Least Absolute Shrinkage And Selection Operator Model Research Articles

Development and external validation of a novel score for predicting postoperative 30‑day mortality in tumor craniotomy patients: A cross‑sectional diagnostic study.

The identification of patients with craniotomy at high risk for postoperative 30-day mortality may contribute to achieving targeted delivery of interventions. The present study aimed to develop a personalized nomogram and scoring system for predicting the risk of postoperative 30-day mortality in such patients. In this retrospective cross-sectional study, 18,642 patients with craniotomy were stratified into a training cohort (n=7,800; year of surgery, 2012-2013) and an external validation cohort (n=10,842; year of surgery, 2014-2015). The least absolute shrinkage and selection operator (LASSO) model was used to select the most important variables among the candidate variables. Furthermore, a stepwise logistic regression model was established to screen out the risk factors based on the predictors chosen by the LASSO model. The model and a nomogram were constructed. The area under the receiver operating characteristic (ROC) curve (AUC) and calibration plot analysis were used to assess the model's discrimination ability and accuracy. The associated risk factors were categorized according to clinical cutoff points to create a scoring model for postoperative 30-day mortality. The total score was divided into four risk categories: Extremely high, high, intermediate and low risk. The postoperative 30-day mortality rates were 2.43 and 2.58% in the training and validation cohort, respectively. A simple nomogram and scoring system were developed for predicting the risk of postoperative 30-day mortality according to the white blood cell count; hematocrit and blood urea nitrogen levels; age range; functional health status; and incidence of disseminated cancer cells. The ROC AUC of the nomogram was 0.795 (95% CI: 0.764 to 0.826) in the training cohort and it was 0.738 (95% CI: 0.7091 to 0.7674) in the validation cohort. The calibration demonstrated a perfect fit between the predicted 30-day mortality risk and the observed 30-day mortality risk. Low, intermediate, high and extremely high risk statuses for 30-day mortality were associated with total scores of (-1.5 to -1), (-0.5 to 0.5), (1 to 2) and (2.5 to 9), respectively. A personalized nomogram and scoring system for predicting postoperative 30-day mortality in adult patients who underwent craniotomy were developed and validated, and individuals at high risk of 30-day mortality were able to be identified.

Construction and validation of a prognostic model for stemness-related genes in lung adenocarcinoma.

Lung adenocarcinoma (LUAD) is the most common histological type of lung cancer with poor overall prognosis. Early identification of high-risk patients and individualized treatment can help extend the survival time of patients. This study aimed to construct and validate a prognostic prediction least absolute shrinkage and selection operator (LASSO) model for stemness-related genes in LUAD. Firstly, LUAD RNA-sequencing data and clinical data were downloaded from The Cancer Genome Atlas (TCGA) database. The tumor stemness index based on mRNA expression (mRNAsi) was calculated, and the relationship between mRNAsi and the survival prognosis as well as clinical features of LUAD patients was analyzed. Then, the weighted gene co-expression network analysis (WGCNA) method was used to screen for gene modules highly correlated with mRNAsi, and functional annotation [Gene Ontology (GO) analysis] and pathway enrichment analysis [Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis] were performed for the selected stemness-related gene module. Furthermore, prognosis-associated genes were determined from the stemness-related genes through univariate Cox analysis, and a prognostic model was constructed using LASSO analysis. Finally, a series of validations including survival curve analysis, receiver operating characteristic (ROC) curve analysis, and risk analysis were conducted for the prognostic model, and nomogram based on the risk model and various clinicopathological features were constructed. LUAD patients with high mRNAsi had a higher mortality rate than those with low mRNAsi. GO analysis showed that stemness-related genes were mainly involved in mRNA processing and extracellular matrix organization, while KEGG analysis revealed their involvement in cell cycle and PI3K-Akt signaling pathways. A prognostic model based on 12 stemness-related genes was constructed using LASSO regression. Validation of the prognostic model demonstrated its good accuracy in predicting the prognosis of LUAD patients. mRNAsi plays an important role in the occurrence and development of LUAD. This study successfully constructed a prognostic prediction LASSO model for stemness-related genes in LUAD, which can serve as a novel prognostic indicator for LUAD and may be an effective complement to the current Tumor Node Metastasis (TNM) clinical staging of LUAD.

Value of serum brain-derived neurotrophic factor and glial fibrillary acidic protein for detecting depression in patients with Helicobacter pylori infection

ObjectiveInfection with helicobacter pylori (H. pylori) is associated with depression, and depression can affect the outcome of H. pylori treatment. This study aimed to evaluate the value of serum brain-derived neurotrophic factor (BDNF) and glial fibrillary acidic protein (GFAP) for predicting depression in H. pylori-positive patients. MethodA total of 82H. pylori-positive and 82H. pylori-negative patients were recruited for this study. All patients underwent neuropsychological and gastrointestinal assessments and blood sampling. BDNF and GFAP levels were measured in serum. The least absolute shrinkage and selection operator (LASSO) model was used to determine a composite marker. ResultsH. pylori-positive patients showed significantly increased serum GFAP levels and significantly decreased serum BDNF levels compared to H. pylori-negative patients. Among H. pylori-positive patients, serum levels of gastrin 17 (G-17), pepsinogen (PG) I/PGII, BDNF, and GFAP, as well as Gastrointestinal Symptom Rating Scale (GSRS) scores, were significantly correlated with Hamilton Depression Scale (HAMD-24) overall scores and factor scores. Interactions between serum BDNF/GFAP and gastrointestinal serum indices or GSRS scores were significantly associated with HAMD-24 scores in H. pylori-positive patients. The LASSO model indicated that the combination of serum BDNF, GFAP, and G-17 and GSRS scores could identify H. pylori-positive patients with depression with an area under the curve of 0.879. ConclusionCirculating changes in BDNF and GFAP were associated with the occurrence of depression in H. pylori-positive patients. A composite marker including neural and gastrointestinal function-related indices may be of value for identifying depression among H. pylori-positive patients.

LASSO-derived model for the prediction of lean-non-alcoholic fatty liver disease in examinees attending a routine health check-up

Background Lean individuals with non-alcohol fatty liver disease (NAFLD) often have normal body size but abnormal visceral fat. Therefore, an alternative to body mass index should be considered for prediction of lean-NAFLD. This study aimed to use representative visceral fat links with other laboratory parameters using the least absolute shrinkage and selection operator (LASSO) method to construct a predictive model for lean-NAFLD. Methods This retrospective cross-sectional analysis enrolled 2325 subjects with BMI < 24 kg/m2 from medical records of 51,271 examinees who underwent a routine health check-up. They were randomly divided into training and validation cohorts at a ratio of 1:1. The LASSO-derived prediction model used LASSO regression to select 23 clinical and laboratory factors. The discrimination and calibration abilities were evaluated using the Hosmer–Lemeshow test and calibration curves. The performance of the LASSO model was compared with the fatty liver index (FLI) model. Results The LASSO-derived model included four variables—visceral fat, triglyceride levels, HDL-C-C levels, and waist hip ratio—and demonstrated superior performance in predicting lean-NAFLD with high discriminatory ability (AUC, 0.8416; 95% CI: 0.811–0.872) that was comparable with the FLI model. Using a cut-off of 0.1484, moderate sensitivity (75.69%) and specificity (79.86%), as well as high negative predictive value (95.9%), were achieved in the LASSO model. In addition, with normal WC subgroup analysis, the LASSO model exhibits a trend of higher accuracy compared to FLI (cut-off 15.45). Conclusions We developed a LASSO-derived predictive model with the potential for use as an alternative tool for predicting lean-NAFLD in clinical settings.

The Ratio Between the Infrarenal and Suprarenal Aortic Diameter Is a Predictor of Durable Proximal Seal After Endovascular Aneurysm Repair.

The aim of this study was to assess whether the ratio of the mean infrarenal neck diameter to the suprarenal aortic diameter is a predictor for a durable proximal seal after endovascular aneurysm repair (EVAR). A total of 439 patients who underwent elective EVAR between 2004 and 2018 in a single vascular referral center met our inclusion criteria. Ratios were calculated by dividing the mean infrarenal neck diameter by 4 different suprarenal aortic diameters. Patients who developed a late type 1A endoleak (n=20) or proximal neck dilatation mandating revision (n=8) were compared with the 411 patients without long-term proximal seal complications. Patients who developed a late type 1A endoleak had more frequently hypertension, a shorter infrarenal neck length, and a larger mean infrarenal neck diameter. The ratio of the mean infrarenal neck diameter to all 4 suprarenal aortic diameters was higher in the late type 1A endoleak group compared with the group without a late type 1A endoleak. Least absolute shrinkage and selection operator (LASSO) logistic regression identified a combination of 6 variables as the best combination to predict a late type 1A endoleak: presence of hypertension, increased mean infrarenal neck diameter, decreased aneurysm neck length, larger ratio of the mean infrarenal neck diameter to the aortic diameter proximal to the superior mesenteric artery (SMA), larger ratio of the mean infrarenal neck diameter to the aortic diameter proximal to the upper renal artery, and increased β-angle. Of these, based on both the univariate area under the curve (AUC) and optimal LASSO model restricted to a single predictor, the ratio of the mean infrarenal neck diameter to the aortic diameter proximal to the SMA (AUC, 0.770; cutoff value, 0.997) was considered the best prognostic variable. The ratio of the mean infrarenal neck diameter to the aortic diameter proximal to the SMA is a good predictor for a late type 1A endoleak. Patients with mean infrarenal neck diameter larger than the diameter proximal to the SMA (ratio >1) are at risk for a late type 1A endoleak. In this single-center, retrospective cohort study, we found that the ratio of the mean infrarenal neck diameter to the aortic diameter proximal to the SMA is a good predictor for a late type 1A endoleak. We conclude that the suprarenal diameter must be taken into account before assessing endovascular aortic aneurysm repair eligibility. Patients with a ratio >1 may not be the best candidates for a durable result after EVAR and may be better off with fenestrated EVAR or open repair.

Plasma-derived exosomal miRNA profiles associated with type 1 diabetes.

Recently, exosomal miRNAs have been shown to play important roles in multiple diseases, including type 1 diabetes (T1D). To assess the biomarker potential of exosomal miRNAs for T1D, we measured the expression profiles of plasma-derived exosomal miRNAs in T1D and explored their potential functions by bioinformatic analysis. In the discovery phase, exosome samples were isolated from plasma by size exclusion chromatography from 10 T1D patients and 10 sex- (p=0.36), age- (p=0.97), and body mass index-matched (p=0.47) healthy control subjects. Exosomal miRNA expression profiles were measured using the Illumina NovaSeq 6000 platform. With verification by quantitative real-time PCR (qRT-PCR), we used multiple bioinformatics approaches to explore the potential biological functions of the identified differentially expressed miRNAs. The diagnostic signature of exosomal miRNAs was evaluated by least absolute shrinkage and selection operator (LASSO) regression and evaluated based on the area under the receiver operating characteristic curve (AUC). In total, 43 differentially expressed miRNAs, among which 34 were upregulated and 9 were downregulated, were identified in T1D. After correcting for multiple testing using false discovery rate, 11 identified exosomal miRNAs still showed statistical significance. Among the 5 selected miRNAs, 3 miRNAs (miR-103a-3p, miR-144-5p and miR-454-3p) were successfully validated by qRT-PCR. The biological analysis-enriched terms included protein autophosphorylation and the Hedgehog signalling pathway. The highest AUC of exosomal miRNA was 0.889 under the LASSO model. The expression levels of 5 selected exosomal miRNAs were correlated with multiple clinical characteristics such as fasting C-peptide and postprandial C-peptide. Our results indicated that plasma-derived exosomal miRNAs could serve as promising diagnostic biomarkers of T1D.

P216 Urine Metabolomic Characteristics of Anxiety and Depression Disorders in Patients with Inflammatory Bowel Disease

Abstract Background Up to 25%-35% of patients with inflammatory bowel disease (IBD) suffer from anxiety or depression. We aimed to depict the urine metabolic profile of IBD patients with anxiety and depression disorders and investigate the urinary metabolites that could characterize the mental disorders in IBD patients. Methods Urine samples were collected from IBD patients, and the Liquid Chromatography-Mass Spectrometry (LC-MS) was used to detect the metabolites. The Hospital Anxiety and Depression Scale (HADS) was used to assess the anxiety and depression symptoms. Least absolute shrinkage and selection operator (LASSO) regression model with 10-fold cross-validation was used to establish a metabolite-based prediction model for mental disorders in patients with IBD. Results Among the involved IBD patients (n=23, median age 41, women 61%), 35% had anxiety and 35% had depression. The urine metabolic profiles of IBD patients with anxiety or depression disorders were largely separated from those without mental symptoms (Figure 1 a-b). There were 36 metabolites up-regulated and 25 metabolites down-regulated in the anxiety group, while 10 metabolites up-regulated and 35 metabolites down-regulated in the depression group (Figure 1 c-d). The LASSO model identified the metabolites predicting IBD patients with mental disorders (Figure 2 a-b). Six metabolites (L-Carnitine, Riboflavin-5-phosphate, 2-Ethyl-2-phenylmalonamide, 3-Indoxyl sulphate, Sodium cholate, Tritylalcohol) were selected to construct the prediction model for anxiety, and another six metabolites (D-(-)-Glutamine, dAMP, DL-Stachydrine, dCMP, 5'-Deoxy-5'-(Methylthio)Adenosine, Sakuranin) were selected to construct the prediction model for depression in IBD patients. Both models showed excellent performances with an AUC of 0.992 (95%CI:0.97-1.00) and 0.933 (95%CI: 0.83-1.00) for predicting anxiety and depression, respectively, among patients with IBD (Figure 2 c-d). Conclusion We found significantly different urine metabolic profiles between IBD patients with and without anxiety and depression disorders. The specific urinary metabolites panel may become effective biomarkers for the prediction of anxiety and depression disorders in IBD patients.

Radiomics analysis to predict pulmonary nodule malignancy using machine learning approaches

BackgroundLow-dose CT screening can reduce lung cancer-related mortality. However, most screen-detected pulmonary abnormalities do not develop into cancer and it often remains challenging to identify malignant nodules, particularly among indeterminate...

Coronary heart disease classification using deep learning approach with feature selection for improved accuracy.

Coronary heart disease (CHD) is one of the deadliest diseases and a risk prediction model for cardiovascular conditions is needed. Due to the huge number of features that lead to heart problems, it is often difficult for an expert to evaluate these huge features into account. So, there is a need of appropriate feature selection for the given CHD dataset. For early CHD detection, deep learning modes (DL) show promising results in the existing studies. This study aimed to develop a deep convolution neural network (CNN) model for classification with a selected number of efficient features using the LASSO (least absolute shrinkage and selection operator) technique. Also, aims to compare the model with similar studies and analyze the performance of the proposed model using accuracy measures. The CHD dataset of NHANES (National Health and Nutritional Examination Survey) was examined with 49 features using LASSO technique. This research work is an attempt to apply an improved CNN model for the classification of the CHD dataset with huge features CNN model with feature extractor consists of a fully connected layer with two convolution 1D layers, and classifier part consists of two fully connected layers with SoftMax function was trained on this dataset. Metrics like accuracy recall, specificity, and ROC were used for the evaluation of the proposed model. The feature selection was performed by applying the LASSO model. The proposed CNN model achieved 99.36% accuracy, while previous studies model achieved over 80 to 92% accuracy. The application of the proposed CNN with the LASSO model for the classification of CHD can speed up the diagnosis of CHD and appears to be effective in predicting cardiovascular disease based on risk features.

Use of Machine Learning for the Identification and Validation of Immunogenic Cell Death Biomarkers and Immunophenotypes in Coronary Artery Disease.

Immunogenic cell death (ICD) is part of the immune system's response to coronary artery disease (CAD). In this study, we bioinformatically evaluated the diagnostic and therapeutic utility of immunogenic cell death-related genes (IRGs) and their relationship with immune infiltration features in CAD. We acquired the CAD-related datasets GSE12288, GSE71226, and GSE120521 from the Gene Expression Omnibus (GEO) database and the IRGs from the GeneCards database. After identifying the immune cell death-related differentially expressed genes (IRDEGs), we developed a risk model and detected immune subtypes in CAD. IRDEGs were identified using least absolute shrinkage and selection operator (LASSO) analysis. Using a nomogram, we confirmed that both the LASSO model and ICD signature genes had good diagnostic performance. There was a high degree of coincidence and immune representativeness between two CAD groups based on characteristic genes and hub genes. Hub genes were associated with the interaction of neuroactive ligands with receptors and cell adhesion receptors. The two groups differed in terms of adipogenesis, allograft rejection, and apoptosis, as well as the ICD signature and hub gene expression levels. The two CAD-ICD subtypes differed in terms of immune infiltration. Quantitative real-time PCR (qRT-PCR) correlated CAD with the expression of OAS3, ITGAV, and PIBF1. The ICD signature genes are candidate biomarkers and reference standards for immune grouping in CAD and can be beneficial in precise immune-targeted therapy.

A nomogram model for predicting 5-year risk of prediabetes in Chinese adults

Early identification is crucial to effectively intervene in individuals at high risk of developing pre-diabetes. This study aimed to create a personalized nomogram to determine the 5-year risk of pre-diabetes among Chinese adults. This retrospective cohort study included 184,188 participants without prediabetes at baseline. Training cohorts (92,177) and validation cohorts (92,011) were randomly assigned (92,011). We compared five prediction models on the training cohorts: full cox proportional hazards model, stepwise cox proportional hazards model, multivariable fractional polynomials (MFP), machine learning, and least absolute shrinkage and selection operator (LASSO) models. At the same time, we validated the above five models on the validation set. And we chose the LASSO model as the final risk prediction model for prediabetes. We presented the model with a nomogram. The model's performance was evaluated in terms of its discriminative ability, clinical utility, and calibration using the area under the receiver operating characteristic (ROC) curve, decision curve analysis, and calibration analysis on the training cohorts. Simultaneously, we also evaluated the above nomogram on the validation set. The 5-year incidence of prediabetes was 10.70% and 10.69% in the training and validation cohort, respectively. We developed a simple nomogram that predicted the risk of prediabetes by using the parameters of age, body mass index (BMI), fasting plasma glucose (FBG), triglycerides (TG), systolic blood pressure (SBP), and serum creatinine (Scr). The nomogram's area under the receiver operating characteristic curve (AUC) was 0.7341 (95% CI 0.7290–0.7392) for the training cohort and 0.7336 (95% CI 0.7285–0.7387) for the validation cohort, indicating good discriminative ability. The calibration curve showed a perfect fit between the predicted prediabetes risk and the observed prediabetes risk. An analysis of the decision curve presented the clinical application of the nomogram, with alternative threshold probability spectrums being presented as well. A personalized prediabetes prediction nomogram was developed and validated among Chinese adults, identifying high-risk individuals. Doctors and others can easily and efficiently use our prediabetes prediction model when assessing prediabetes risk.

Identification of central genes for endometriosis through integration of single-cell RNA sequencing and bulk RNA sequencing analysis.

This study aimed to identify the key genes involved in the development of endometriosis and construct an accurate predictive model to provide new directions for the diagnosis and treatment of endometriosis. Using bioinformatics analysis, we employed the single-cell cell communication method to identify the key cell subtypes. By combining chip data and integrating differential analysis, WGCNA analysis, and the least absolute shrinkage and selection operator (LASSO) model, key genes were identified for immune infiltration and functional enrichment analyses. Cell communication analysis identified tissue stem cells as the key subtype. Differential analysis revealed 1879 differentially expressed genes, whereas WGCNA identified 357 module genes. The LASSO model further selects 4 key genes: Adipocyte Enhancer Binding Protein 1(AEBP1), MBNL1, GREM1, and DES. All 4 key genes showed significant correlations with immune cell content. Moreover, these genes were significantly expressed in single cells. The predictive model demonstrated good diagnostic performance. Through scRNA-seq, WGCNA, and LASSO methodologies, DES, GREM1, MBNL1, and AEBP1 emerged as crucial core genes linked to tissue stem cell markers in endometriosis. These genes have promising applications as diagnostic markers and therapeutic targets for endometriosis.

Developing non-response weights to account for attrition-related bias in a longitudinal pregnancy cohort

BackgroundProspective cohorts may be vulnerable to bias due to attrition. Inverse probability weights have been proposed as a method to help mitigate this bias. The current study used the “All Our Families” longitudinal pregnancy cohort of 3351 maternal-infant pairs and aimed to develop inverse probability weights using logistic regression models to predict study continuation versus drop-out from baseline to the three-year data collection wave.MethodsTwo methods of variable selection took place. One method was a knowledge-based a priori variable selection approach, while the second used Least Absolute Shrinkage and Selection Operator (LASSO). The ability of each model to predict continuing participation through discrimination and calibration for both approaches were evaluated by examining area under the receiver operating curve (AUROC) and calibration plots, respectively. Stabilized inverse probability weights were generated using predicted probabilities. Weight performance was assessed using standardized differences of baseline characteristics for those who continue in study and those that do not, with and without weights (unadjusted estimates).ResultsThe a priori and LASSO variable selection method prediction models had good and fair discrimination with AUROC of 0.69 (95% Confidence Interval [CI]: 0.67–0.71) and 0.73 (95% CI: 0.71–0.75), respectively. Calibration plots and non-significant Hosmer-Lemeshow Goodness of Fit Tests indicated that both the a priori (p = 0.329) and LASSO model (p = 0.242) were well-calibrated. Unweighted results indicated large (> 10%) standardized differences in 15 demographic variables (range: 11 − 29%), when comparing those who continued in the study with those that did not. Weights derived from the a priori and LASSO models reduced standardized differences relative to unadjusted estimates, with the largest differences of 13% and 5%, respectively. Additionally, when applying the same LASSO variable selection method to develop weights in future data collection waves, standardized differences remained below 10% for each demographic variable.ConclusionThe LASSO variable selection approach produced robust weights that addressed non-response bias more than the knowledge-driven approach. These weights can be applied to analyses across multiple longitudinal waves of data collection to reduce bias.

Early Plasma Levels of IL7 and IL15 Predict Response to Chimeric Antigen Receptor T Cell Therapy (CAR-T) in Large B-Cell Lymphoma (LBCL) Patients

Introduction Anti-CD19 chimeric antigen receptor T cells (CAR-T) have rapidly shifted the treatment paradigm for relapsed/refractory large B-cell lymphoma (LBCL) but produce durable complete response (CR) in only 30-40% of patients. In addition, CAR-T-mediated toxicities such as cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) remain a major concern. To investigate the relationships of circulating cytokines and chemokines on CAR-T outcomes, we applied specialized comprehensive cytokine panels to peripheral blood samples serially collected from LBCL patients undergoing treatment with anti-CD19 CAR-T cells. Methods LBCL patients were consented to our IRB-approved study and longitudinally followed for sample collection at pre-specified timepoints: prior to lymphodepleting chemotherapy (LD chemo), day 0 (day of [and prior to] CAR-T infusion), and days 3, 7, 10, 14, 21, and 90. MSD R-PLEX and U-PLEX assays were used to detect 47 cytokines and chemokines in plasma; vasoactive intestinal peptide (VIP) levels were quantified via a VIP-specific enzyme immunoassay. Mann-Whitney U test, Spearman's correlation coefficient, and a LASSO (least absolute shrinkage and selection operator) model were utilized to identify key cytokines contributing to critical clinical outcomes. Results Peri-CAR-T blood samples were collected from 24 LBCL patients, including one patient who received 2 infusions of an investigational allogeneic anti-CD19 product; all other patients received commercially available products, each associated with their distinct LD chemo dosing regimens. With median follow-up of 11.2 months, 54% patients achieved CR as best response, 50% experienced CRS (maximum Grade 2 in 25%), and 29% experienced ICANS (Grade ≥3 in 8%). Most cytokines demonstrated characteristic peak patterns over time that were not dependent on whether cytokines were produced by hematopoietic cells ( Table). Among most cytokines with peak values at Day 0, the mean percentage change from Pre-LD Chemo to Day 0 increased as chemo intensity increased (i.e., fludarabine dose from 25 to 30 mg/m 2, and cyclophosphamide dose from 250 to 500 mg/m 2). Notably, levels of the non-hematopoietic cytokines IL7 and IL15 were highest at day 0, reflecting effects of LD chemo. Using statistical models to filter relevant values, we developed a signature of 14 key cytokines that could be used to predict clinical response at 3 months. Half of these were stromal-produced cytokines from early timepoints: Cytokines positively correlated with CR or partial response (PR) included IL7 at Day 0 (Mann-Whitney U p&amp;lt;0.001) and Day 7 (p=0.044), and IL15 at Day 0 (p=0.016). Negatively correlated cytokines included IL4 at Day 3 (p=0.041), IL10 at Day 7 (p=0.027), CCL19 at Day 10 (p=0.017), and MIP1α at Day 14 (p=0.03). Intriguingly, several Type 1 T helper cell (Th1)-related cytokines with peaks at Day 3 or Day 7 (e.g., IL2, IFN-γ, TNF-α, and IL18) were not associated with response. Strong associations of timepoint-specific cytokine levels with patient outcomes are exemplified in the Figure: all patients achieving CR had IL7 levels above average on Day 0, while most patients (9/10) who failed to achieve CR had IL7 levels below average on Day 0. This same striking pattern was observed for IL15 at Day 0. Similar analyses gauged which timepoint-specific cytokines were associated with CRS and ICANS. Of 14 values associated with CRS, 13 were from Day 3 or later, likely reflecting sequelae of CRS rather than representing predictive biomarkers. Pre-LD Chemo CCL18 was the exception, positively correlated with CRS (p=0.035). Only 5 cytokine values were associated with development of ICANS, 3 from Day 10 or later; IL7 at Day 0 (p=0.011) and Day 3 (p&amp;lt;0.001) was positively correlated with ICANS . Conclusions Comprehensive cytokine profiling revealed that early measurements of several stromal-produced cytokines can predict response in LBCL patients who receive CAR-T-most notably IL7 and IL15 levels on Day 0, prior to CAR-T infusion. Conversely, few early cytokine measurements were helpful in predicting CRS and ICANS, limited to CCL18 and IL7. Analyses are ongoing to characterize associations of patient characteristics, product differences, and circulating immune cell subsets with changes in cytokine profile over time. Further studies are warranted to validate our findings in a larger data set.

Research on Carbon Intensity Prediction Method for Ships Based on Sensors and Meteorological Data

The Carbon Intensity Index (CII) exerts a substantial impact on the operations and valuation of international shipping vessels. Accurately predicting the CII of ships could help ship operators dynamically evaluate the possible CII grate of a ship at the end of the year and choose appropriate methods to improve its CII grade to meet the IMO requirement with minimum cost. This study developed and compared five CII predicting models with multiple data sources. It integrates diverse data sources, including Automatic Identification System (AIS) data, sensor data, meteorological data, and sea state data from 2022, and extracts 21 relevant features for the vessel CII prediction. Five machine learning methods, including Artificial Neural Network (ANN), Support Vector Regression (SVR), Least Absolute Shrinkage and Selection Operator (LASSO), Extreme Gradient Boosting (XGBoost), and Random Forest (RF), are employed to construct the CII prediction model, which is then applied to a 2400 TEU container ship. Features such as the mean period of total swell, mean period of wind waves, and seawater temperature were considered for inclusion as inputs in the model. The results reveal significant correlations between cumulative carbon emissions intensity and features like cumulative distance, seawater temperature, wave period, and swell period. Among these, the strongest correlations are observed with cumulative distance and seawater temperature, having correlation coefficients of 0.45 and 0.34, respectively. Notably, the ANN model demonstrates the highest accuracy in CII prediction, with an average absolute error of 0.0336, whereas the LASSO model exhibits the highest error of 0.2817. Similarly, the ANN model provides more accurate annual CII ratings for the vessel. Consequently, the ANN model proves to be the most suitable choice for cumulative CII prediction.

Numerical optimization of Lenalidomide immunomodulatory drug inside the supercritical carbon dioxide system using different machine learning models

supercritical fluids (SCFs) have been introduced as a reliable alternative for conventional organic solvents due to their unique gases-like and liquids-like behavior above the critical point. Nowadays, industrial-based employment of SCFs (especially SCCO2) for increasing the solubility of low water-soluble orally-taken medicines has increased significantly due to their noteworthy advantages like eco-friendliness, low carcinogenicity/toxicity, high technical efficiency, accessible critical points and affordable cost. The main aim is to design, optimize and modify the solubility of the Lenalidomide orally-administered immunomodulatory drug in SC-CO2 as a function of temperature (T) and pressure (P). Models are built as a novel model for using a variety of machine learning techniques, such as the Least Absolute Shrinkage and Selection Operator (LASSO), Multilayer Perceptron (MLP), and Support Vector Model (SVM). These models tuned on their hyper-parameters and then evaluated using different metrics. In terms of RMSE, LASSO, MLP, and SVM have error values equal to 9.978 × 10-2, 2.545 × 10-2, and 1.259 × 10-1, respectively. Also, the LASSO model has an R2 score of 0.9043, this score is 0.9950 for MLP and 0.9258 for SVM models. Accordingly, the MLP is the model that fits best on the solubility of Lenalidomide among these models and selected for rest of analysis in this study.

Development and validation of prediction models for poor sleep quality among older adults in the post-COVID-19 pandemic era

Background Corona Virus Disease 2019 (COVID-19) has a significant impact on sleep quality. However, the effects on sleep quality in the post-COVID-19 pandemic era remain unclear, and there is a lack of a screening tool for Chinese older adults. This study aimed to understand the prevalence of poor sleep quality and determine sensitive variables to develop an effective prediction model for screening sleep problems during infectious diseases outbreaks. Materials and Methods The Peking University Health Cohort included 10,156 participants enrolled from April to May 2023. The Pittsburgh Sleep Quality Index (PSQI) scale was used to assess sleep quality. The data were randomly divided into a training-testing cohort (n = 7109, 70%) and an independent validation cohort (n = 3027, 30%). Five prediction models with 10-fold cross validation including the Least Absolute Shrinkage and Selection Operator (LASSO), Stochastic Volatility Model (SVM), Random Forest (RF), Artificial Neural Network (ANN), and XGBoost model based on the area under curve (AUC) were used to develop and validate predictors. Results The prevalence of poor sleep quality (PSQI >7) was 30.69% (3117/10,156). Among the generated models, the LASSO model outperformed SVM (AUC 0.579), RF (AUC 0.626), ANN (AUC 0.615) and XGBoost (AUC 0.606), with an AUC of 0.7. Finally, a total of 12 variables related to sleep quality were used as parameters in the prediction models. These variables included age, gender, ethnicity, educational level, residence, marital status, history of chronic diseases, SARS-CoV-2 infection, COVID-19 vaccination, social support, depressive symptoms, and cognitive impairment among older adults during the post-COVID-19 pandemic. The nomogram illustrated that depressive symptoms contributed the most to the prediction of poor sleep quality, followed by age and residence. Conclusions This nomogram, based on twelve-variable, could potentially serve as a practical and reliable tool for early identification of poor sleep quality among older adults during the post-pandemic period.

Novel nomogram for predicting paradoxical chest wall movement in patients with flail segment of traumatic rib fracture: a retrospective cohort study

Flail chest is a severe injury to the chest wall and is related to adverse outcomes. A flail chest is classified as the physiologic, paradoxical motion of a chest wall or flail segment of rib fracture (RFX). We hypothesized that patients with paradoxical chest wall movement would present different clinical features from patients with a flail segment. This retrospective observational study included patients with blunt chest trauma who visited our level 1 trauma center between January 2019 and October 2022 and were diagnosed with one or more flail segments by computed tomography. The primary outcome of our study was a clinically diagnosed visible, paradoxical chest wall motion. We used the least absolute shrinkage and selection operator (LASSO) logistic regression model to minimize overfitting. After a feature selection using the LASSO regression model, we constructed a multivariable logistic regression (MLR) model and nomogram. A total of five risk factors were selected in the LASSO model and applied to the multivariable logistic regression model. Of these, four risk factors were statistically significant: the total number of RFX (adjusted OR [aOR], 1.28; 95% confidence interval [CI], 1.09–1.49; p = 0.002), number of segmental RFX including Grade III fractures (aOR, 1.78; 95% CI, 1.14–2.79; p = 0.012), laterally located primary fracture lines (aOR, 4.00; 95% CI, 1.69–9.43; p = 0.002), and anterior–lateral flail segments (aOR, 4.20; 95% CI, 1.60–10.99; p = 0.004). We constructed a nomogram to predict the personalized probability of the flail motion. A novel nomogram was developed in patients with flail segments of traumatic RFX to predict paradoxical chest wall motion. The number of RFX, Grade III segmental RFX, and the location of the RFX were significant risk factors.

Clinical and vessel characteristics associated with hard outcomes after PCI and their combined prognostic implications

Abstract Background While clinical outcomes after percutaneous coronary intervention (PCI) have improved in recent decades, cardiac death or myocardial infarction (MI) still occurs in patients undergoing contemporary PCI. An integrative risk model for post-PCI hard clinical endpoints incorporating clinical risk factors, angiographic and physiological vessel characteristics has not been established. Purpose To identify adverse clinical and vessel characteristics related to hard outcomes after PCI and to investigate their individual and combined prognostic implications. Methods From an individual patient data meta-analysis of 17 cohorts of patients who underwent post-PCI fractional flow reserve (FFR) measurement after drug eluting stent (DES) implantation, 2,081 patients with available 10 clinical characteristics (age, sex, clinical diagnosis at index procedure, diabetes, hypertension, dyslipidemia, smoking history, chronic kidney disease [CKD], prior myocardial infarction, left ventricular ejection fraction [LVEF]) and 8 vessel characteristics (target vessel, reference vessel diameter, % diameter stenosis, minimum lumen diameter [MLD], post-PCI % diameter stenosis, post-PCI MLD, total stent length in the target vessel, and post-PCI FFR) were analyzed. The least absolute shrinkage and selection operator (LASSO) model was applied for feature selection. The primary outcome was cardiac death or target vessel-MI at 2-years. Results The mean age of patients was 64.2±10.2 years, and the mean angiographic % diameter stenosis was 63.9±14.3%. Among 10 clinical and 8 vessel features, 4 adverse clinical characteristics (ACCs; age ≥65 years, diabetes, chronic kidney disease, and left ventricular ejection fraction &amp;lt;50%) and 2 adverse vessel characteristics (AVCs; post-PCI FFR ≤0.80 and total stent length ≥54 mm) were selected by the LASSO model. All ACCs and AVCs were independent predictors for the primary outcome (HR 3.33 [95% CI 1.34 – 8.25], P =0.009 for age ≥65 years; HR 2.62 [1.22 – 5.65], P =0.014 for diabetes; HR 5.95 [2.52 – 14.1], P &amp;lt;0.001 for CKD; HR 2.91 [1.17 – 7.20], P =0.021 for LVEF &amp;lt;50%; HR 3.79 [1.73 – 8.28], P &amp;lt;0.001 for post-PCI FFR ≤0.80; HR 2.42 [1.08 – 5.39], P =0.031 for total stent length ≥54 mm). The number of AVCs had additive predictability for the primary endpoint to that of ACCs (AUC 0.72 vs 0.78, P =0.03), and vice versa (AUC 0.68 vs 0.78, P =0.03). The cumulative event rate increased in the order of none, either, and both of ACCs ≥2 and AVCs ≥1 (0.3%, 2.4%, and 5.3%, P-for-trend &amp;lt;0.01) (Figure). Conclusions In patients undergoing DES implantation, adverse clinical and vessel characteristics were associated with the risk of cardiac death or target vessel-MI. Since these characteristics showed independent and additive prognostic value, their integrative assessment can optimize post-PCI risk stratification.

Temporal Generalizability of Machine Learning Models for Predicting Postoperative Delirium Using Electronic Health Record Data: Model Development and Validation Study.

Although machine learning models demonstrate significant potential in predicting postoperative delirium, the advantages of their implementation in real-world settings remain unclear and require a comparison with conventional models in practical applications. The objective of this study was to validate the temporal generalizability of decision tree ensemble and sparse linear regression models for predicting delirium after surgery compared with that of the traditional logistic regression model. The health record data of patients hospitalized at an advanced emergency and critical care medical center in Kumamoto, Japan, were collected electronically. We developed a decision tree ensemble model using extreme gradient boosting (XGBoost) and a sparse linear regression model using least absolute shrinkage and selection operator (LASSO) regression. To evaluate the predictive performance of the model, we used the area under the receiver operating characteristic curve (AUROC) and the Matthews correlation coefficient (MCC) to measure discrimination and the slope and intercept of the regression between predicted and observed probabilities to measure calibration. The Brier score was evaluated as an overall performance metric. We included 11,863 consecutive patients who underwent surgery with general anesthesia between December 2017 and February 2022. The patients were divided into a derivation cohort before the COVID-19 pandemic and a validation cohort during the COVID-19 pandemic. Postoperative delirium was diagnosed according to the confusion assessment method. A total of 6497 patients (68.5, SD 14.4 years, women n=2627, 40.4%) were included in the derivation cohort, and 5366 patients (67.8, SD 14.6 years, women n=2105, 39.2%) were included in the validation cohort. Regarding discrimination, the XGBoost model (AUROC 0.87-0.90 and MCC 0.34-0.44) did not significantly outperform the LASSO model (AUROC 0.86-0.89 and MCC 0.34-0.41). The logistic regression model (AUROC 0.84-0.88, MCC 0.33-0.40, slope 1.01-1.19, intercept -0.16 to 0.06, and Brier score 0.06-0.07), with 8 predictors (age, intensive care unit, neurosurgery, emergency admission, anesthesia time, BMI, blood loss during surgery, and use of an ambulance) achieved good predictive performance. The XGBoost model did not significantly outperform the LASSO model in predicting postoperative delirium. Furthermore, a parsimonious logistic model with a few important predictors achieved comparable performance to machine learning models in predicting postoperative delirium.