60 layers, which can be attributed to limited thickness and similar electron coupling strength. However, with the thickness of 2D perovskite increasing, electron transfer efficiency in the (BA)2PbI4/C60 heterostructure exhibits the most rapid decrease due to poor carrier diffusion of (BA)2PbI4 caused by stronger exciton-phonon interactions compared to (PEA)2PbI4 and (m-F-PEA)2PbI4 in thickness-dependent charge transfer research. Meanwhile, the fill factor of 2D/3D PSC treated with BAI exhibits the most rapid decrease compared to PEAI- and m-F-PEAI-treated 2D/3D PSCs with the concentration increase of passivators. This study indicates that it is easier to enhance open-circuit voltages and minimize the decrease of fill factor by increasing the concentration of passivators in 2D/3D PSCs when using passivators with a rigid molecular structure.

Polar metals in strain-engineered KNbO3/CaNbO3 superlattices: a first-principles study.

Polar metals have generated significant interest since the ferroelectric-like structural transition in metallic LiOsO3 was discovered. Herein, we report on a strain-modulated polar metal in the ferroelectric/metal superlattice of 1 : 1 KNbO3/CaNbO3. Using first-principles calculations, we have investigated the structural distortions, including polar distortions and octahedral rotations, and layer-by-layer electronic structures in the KNbO3/CaNbO3 superlattice under different epitaxial strains. Along the stacking direction, the superlattice has almost parallel polar displacements under compressive strain, whereas both in-plane and out-of-plane antiferroelectric-like polar displacements are robust under intermediate strain, which is connected to the octahedral tilting pattern and interlayer electron transfer. In addition, the in-plane polar distortions are enhanced by tensile strains and have a sudden increase at 4% tensile strain. The metallicity is mainly contributed by d electrons from Nb atoms. And orbital-resolved electron distributions in each layer show that d-orbital splitting is related not only to the epitaxial strain but also to the direction of polar displacements. Our results suggest an efficient way to tune polar distortions as well as local metallicity via epitaxial strains in the superlattice.

Multi-center evaluation of machine learning-based radiomic model in predicting disease free survival and adjuvant chemotherapy benefit in stage II colorectal cancer patients.

BACKGROUND: Our study aimed to explore the potential of radiomics features derived from CT images in predicting the prognosis and response to adjuvant chemotherapy (ACT) in patients with Stage II colorectal cancer (CRC). METHODS: A total of 478 patients with confirmed stage II CRC, with 313 from Shanghai (Training set) and 165 from Beijing (Validation set) were enrolled. Optimized features were selected using GridSearchCV and Iterative Feature Elimination (IFE) algorithm. Subsequently, we developed an ensemble random forest classifier to predict the probability of disease relapse.We evaluated the performance of the model using the concordance index (C-index), precision-recall curves, and area under the precision-recall curves (AUCPR). RESULTS: A radiomic model (namely the RF5 model) consisting of four radiomics features and T stage were developed. The RF5 model performed better than simple radiomics features or T stage alone, with higher C-index and AUCPR, as well as better sensitivity and specificity (C-indexRF5: 0.836; AUCPR = 0.711; Sensitivity = 0.610; Specificity = 0.935). We identified an optimal cutoff value of 0.1215 to split patients into high- or low-score subgroups, with those in the low-score group having better disease-free survival (DFS) (Training Set: P = 1.4e-11; Validation Set: P = 0.015). Furthermore, patients in the high-score group who received ACT had better DFS compared to those who did not receive ACT (P = 0.04). However, no statistical difference was found in low-score patients (P = 0.17). CONCLUSION: The radiomic model can serve as a reliable tool for assessing prognosis and identifying the optimal candidates for ACT in Stage II CRC patients. TRIAL REGISTRATION: Retrospectively registered.

Size and number of lymph nodes were risk factors of recurrence in stage II colorectal cancer.

BACKGROUND: Size and number of lymph nodes (LNs) were reported to be associated with the prognosis of stage II colorectal cancer (CRC). The purpose of this study was to determine the prognostic role of the size of LNs (SLNs) measured by computer tomography (CT) and the number of retrieved LNs (NLNs) in the relapse-free survival (RFS) and overall survival (OS) among stage II CRC patients. METHODS: Consecutive patients diagnosed with stage II CRC at Fudan University Shanghai Cancer Center (FUSCC) from January 2011 to December 2015 were reviewed, and 351 patients were randomly divided into two cohorts for cross-validation. The optimal cut-off values were obtained using X-tile program. Kaplan-Meier curves and Cox regression analyses were conducted for the two cohorts. RESULTS: Data from 351 stage II CRC patients were analyzed. The cut-off values for SLNs and NLNs were 5.8 mm and 22, respectively, determined by the X-tile in the training cohort. In the validation cohort, Kaplan-Meier curves demonstrated SLNs (P = 0.0034) and NLNs (P = 0.0451) were positively correlated with RFS but not with OS. The median follow-up time in the training cohort and the validation cohort were 60.8 months and 61.0 months respectively. Univariate and multivariate analysis revealed that both SLNs (training cohort: Hazard Ratio (HR) = 2.361, 95% Confidence interval (CI): 1.044-5.338, P = 0.039; validation cohort: HR = 2.979, 95%CI: 1.435-5.184, P = 0.003) and NLNs (training cohort: HR = 0.335, 95%CI: 0.113-0.994, P = 0.049; validation cohort: HR = 0.375, 95%CI: 0.156-0.900, P = 0.021) were independent prognostic factors for RFS whereas not for OS. CONCLUSION: SLNs and NLNs are independent prognostic factors for patients with stage II CRC. Patients with SLNs > 5.8 mm and NLNs ≤ 22 are apt to have higher risk of recurrence.

Reservoir Lithology Identification Based on Multicore Ensemble Learning and Multiclassification Algorithm Based on Noise Detection Function.

Reservoir lithology identification is an important part of well logging interpretation. The accuracy of identification affects the subsequent exploration and development work, such as reservoir division and reserve prediction. Correct reservoir lithology identification has important geological significance. In this paper, the wavelet threshold method will be used to preliminarily reduce the noise of the curve, and then the MKBoost-MC model will be used to identify the reservoir lithology. It is found that the prediction accuracy of MKBoost-MC is higher than that of the traditional SVM algorithm, and though the operation of MKBoost-MC takes a long time, the speed of MKBoost-MC reservoir lithology identification is much higher than that of manual processing. The accuracy of MKBoost-MC for reservoir lithology recognition can reach the application standard. For the unbalanced distribution of lithology types, the MKBoost-MC algorithm can be effectively suppressed. Finally, the MKBoost-MC reservoir lithology identification method has good applicability and practicality to the lithology identification problem.

Special issue "The advance of solid tumor research in China": Prognosis prediction for stage II colorectal cancer by fusing computed tomography radiomics and deep-learning features of primary lesions and peripheral lymph nodes.

Currently, the prognosis assessment of stage II colorectal cancer (CRC) remains a difficult clinical problem; therefore, more accurate prognostic predictors must be developed. In our study, we developed a prognostic prediction model for stage II CRC by fusing radiomics and deep-learning (DL) features of primary lesions and peripheral lymph nodes (LNs) in computed tomography (CT) scans. First, two CT radiomics models were built using primary lesion and LN image features. Subsequently, an information fusion method was used to build a fusion radiomics model by combining the tumor and LN image features. Furthermore, a transfer learning method was applied to build a deep convolutional neural network (CNN) model. Finally, the prediction scores generated by the radiomics and CNN models were fused to improve the prognosis prediction performance. The disease-free survival (DFS) and overall survival (OS) prediction areas under the curves (AUCs) generated by the fusion model improved to 0.76 ± 0.08 and 0.91 ± 0.05, respectively. These were significantly higher than the AUCs generated by the models using the individual CT radiomics and deep image features. Applying the survival analysis method, the DFS and OS fusion models yielded concordance index (C-index) values of 0.73 and 0.9, respectively. Hence, the combined model exhibited good predictive efficacy; therefore, it could be used for the accurate assessment of the prognosis of stage II CRC patients. Moreover, it could be used to screen out high-risk patients with poor prognoses, and assist in the formulation of clinical treatment decisions in a timely manner to achieve precision medicine.

Bioinformatics analysis on differentially expressed genes between colorectal adenoma and colorectal adenocarcinoma.

BACKGROUND: Colorectal adenoma (CRA) is the main cause of the progression of Colorectal adenocarcinoma (COAD). Therefore, it is very important to accurately reveal its developmental mechanism. METHODS: Differential expression genes (DEGs) in three microarray datasets were screened using GEO and GEO2R. R packages were used for gene ontology (GO) functional annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) path enrichment analysis. Hub genes screened by STRING, Cytoscape and CytoHubba were used. R was used for DEGs of hub genes, and Gene Expression Profiling Interactive Analysis (GEPIA2) database was used for prognostic Analysis. R-packet were used to analyze tumor pathology, tumour, lymph-nodes, and metastases (TNM) staging, enrichment, immune invasion and prognosis. RESULTS: Among the 66 genes, including 36 up-regulated and 30 down-regulated genes. Survival analysis showed that COL1A1, COL5A2, COL5A1 and secreted protein acidic and rich in cysteine (SPARC) were associated with disease-free survival in patients. The four genes were related to tumor pathological stage, TNM stage and immune invasion. COL1A1 and COL5A2 were highly expressed in chromatin modification and cellular senescence. Low expression of COL5A1 and SPARC was significantly enriched in neutrophil degranulation and Wp VegfavegFR2 signaling pathways. CONCLUSIONS: Obviously, these four key genes can serve as important targets for early diagnosis, treatment, immunity and prognosis of CRA to COAD.

An Attention-Guided Spatiotemporal Graph Convolutional Network for Sleep Stage Classification.

Sleep staging has been widely used as an approach in sleep diagnoses at sleep clinics. Graph neural network (GNN)-based methods have been extensively applied for automatic sleep stage classifications with significant results. However, the existing GNN-based methods rely on a static adjacency matrix to capture the features of the different electroencephalogram (EEG) channels, which cannot grasp the information of each electrode. Meanwhile, these methods ignore the importance of spatiotemporal relations in classifying sleep stages. In this work, we propose a combination of a dynamic and static spatiotemporal graph convolutional network (ST-GCN) with inter-temporal attention blocks to overcome two shortcomings. The proposed method consists of a GCN with a CNN that takes into account the intra-frame dependency of each electrode in the brain region to extract spatial and temporal features separately. In addition, the attention block was used to capture the long-range dependencies between the different electrodes in the brain region, which helps the model to classify the dynamics of each sleep stage more accurately. In our experiments, we used the sleep-EDF and the subgroup III of the ISRUC-SLEEP dataset to compare with the most current methods. The results show that our method performs better in accuracy from 4.6% to 5.3%, in Kappa from 0.06 to 0.07, and in macro-F score from 4.9% to 5.7%. The proposed method has the potential to be an effective tool for improving sleep disorders.