Development and validation of novel interpretable survival prediction models based on drug exposures for severe heart failure during vulnerable period.

BACKGROUND: Severe heart failure (HF) has a higher mortality during vulnerable period while targeted predictive tools, especially based on drug exposures, to accurately assess its prognoses remain largely unexplored. Therefore, this study aimed to utilize drug information as the main predictor to develop and validate survival models for severe HF patients during this period. METHODS: We extracted severe HF patients from the MIMIC-IV database (as training and internal validation cohorts) as well as from the MIMIC-III database and local hospital (as external validation cohorts). Three algorithms, including Cox proportional hazards model (CoxPH), random survival forest (RSF), and deep learning survival prediction (DeepSurv), were applied to incorporate the parameters (partial hospitalization information and exposure durations of drugs) for constructing survival prediction models. The model performance was assessed mainly using area under the receiver operator characteristic curve (AUC), brier score (BS), and decision curve analysis (DCA). The model interpretability was determined by the permutation importance and Shapley additive explanations values. RESULTS: A total of 11,590 patients were included in this study. Among the 3 models, the CoxPH model ultimately included 10 variables, while RSF and DeepSurv models incorporated 24 variables, respectively. All of the 3 models achieved respectable performance metrics while the DeepSurv model exhibited the highest AUC values and relatively lower BS among these models. The DCA also verified that the DeepSurv model had the best clinical practicality. CONCLUSIONS: The survival prediction tools established in this study can be applied to severe HF patients during vulnerable period by mainly inputting drug treatment duration, thus contributing to optimal clinical decisions prospectively.

A prediction model of elderly hip fracture mortality including preoperative red cell distribution width constructed based on the random survival forest (RSF) and Cox risk ratio regression.

An independent correlation between pre-RDW and 1-year mortality after surgery in elderly hip fracture can be used to predict mortality in elderly hip fracture patients and has predictive significance in anemia patients. With further research, a treatment algorithm can be developed to potentially identify patients at high risk of preoperative mortality. INTRODUCTION: Red blood cell distribution width (RDW) is an independent predictor of various disease states in elderly individuals, but its association with the prognosis of elderly hip fracture patients is controversial. This study aimed to evaluate the prognostic value of RDW in such patients, construct a prediction model containing RDW using random survival forest (RSF) and Cox regression analysis, and compare RDW in patients with and without anemia. METHODS: We retrospectively analyzed the data of elderly patients who underwent hip fracture surgery, selected the best variables using RSF, stratified the independent variables by Cox regression analysis, constructed a 1-year mortality prediction model of elderly hip fracture with RDW, and conducted internal validation and external validation. RESULTS: Two thousand one hundred six patients were included in this study. The RSF algorithm selects 12 important influencing factors, and Cox regression analysis showed that eight variables including preoperative RDW (pre-RDW) were independent risk factors for death within 1-year after hip fracture surgery in elderly patients. Stratified analysis showed that pre-RDW was still independently associated with 1-year mortality in the non-anemia group and not in the anemia group. The nomogram prediction model had high differentiation and fit, and the prediction model constructed by the total cohort of patients was also used for validation of patients in the anemia patients and obtained good clinical benefits. CONCLUSION: An independent correlation between pre-RDW and 1-year mortality after surgery in elderly hip fracture can be used to predict mortality in elderly hip fracture patients and has predictive significance in anemia patients.

Development and Validation of a Novel Machine Learning Model to Predict the Survival of Patients with Gastrointestinal Neuroendocrine Neoplasms.

INTRODUCTION: Well-calibrated models for personalized prognostication of patients with gastrointestinal neuroendocrine neoplasms (GINENs) are limited. This study aimed to develop and validate a machine-learning model to predict the survival of patients with GINENs. METHODS: Oblique random survival forest (ORSF) model, Cox proportional hazard risk model, Cox model with least absolute shrinkage and selection operator penalization, CoxBoost, Survival Gradient Boosting Machine, Extreme Gradient Boosting survival regression, DeepHit, DeepSurv, DNNSurv, logistic-hazard model, and PC-hazard model were compared. We further tuned hyperparameters and selected variables for the best-performing ORSF. Then, the final ORSF model was validated. RESULTS: A total of 43,444 patients with GINENs were included. The median (interquartile range) survival time was 53 (19-102) months. The ORSF model performed best, in which age, histology, M stage, tumor size, primary tumor site, sex, tumor number, surgery, lymph nodes removed, N stage, race, and grade were ranked as important variables. However, chemotherapy and radiotherapy were not necessary for the ORSF model. The ORSF model had an overall C index of 0.86 (95% confidence interval, 0.85-0.87). The area under the receiver operation curves at 1, 3, 5, and 10 years were 0.91, 0.89, 0.87, and 0.80, respectively. The decision curve analysis showed superior clinical usefulness of the ORSF model than the American Joint Committee on Cancer Stage. A nomogram and an online tool were given. CONCLUSION: The machine learning ORSF model could precisely predict the survival of patients with GINENs, with the ability to identify patients at high risk for death and probably guide clinical practice.

Model-agnostic explanations for survival prediction models.

Advanced machine learning methods capable of capturing complex and nonlinear relationships can be used in biomedical research to accurately predict time-to-event outcomes. However, these methods have been criticized as "black boxes" that are not interpretable and thus are difficult to trust in making important clinical decisions. Explainable machine learning proposes the use of model-agnostic explainers that can be applied to predictions from any complex model. These explainers describe how a patient's characteristics are contributing to their prediction, and thus provide insight into how the model is arriving at that prediction. The specific application of these explainers to survival prediction models can be used to obtain explanations for (i) survival predictions at particular follow-up times, and (ii) a patient's overall predicted survival curve. Here, we present a model-agnostic approach for obtaining these explanations from any survival prediction model. We extend the local interpretable model-agnostic explainer framework for classification outcomes to survival prediction models. Using simulated data, we assess the performance of the proposed approaches under various settings. We illustrate application of the new methodology using prostate cancer data.

Developing a Prognostic Model for Primary Biliary Cholangitis Based on a Random Survival Forest Model.

Background: Primary biliary cholangitis (PBC) is a rare autoimmune liver disease with few effective treatments and a poor prognosis, and its incidence is on the rise. There is an urgent need for more targeted treatment strategies to accurately identify high-risk patients. The use of stochastic survival forest models in machine learning is an innovative approach to constructing a prognostic model for PBC that can improve the prognosis by identifying high-risk patients for targeted treatment. Method: Based on the inclusion and exclusion criteria, the clinical data and follow-up data of patients diagnosed with PBC-associated cirrhosis between January 2011 and December 2021 at Taizhou Hospital of Zhejiang Province were retrospectively collected and analyzed. Data analyses and random survival forest model construction were based on the R language. Result: Through a Cox univariate regression analysis of 90 included samples and 46 variables, 17 variables with p-values <0.1 were selected for initial model construction. The out-of-bag (OOB) performance error was 0.2094, and K-fold cross-validation yielded an internal validation C-index of 0.8182. Through model selection, cholinesterase, bile acid, the white blood cell count, total bilirubin, and albumin were chosen for the final predictive model, with a final OOB performance error of 0.2002 and C-index of 0.7805. Using the final model, patients were stratified into high- and low-risk groups, which showed significant differences with a P value <0.0001. The area under the curve was used to evaluate the predictive ability for patients in the first, third, and fifth years, with respective results of 0.9595, 0.8898, and 0.9088. Conclusion: The present study constructed a prognostic model for PBC-associated cirrhosis patients using a random survival forest model, which accurately stratified patients into low- and high-risk groups. Treatment strategies can thus be more targeted, leading to improved outcomes for high-risk patients.

Risk factors for the time to development of retinopathy of prematurity in premature infants in Iran: a machine learning approach.

BACKGROUND: Retinopathy of prematurity (ROP), is a preventable leading cause of blindness in infants and is a condition in which the immature retina experiences abnormal blood vessel growth. The development of ROP is multifactorial; nevertheless, the risk factors are controversial. This study aimed to identify risk factors of time to development of ROP in Iran. METHODS: This historical cohort study utilized data from the hospital records of all newborns referred to the ROP department of Farabi Hospital (from 2017 to 2021) and the NICU records of infants referred from Mahdieh Hospital to Farabi Hospital. Preterm infants with birth weight (BW) ≤ 2000 g or gestational age (GA) < 34 wk, as well as selected infants with an unstable clinical course, as determined by their pediatricians or neonatologists, with BW > 2000 g or GA ≥ 34 wk. The outcome variable was the time to development of ROP (in weeks). Random survival forest was used to analyze the data. RESULTS: A total of 338 cases, including 676 eyes, were evaluated. The mean GA and BW of the study group were 31.59 ± 2.39 weeks and 1656.72 ± 453.80 g, respectively. According to the criteria of minimal depth and variable importance, the most significant predictors of the time to development of ROP were duration of ventilation, GA, duration of oxygen supplementation, bilirubin levels, duration of antibiotic administration, duration of Total Parenteral Nutrition (TPN), mother age, birth order, number of surfactant administration, and on time screening. The concordance index for predicting survival of the fitted model was 0.878. CONCLUSION: Our findings indicated that the duration of ventilation, GA, duration of oxygen supplementation, bilirubin levels, duration of antibiotic administration, duration of TPN, mother age, birth order, number of surfactant administrations, and on time screening are potential risk factors of prognosis of ROP. The associations between identified risk factors were mostly nonlinear. Therefore, it is recommended to consider the nature of these relationships in managing treatment and designing early interventions.

Development of machine learning models for patients in the high intrahepatic cholangiocarcinoma incidence age group.

BACKGROUND: Intrahepatic cholangiocarcinoma (ICC) has a poor prognosis and is understudied. Based on the clinical features of patients with ICC, we constructed machine learning models to understand their importance on survival and to accurately determine patient prognosis, aiming to develop reference values to guide physicians in developing more effective treatment plans. METHODS: This study used machine learning (ML) algorithms to build prediction models using ICC data on 1,751 patients from the SEER (Surveillance, Epidemiology, and End Results) database and 58 hospital cases. The models' performances were compared using receiver operating characteristic curve analysis, C-index, and Brier scores. RESULTS: A total of eight variables were used to construct the ML models. Our analysis identified the random survival forest model as the best for prognostic prediction. In the training cohort, its C-index, Brier score, and Area Under the Curve values were 0.76, 0.124, and 0.882, respectively, and it also performed well in the test cohort. Kaplan-Meier survival analysis revealed that the model could effectively determine patient prognosis. CONCLUSIONS: To our knowledge, this is the first study to develop ML prognostic models for ICC in the high-incidence age group. Of the ML models, the random survival forest model was best at prognosis prediction.

Development and validation of a prospective study to predict the risk of readmission within 365 days of respiratory failure: based on a random survival forest algorithm combined with COX regression modeling.

BACKGROUND: There is a need to develop and validate a widely applicable nomogram for predicting readmission of respiratory failure patients within 365 days. METHODS: We recruited patients with respiratory failure at the First People's Hospital of Yancheng and the People's Hospital of Jiangsu. We used the least absolute shrinkage and selection operator regression to select significant features for multivariate Cox proportional hazard analysis. The Random Survival Forest algorithm was employed to construct a model for the variables that obtained a coefficient of 0 following LASSO regression, and subsequently determine the prediction score. Independent risk factors and the score were used to develop a multivariate COX regression for creating the line graph. We used the Harrell concordance index to quantify the predictive accuracy and the receiver operating characteristic curve to evaluate model performance. Additionally, we used decision curve analysiso assess clinical usefulness. RESULTS: The LASSO regression and multivariate Cox regression were used to screen hemoglobin, diabetes and pneumonia as risk variables combined with Score to develop a column chart model. The C index is 0.927 in the development queue, 0.924 in the internal validation queue, and 0.922 in the external validation queue. At the same time, the predictive model also showed excellent calibration and higher clinical value. CONCLUSIONS: A nomogram predicting readmission of patients with respiratory failure within 365 days based on three independent risk factors and a jointly developed random survival forest algorithm has been developed and validated. This improves the accuracy of predicting patient readmission and provides practical information for individualized treatment decisions.

Improved nonparametric survival prediction using CoxPH, Random Survival Forest & DeepHit Neural Network.

In recent times, time-to-event data such as time to failure or death is routinely collected alongside high-throughput covariates. These high-dimensional bioinformatics data often challenge classical survival models, which are either infeasible to fit or produce low prediction accuracy due to overfitting. To address this issue, the focus has shifted towards introducing a novel approaches for feature selection and survival prediction. In this article, we propose a new hybrid feature selection approach that handles high-dimensional bioinformatics datasets for improved survival prediction. This study explores the efficacy of four distinct variable selection techniques: LASSO, RSF-vs, SCAD, and CoxBoost, in the context of non-parametric biomedical survival prediction. Leveraging these methods, we conducted comprehensive variable selection processes. Subsequently, survival analysis models-specifically CoxPH, RSF, and DeepHit NN-were employed to construct predictive models based on the selected variables. Furthermore, we introduce a novel approach wherein only variables consistently selected by a majority of the aforementioned feature selection techniques are considered. This innovative strategy, referred to as the proposed method, aims to enhance the reliability and robustness of variable selection, subsequently improving the predictive performance of the survival analysis models. To evaluate the effectiveness of the proposed method, we compare the performance of the proposed approach with the existing LASSO, RSF-vs, SCAD, and CoxBoost techniques using various performance metrics including integrated brier score (IBS), concordance index (C-Index) and integrated absolute error (IAE) for numerous high-dimensional survival datasets. The real data applications reveal that the proposed method outperforms the competing methods in terms of survival prediction accuracy.

Random Survival Forests With Competing Events: A Subdistribution-Based Imputation Approach.

Random survival forests (RSF) can be applied to many time-to-event research questions and are particularly useful in situations where the relationship between the independent variables and the event of interest is rather complex. However, in many clinical settings, the occurrence of the event of interest is affected by competing events, which means that a patient can experience an outcome other than the event of interest. Neglecting the competing event (i.e., regarding competing events as censoring) will typically result in biased estimates of the cumulative incidence function (CIF). A popular approach for competing events is Fine and Gray's subdistribution hazard model, which directly estimates the CIF by fitting a single-event model defined on a subdistribution timescale. Here, we integrate concepts from the subdistribution hazard modeling approach into the RSF. We develop several imputation strategies that use weights as in a discrete-time subdistribution hazard model to impute censoring times in cases where a competing event is observed. Our simulations show that the CIF is well estimated if the imputation already takes place outside the forest on the overall dataset. Especially in settings with a low rate of the event of interest or a high censoring rate, competing events must not be neglected, that is, treated as censoring. When applied to a real-world epidemiological dataset on chronic kidney disease, the imputation approach resulted in highly plausible predictor-response relationships and CIF estimates of renal events.

A risk-based postresection follow-up strategy for hepatocellular carcinoma patients.

BACKGROUND: The optimal intervals for follow-up after hepatocellular carcinoma (HCC) patients undergo curative liver resection (LR) remain unclear. This study aimed to establish a risk-based post-resection follow-up strategy. METHODS: Patients that were diagnosed with HCC and received LR from three hospitals in China were included. The risk-based strategy was established based on the random survival forest model and compared with a fixed strategy both internally and externally. RESULTS: In total, 3447 patients from three hospitals were included. The authors' strategy showed superiority in the early detection of tumor relapse compared with fixed surveillance. Under fewer total visits, risk-based strategy achieved analogous survival time compared to the total 20 times follow-ups based on fixed strategy. Twelve total visits (five, three, one, two, and one visits in years 1-5, respectively) for American Joint Committee on Cancer/International Union Against Cancer T1a stage patients, 13 total visits (five, four, one, two, and one visits in years 1-5, respectively) for T1b stage patients, 15 total visits (eight, three, three, zero, and one visits in years 1-5, respectively) for T2 stage patients, and 15 total visits (eight, four, one, one, and one visits in years 1-5, respectively) for T3 stage patients were advocated. The detailed follow-up arrangements were available to the public through an interactive website (https://sysuccfyz.shinyapps.io/RiskBasedFollowUp/). CONCLUSION: This risk-based surveillance strategy was demonstrated to detect relapse earlier and reduce the total number of follow-ups without compromising on survival. Based on the strategy and methodology of the authors, surgeons or patients could choose more intensive or flexible schedules depending on the requirements and economic conditions. PLAIN LANGUAGE SUMMARY: A risk-based post-resection follow-up strategy was established by random survival forest model using a larger hepatocellular carcinoma population The strategy was demonstrated to detect tumor relapse earlier and reduce the total number of follow-ups without compromising on survival Our strategy and methodology could be widely applied by other surgeons and patients.

PCLasso: a protein complex-based, group lasso-Cox model for accurate prognosis and risk protein complex discovery.

For high-dimensional expression data, most prognostic models perform feature selection based on individual genes, which usually lead to unstable prognosis, and the identified risk genes are inherently insufficient in revealing complex molecular mechanisms. Since most genes carry out cellular functions by forming protein complexes-basic representatives of functional modules, identifying risk protein complexes may greatly improve our understanding of disease biology. Coupled with the fact that protein complexes have been shown to have innate resistance to batch effects and are effective predictors of disease phenotypes, constructing prognostic models and selecting features with protein complexes as the basic unit should improve the robustness and biological interpretability of the model. Here, we propose a protein complex-based, group lasso-Cox model (PCLasso) to predict patient prognosis and identify risk protein complexes. Experiments on three cancer types have proved that PCLasso has better prognostic performance than prognostic models based on individual genes. The resulting risk protein complexes not only contain individual risk genes but also incorporate close partners that synergize with them, which may promote the revealing of molecular mechanisms related to cancer progression from a comprehensive perspective. Furthermore, a pan-cancer prognostic analysis was performed to identify risk protein complexes of 19 cancer types, which may provide novel potential targets for cancer research.

Prediabetes as a risk factor for new-onset atrial fibrillation: the propensity-score matching cohort analyzed using the Cox regression model coupled with the random survival forest.

BACKGROUND: The glycemic continuum often indicates a gradual decline in insulin sensitivity leading to an increase in glucose levels. Although prediabetes is an established risk factor for both macrovascular and microvascular diseases, whether prediabetes is independently associated with the risk of developing atrial fibrillation (AF), particularly the occurrence time, has not been well studied using a high-quality research design in combination with statistical machine-learning algorithms. METHODS: Using data available from electronic medical records collected from the National Taiwan University Hospital, a tertiary medical center in Taiwan, we conducted a retrospective cohort study consisting 174,835 adult patients between 2014 and 2019 to investigate the relationship between prediabetes and AF. To render patients with prediabetes as comparable to those with normal glucose test, a propensity-score matching design was used to select the matched pairs of two groups with a 1:1 ratio. The Kaplan-Meier method was used to compare the cumulative risk of AF between prediabetes and normal glucose test using log-rank test. The multivariable Cox regression model was employed to estimate adjusted hazard ratio (HR) for prediabetes versus normal glucose test by stratifying three levels of glycosylated hemoglobin (HbA1c). The machine-learning algorithm using the random survival forest (RSF) method was further used to identify the importance of clinical factors associated with AF in patients with prediabetes. RESULTS: A sample of 14,309 pairs of patients with prediabetes and normal glucose test result were selected. The incidence of AF was 11.6 cases per 1000 person-years during a median follow-up period of 47.1 months. The Kaplan-Meier analysis revealed that the risk of AF was significantly higher in patients with prediabetes (log-rank p < 0.001). The multivariable Cox regression model indicated that prediabetes was independently associated with a significant increased risk of AF (HR 1.24, 95% confidence interval 1.11-1.39, p < 0.001), particularly for patients with HbA1c above 5.5%. The RSF method identified elevated N-terminal natriuretic peptide and altered left heart structure as the two most important risk factors for AF among patients with prediabetes. CONCLUSIONS: Our study found that prediabetes is independently associated with a higher risk of AF. Furthermore, alterations in left heart structure make a significant contribution to this elevated risk, and these structural changes may begin during the prediabetes stage.

Plasma lipidic fingerprint associated with type 2 diabetes in patients with coronary heart disease: CORDIOPREV study.

OBJECTIVE: We aimed to identify a lipidic profile associated with type 2 diabetes mellitus (T2DM) development in coronary heart disease (CHD) patients, to provide a new, highly sensitive model which could be used in clinical practice to identify patients at T2DM risk. METHODS: This study considered the 462 patients of the CORDIOPREV study (CHD patients) who were not diabetic at the beginning of the intervention. In total, 107 of them developed T2DM after a median follow-up of 60 months. They were diagnosed using the American Diabetes Association criteria. A novel lipidomic methodology employing liquid chromatography (LC) separation followed by HESI, and detection by mass spectrometry (MS) was used to annotate the lipids at the isomer level. The patients were then classified into a Training and a Validation Set (60-40). Next, a Random Survival Forest (RSF) was carried out to detect the lipidic isomers with the lowest prediction error, these lipids were then used to build a Lipidomic Risk (LR) score which was evaluated through a Cox. Finally, a production model combining the clinical variables of interest, and the lipidic species was carried out. RESULTS: LC-tandem MS annotated 440 lipid species. From those, the RSF identified 15 lipid species with the lowest prediction error. These lipids were combined in an LR score which showed association with the development of T2DM. The LR hazard ratio per unit standard deviation was 2.87 and 1.43, in the Training and Validation Set respectively. Likewise, patients with higher LR Score values had lower insulin sensitivity (P = 0.006) and higher liver insulin resistance (P = 0.005). The receiver operating characteristic (ROC) curve obtained by combining clinical variables and the selected lipidic isomers using a generalised lineal model had an area under the curve (AUC) of 81.3%. CONCLUSION: Our study showed the potential of comprehensive lipidomic analysis in identifying patients at risk of developing T2DM. In addition, the lipid species combined with clinical variables provided a new, highly sensitive model which can be used in clinical practice to identify patients at T2DM risk. Moreover, these results also indicate that we need to look closely at isomers to understand the role of this specific compound in T2DM development. Trials registration NCT00924937.

Prognostic models for upper urinary tract urothelial carcinoma patients after radical nephroureterectomy based on a novel systemic immune-inflammation score with machine learning.

PURPOSE: This study aimed to evaluate the clinical significance of a novel systemic immune-inflammation score (SIIS) to predict oncological outcomes in upper urinary tract urothelial carcinoma(UTUC) after radical nephroureterectomy(RNU). METHOD: The clinical data of 483 patients with nonmetastatic UTUC underwent surgery in our center were analyzed. Five inflammation-related biomarkers were screened in the Lasso-Cox model and then aggregated to generate the SIIS based on the regression coefficients. Overall survival (OS) was assessed using Kaplan-Meier analyses. The Cox proportional hazards regression and random survival forest model were adopted to build the prognostic model. Then we established an effective nomogram for UTUC after RNU based on SIIS. The discrimination and calibration of the nomogram were evaluated using the concordance index (C-index), area under the time-dependent receiver operating characteristic curve (time-dependent AUC), and calibration curves. Decision curve analysis (DCA) was used to assess the net benefits of the nomogram at different threshold probabilities. RESULT: According to the median value SIIS computed by the lasso Cox model, the high-risk group had worse OS (p<0.0001) than low risk-group. Variables with a minimum depth greater than the depth threshold or negative variable importance were excluded, and the remaining six variables were included in the model. The area under the ROC curve (AUROC) of the Cox and random survival forest models were 0.801 and 0.872 for OS at five years, respectively. Multivariate Cox analysis showed that elevated SIIS was significantly associated with poorer OS (p<0.001). In terms of predicting overall survival, a nomogram that considered the SIIS and clinical prognostic factors performed better than the AJCC staging. CONCLUSION: The pretreatment levels of SIIS were an independent predictor of prognosis in upper urinary tract urothelial carcinoma after RNU. Therefore, incorporating SIIS into currently available clinical parameters helps predict the long-term survival of UTUC.

Prognostic value of preoperative hematological markers in patients with glioblastoma multiforme and construction of random survival forest model.

OBJECTIVE: In recent years, an increasing number of studies have revealed that patients' preoperative inflammatory response, coagulation function, and nutritional status are all linked to the occurrence, development, angiogenesis, and metastasis of various malignant tumors. The goal of this study is to determine the relationship between preoperative peripheral blood neutrophil to lymphocyte ratio (NLR), monocyte to lymphocyte ratio (MLR), systemic immune-inflammatory index (SII), platelet to lymphocyte ratio (PLR), and platelet to fibrinogen ratio (FPR). Prognostic nutritional index (PNI) and the prognosis of glioblastoma multiforme (GBM) patients, as well as establish a forest prediction model that includes preoperative hematological markers to predict the individual GBM patient's 3-year survival status after treatment. METHODS: The clinical and hematological data of 281 GBM patients were analyzed retrospectively; overall survival (OS) was the primary endpoint. X-Tile software was used to determine the best cut-off values for NLR, SII, and PLR, and the survival analysis was carried out by the Kaplan-Meier method as well as univariate and multivariate COX regression. Afterward, we created a random forest model that predicts the individual GBM patient's 3-year survival status after treatment, and the area under the curve (AUC) is used to validate the model's effectiveness. RESULTS: The best cut-off values for NLR, SII, and PLR in GBM patients' preoperative peripheral blood were 2.12, 537.50, and 93.5 respectively. The Kaplan-Meier method revealed that preoperative GBM patients with high SII, high NLR, and high PLR had shorter overall survival, and the difference was statistically significant. In addition to clinical and pathological factors. Univariate Cox showed NLR (HR = 1.456, 95% CI: 1.286 ~ 1.649, P < 0.001) MLR (HR = 1.272, 95% CI: 1.120 ~ 1.649, P < 0.001), FPR (HR = 1.183,95% CI: 1.049 ~ 1.333, P < 0.001), SII (HR = 0.218,95% CI: 1.645 ~ 2.127, P < 0.001) is related to the prognosis and overall survival of GBM. Multivariate Cox proportional hazard regression showed that SII (HR = 1.641, 95% CI: 1.430 ~ 1.884, P < 0.001) is also related to the overall survival of patients with GBM. In the random forest prognostic model with preoperative hematologic markers, the AUC in the test set and the validation set was 0.907 and 0.900, respectively. CONCLUSION: High levels of NLR, MLR, PLR, FPR, and SII before surgery are prognostic risk factors for GBM patients. A high preoperative SII level is an independent risk factor for GBM prognosis. The random forest model that includes preoperative hematological markers has the potential to predict the individual GBM patient's 3-year survival status after treatment,and assist the clinicians for making a good clinical decision.

Machine learning versus regression for prediction of sporadic pancreatic cancer.

BACKGROUND/OBJECTIVES: There is currently no widely accepted approach to identify patients at increased risk for sporadic pancreatic cancer (PC). We aimed to compare the performance of two machine-learning models with a regression-based model in predicting pancreatic ductal adenocarcinoma (PDAC), the most common form of PC. METHODS: This retrospective cohort study consisted of patients 50-84 years of age enrolled in either Kaiser Permanente Southern California (KPSC, model training, internal validation) or the Veterans Affairs (VA, external testing) between 2008 and 2017. The performance of random survival forests (RSF) and eXtreme gradient boosting (XGB) models were compared to that of COX proportional hazards regression (COX). Heterogeneity of the three models were assessed. RESULTS: The KPSC and the VA cohorts consisted of 1.8 and 2.7 million patients with 1792 and 4582 incident PDAC cases within 18 months, respectively. Predictors selected into all three models included age, abdominal pain, weight change, and glycated hemoglobin (A1c). Additionally, RSF selected change in alanine transaminase (ALT), whereas the XGB and COX selected the rate of change in ALT. The COX model appeared to have lower AUC (KPSC: 0.737, 95% CI 0.710-0.764; VA: 0.706, 0.699-0.714), compared to those of RSF (KPSC: 0.767, 0.744-0.791; VA: 0.731, 0.724-0.739) and XGB (KPSC: 0.779, 0.755-0.802; VA: 0.742, 0.735-0.750). Among patients with top 5% predicted risk from all three models (N = 29,663), 117 developed PDAC, of which RSF, XGB and COX captured 84 (9 unique), 87 (4 unique), 87 (19 unique) cases, respectively. CONCLUSIONS: The three models complement each other, but each has unique contributions.

Application of machine learning in predicting survival outcomes involving real-world data: a scoping review.

BACKGROUND: Despite the interest in machine learning (ML) algorithms for analyzing real-world data (RWD) in healthcare, the use of ML in predicting time-to-event data, a common scenario in clinical practice, is less explored. ML models are capable of algorithmically learning from large, complex datasets and can offer advantages in predicting time-to-event data. We reviewed the recent applications of ML for survival analysis using RWD in healthcare. METHODS: PUBMED and EMBASE were searched from database inception through March 2023 to identify peer-reviewed English-language studies of ML models for predicting time-to-event outcomes using the RWD. Two reviewers extracted information on the data source, patient population, survival outcome, ML algorithms, and the Area Under the Curve (AUC). RESULTS: Of 257 citations, 28 publications were included. Random survival forests (N = 16, 57%) and neural networks (N = 11, 39%) were the most popular ML algorithms. There was variability across AUC for these ML models (median 0.789, range 0.6-0.950). ML algorithms were predominately considered for predicting overall survival in oncology (N = 12, 43%). ML survival models were often used to predict disease prognosis or clinical events (N = 27, 96%) in the oncology, while less were used for treatment outcomes (N = 1, 4%). CONCLUSIONS: The ML algorithms, random survival forests and neural networks, are mainly used for RWD to predict survival outcomes such as disease prognosis or clinical events in the oncology. This review shows that more opportunities remain to apply these ML algorithms to inform treatment decision-making in clinical practice. More methodological work is also needed to ensure the utility and applicability of ML models in survival outcomes.

A comparison of random survival forest and Cox regression for prediction of mortality in patients with hemorrhagic stroke.

OBJECTIVE: To evaluate RSF and Cox models for mortality prediction of hemorrhagic stroke (HS) patients in intensive care unit (ICU). METHODS: In the training set, the optimal models were selected using five-fold cross-validation and grid search method. In the test set, the bootstrap method was used to validate. The area under the curve(AUC) was used for discrimination, Brier Score (BS) was used for calibration, positive predictive value(PPV), negative predictive value(NPV), and F1 score were combined to compare. RESULTS: A total of 2,990 HS patients were included. For predicting the 7-day mortality, the mean AUCs for RSF and Cox regression were 0.875 and 0.761, while the mean BS were 0.083 and 0.108. For predicting the 28-day mortality, the mean AUCs for RSF and Cox regression were 0.794 and 0.649, while the mean BS were 0.129 and 0.174. The mean AUCs of RSF and Cox versus conventional scores for predicting patients' 7-day mortality were 0.875 (RSF), 0.761 (COX), 0.736 (SAPS II), 0.723 (OASIS), 0.632 (SIRS), and 0.596 (SOFA), respectively. CONCLUSIONS: RSF provided a better clinical reference than Cox. Creatine, temperature, anion gap and sodium were important variables in both models.

Development and validation of a model for the prediction of disease-specific survival in patients with oral squamous cell carcinoma: based on random survival forest analysis.

OBJECTIVE: To establish a model for predicting the disease-specific survival (DSS) of patients with oral squamous cell carcinoma (OSCC). METHODS: Patients diagnosed with OSCC from the Surveillance, Epidemiology, and End Results (SEER) database were enrolled and randomly divided into development (n = 14,495) and internal validation cohort (n = 9625). Additionally, a cohort from a hospital located in Southeastern China was utilized for external validation (n = 582). RESULTS: TNM stage, adjuvant treatment, surgery, tumor sites, age, grade, and gender were used for RSF model construction based on the development cohort. The effectiveness of the model was confirmed through time-dependent ROC curves in different cohorts. The risk score exhibited an almost exponential increase in the hazard ratio of death due to OSCC. In development, internal, and external validation cohorts, the prognosis was significantly worse for patients in groups with higher risk scores (all log-rank P < 0.05). CONCLUSION: Based on RSF, a high-performance prediction model for OSCC prognosis was created and verified in this study.