Multitask Logistic Regression Research Articles

Neutral Benchmarking of Survival Models in Health Sciences: Comparative Study of Classical and Machine Learning Techniques

Survival analysis plays a central role in diverse research fields, especially in health sciences. As an analytical tool, it can be used to help improve patients’ survival time, or at least, reduce the prospects of recurrence in cancer studies. However, approaches to the predictive performance of the current survival models mainly center on clinical data along with the classical survival methods. For censored “omics” data, the performance of survival models has not been thoroughly studied, either often due to their high dimensionality issues or reliance on binarizing the survival time for classification analysis. We aim to present a neural benchmark approach that analyzes and compares a broad range of classical and state-of-the-art machine learning survival models for “omics” and clinical datasets. All the methods considered in our study are evaluated using predictability as a performance measure. The study is systematically designed to make 36 comparisons (9 methods over 4 datasets, i.e., 2 clinical and 2 omics), and shows that, in practice, predictability of survival models does vary across real-world datasets, model choice, as well as the evaluation metric. From our results, we emphasize that performance criteria can play a key role in a balanced assessment of diverse survival models. Moreover, the Multitask Logistic Regression (MTLR) showed remarkable predictability for almost all the datasets. We believe this outstanding performance presents a unique opportunity for a wider use of MTLR for survival risk factors. For translational clinicians and scientists, we hope our findings provide practical guidance for benchmark studies of survival models, as well as highlight potential areas of research interest.

Machine learning-based individualized survival prediction model for prognosis in osteosarcoma: Data from the SEER database.

Patient outcomes of osteosarcoma vary because of tumor heterogeneity and treatment strategies. This study aimed to compare the performance of multiple machine learning (ML) models with the traditional Cox proportional hazards (CoxPH) model in predicting prognosis and explored the potential of ML models in clinical decision-making. From 2000 to 2018, 1243 patients with osteosarcoma were collected from the Surveillance, Epidemiology, and End Results (SEER) database. Three ML methods were chosen for model development (DeepSurv, neural multi-task logistic regression [NMTLR]) and random survival forest [RSF]) and compared them with the traditional CoxPH model and TNM staging systems. 871 samples were used for model training, and the rest were used for model validation. The models' overall performance and predictive accuracy for 3- and 5-year survival were assessed by several metrics, including the concordance index (C-index), the Integrated Brier Score (IBS), receiver operating characteristic curves (ROC), area under the ROC curves (AUC), calibration curves, and decision curve analysis. The efficacy of personalized recommendations by ML models was evaluated by the survival curves. The performance was highest in the DeepSurv model (C-index, 0.77; IBS, 0.14; 3-year AUC, 0.80; 5-year AUC, 0.78) compared with other methods (C-index, 0.73-0.74; IBS, 0.16-0.17; 3-year AUC, 0.73-0.78; 5-year AUC, 0.72-0.78). There are also significant differences in survival outcomes between patients who align with the treatment option recommended by the DeepSurv model and those who do not (hazard ratio, 1.88; P < .05). The DeepSurv model is available in an approachable web app format at https://survivalofosteosarcoma.streamlit.app/. We developed ML models capable of accurately predicting the survival of osteosarcoma, which can provide useful information for decision-making regarding the appropriate treatment.

Development and validation of a deep learning-based survival prediction model for pediatric glioma patients: A retrospective study using the SEER database and Chinese data

ObjectiveDevelop a time-dependent deep learning model to accurately predict the prognosis of pediatric glioma patients, which can assist clinicians in making precise treatment decisions and reducing patient risk. Study designThe study involved pediatric glioma patients from the Surveillance, Epidemiology, and End Results (SEER) Registry (2000–2018) and Tangdu Hospital in China (2010–2018) within specific time frames. For training, we selected two neural network-based algorithms (DeepSurv, neural multi-task logistic regression [N-MTLR]) and one ensemble learning-based algorithm (random survival forest [RSF]). Additionally, a multivariable Cox proportional hazard (CoxPH) model was developed for comparison purposes. The SEER dataset was randomly divided into 80 % for training and 20 % for testing, while the Tangdu Hospital dataset served as an external validation cohort. Super-parameters were fine-tuned through 1000 repeated random searches and 5-fold cross-validation on the training cohort. Model performance was assessed using the concordance index (C-index), Brier score, and Integrated Brier Score (IBS). Furthermore, the accuracy of predicting survival at 1, 3, and 5 years was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and the area under the ROC curves (AUC). The generalization ability of the model was assessed using the C-index of the Tangdu Hospital data, ROC curves for 1, 3, and 5 years, and AUC values. Lastly, decision curve analysis (DCA) curves for 1, 3, and 5-year time frames are provided to assess the net benefits across different models. ResultsA total of 9532 patients with pediatric glioma were included in this study, comprising 9274 patients from the SEER database and 258 patients from Tangdu Hospital in China. The average age at diagnosis was 9.4 ± 6.2 years, and the average survival time was 96 ± 66 months. Through comprehensive performance comparison, the DeepSurv model demonstrated the highest effectiveness, with a C-index of 0.881 on the training cohort. Furthermore, it exhibited excellent accuracy in predicting the 1-year, 3-year, and 5-year survival rates (AUC: 0.903–0.939). Notably, the DeepSurv model also achieved remarkable performance and accuracy on the Chinese dataset (C-index: 0.782, AUC: 0.761–0.852). Comprehensive analysis of DeepSurv, N-MTLR, and RSF revealed that tumor stage, radiotherapy, histological type, tumor size, chemotherapy, age, and surgical method are all significant factors influencing the prognosis of pediatric glioma. Finally, an online version of the pediatric glioma survival predictor based on the DeepSurv model has been established and can be accessed through https://pediatricglioma-tangdu.streamlit.app. ConclusionsThe DeepSurv model exhibits exceptional efficacy in predicting the survival of pediatric glioma patients, demonstrating strong performance in discrimination, calibration, stability, and generalization. By utilizing the online version of the pediatric glioma survival predictor, which is based on the DeepSurv model, clinicians can accurately predict patient survival and offer personalized treatment options.

Deep learning models for predicting the survival of patients with hepatocellular carcinoma based on a surveillance, epidemiology, and end results (SEER) database analysis

Hepatocellular carcinoma (HCC) is a common malignancy with poor survival and requires long-term follow-up. Hence, we collected information on patients with Primary Hepatocellular Carcinoma in the United States from the Surveillance, Epidemiology, and EndResults (SEER) database. We used this information to establish a deep learning with a multilayer neural network (the NMTLR model) for predicting the survival rate of patients with Primary Hepatocellular Carcinoma. HCC patients pathologically diagnosed between January 2011 and December 2015 in the SEER (Surveillance, Epidemiology, and End Results) database of the National Cancer Institute of the United States were selected as study subjects. We utilized two deep learning-based algorithms (DeepSurv and Neural Multi-Task Logistic Regression [NMTLR]) and a machine learning-based algorithm (Random Survival Forest [RSF]) for model training. A multivariable Cox Proportional Hazards (CoxPH) model was also constructed for comparison. The dataset was randomly divided into a training set and a test set in a 7:3 ratio. The training dataset underwent hyperparameter tuning through 1000 iterations of random search and fivefold cross-validation. Model performance was assessed using the concordance index (C-index), Brier score, and Integrated Brier Score (IBS). The accuracy of predicting 1-year, 3-year, and 5-year survival rates was evaluated using Receiver Operating Characteristic (ROC) curves, calibration plots, and Area Under the Curve (AUC). The primary outcomes were the 1-year, 3-year, and 5-year overall survival rates. Models were developed using DeepSurv, NMTLR, RSF, and Cox Proportional Hazards regression. Model differentiation was evaluated using the C-index, calibration with concordance plots, and risk stratification capability with the log-rank test. The study included 2197 HCC patients, randomly divided into a training cohort (70%, n = 1537) and a testing cohort (30%, n = 660). Clinical characteristics between the two cohorts showed no significant statistical difference (p > 0.05). The deep learning models outperformed both RSF and CoxPH models, with C-indices of 0.735 (NMTLR) and 0.731 (DeepSurv) in the test dataset. The NMTLR model demonstrated enhanced accuracy and well-calibrated survival estimates, achieving an Area Under the Curve (AUC) of 0.824 for 1-year survival predictions, 0.813 for 3-year, and 0.803 for 5-year survival rates. This model's superior calibration and discriminative ability enhance its utility for clinical prognostication in Primary Hepatocellular Carcinoma. We deployed the NMTLR model as a web application for clinical practice. The NMTLR model have potential advantages over traditional linear models in prognostic assessment and treatment recommendations. This novel analytical approach may provide reliable information on individual survival and treatment recommendations for patients with primary liver cancer.

Individual Survival Distributions Generated by Multi-Task Logistic Regression Yield a New Perspective on Molecular and Clinical Prognostic Factors in Gastric Adenocarcinoma.

Recent advances in our understanding of gastric cancer biology have prompted a shift towards more personalized therapy. However, results are based on population-based survival analyses, which evaluate the average survival effects of entire treatment groups or single prognostic variables. This study uses a personalized survival modelling approach called individual survival distributions (ISDs) with the multi-task logistic regression (MTLR) model to provide novel insight into personalized survival in gastric adenocarcinoma. We performed a pooled analysis using 1043 patients from a previously characterized database annotated with molecular subtypes from the Cancer Genome Atlas, Asian Cancer Research Group, and tumour microenvironment (TME) score. The MTLR model achieved a 5-fold cross-validated concordance index of 72.1 ± 3.3%. This model found that the TME score and chemotherapy had similar survival effects over the entire study time. The TME score provided the greatest survival benefit beyond a 5-year follow-up. Stage III and Stage IV disease contributed the greatest negative effect on survival. The MTLR model weights were significantly correlated with the Cox model coefficients (Pearson coefficient = 0.86, p < 0.0001). We illustrate how ISDs can accurately predict the survival time for each patient, which is especially relevant in cases of molecular subtype heterogeneity. This study provides evidence that the TME score is principally associated with long-term survival in gastric adenocarcinoma. Additional external validation and investigation into the clinical utility of this ISD model in gastric cancer is an area of future research.

Vision Transformer-Based Multilabel Survival Prediction for Oropharynx Cancer After Radiation Therapy

A reliable and comprehensive cancer prognosis model for oropharyngeal cancer (OPC) could better assist in personalizing treatment. In this work, we developed a vision transformer-based (ViT-based) multilabel model with multimodal input to learn complementary information from available pretreatment data and predict multiple associated endpoints for radiation therapy for patients with OPC. A publicly available data set of 512 patients with OPC was used for both model training and evaluation. Planning computed tomography images, primary gross tumor volume masks, and 16 clinical variables representing patient demographics, diagnosis, and treatment were used as inputs. To extract deep image features with global attention, we used a ViT module. Clinical variables were concatenated with the learned image features and fed into fully connected layers to incorporate cross-modality features. To learn the mapping between the features and correlated survival outcomes, including overall survival, local failure-free survival, regional failure-free survival, and distant failure-free survival, we employed 4 multitask logistic regression layers. The proposed model was optimized by combining the multitask logistic regression negative-log likelihood losses of different prediction targets. We employed the C-index and area under the curve metrics to assess the performance of our model for time-to-event prediction and time-specific binary prediction, respectively. Our proposed model outperformed corresponding single-modality and single-label models on all prediction labels, achieving C-indices of 0.773, 0.765, 0.776, and 0.773 for overall survival, local failure-free survival, regional failure-free survival, and distant failure-free survival, respectively. The area under the curve values ranged between 0.799 and 0.844 for different tasks at different time points. Using the medians of predicted risks as the thresholds to identify high-risk and low-risk patient groups, we performed the log-rank test, the results of which showed significantly larger separations in different event-free survivals. We developed the first model capable of predicting multiple labels for OPC simultaneously. Our model demonstrated better prognostic ability for all the prediction targets compared with corresponding single-modality models and single-label models.

Predicting Colorectal Cancer Survival Using Time-to-Event Machine Learning: Retrospective Cohort Study.

Machine learning (ML) methods have shown great potential in predicting colorectal cancer (CRC) survival. However, the ML models introduced thus far have mainly focused on binary outcomes and have not considered the time-to-event nature of this type of modeling. This study aims to evaluate the performance of ML approaches for modeling time-to-event survival data and develop transparent models for predicting CRC-specific survival. The data set used in this retrospective cohort study contains information on patients who were newly diagnosed with CRC between December 28, 2012, and December 27, 2019, at West China Hospital, Sichuan University. We assessed the performance of 6 representative ML models, including random survival forest (RSF), gradient boosting machine (GBM), DeepSurv, DeepHit, neural net-extended time-dependent Cox (or Cox-Time), and neural multitask logistic regression (N-MTLR) in predicting CRC-specific survival. Multiple imputation by chained equations method was applied to handle missing values in variables. Multivariable analysis and clinical experience were used to select significant features associated with CRC survival. Model performance was evaluated in stratified 5-fold cross-validation repeated 5 times by using the time-dependent concordance index, integrated Brier score, calibration curves, and decision curves. The SHapley Additive exPlanations method was applied to calculate feature importance. A total of 2157 patients with CRC were included in this study. Among the 6 time-to-event ML models, the DeepHit model exhibited the best discriminative ability (time-dependent concordance index 0.789, 95% CI 0.779-0.799) and the RSF model produced better-calibrated survival estimates (integrated Brier score 0.096, 95% CI 0.094-0.099), but these are not statistically significant. Additionally, the RSF, GBM, DeepSurv, Cox-Time, and N-MTLR models have comparable predictive accuracy to the Cox Proportional Hazards model in terms of discrimination and calibration. The calibration curves showed that all the ML models exhibited good 5-year survival calibration. The decision curves for CRC-specific survival at 5 years showed that all the ML models, especially RSF, had higher net benefits than default strategies of treating all or no patients at a range of clinically reasonable risk thresholds. The SHapley Additive exPlanations method revealed that R0 resection, tumor-node-metastasis staging, and the number of positive lymph nodes were important factors for 5-year CRC-specific survival. This study showed the potential of applying time-to-event ML predictive algorithms to help predict CRC-specific survival. The RSF, GBM, Cox-Time, and N-MTLR algorithms could provide nonparametric alternatives to the Cox Proportional Hazards model in estimating the survival probability of patients with CRC. The transparent time-to-event ML models help clinicians to more accurately predict the survival rate for these patients and improve patient outcomes by enabling personalized treatment plans that are informed by explainable ML models.

Learning Individual Survival Models from PanCancer Whole Transcriptome Data.

Personalized medicine attempts to predict survival time for each patient, based on their individual tumor molecular profile. We investigate whether our survival learner in combination with a dimension reduction method can produce useful survival estimates for a variety of patients with cancer. This article provides a method that learns a model for predicting the survival time for individual patients with cancer from the PanCancer Atlas: given the (16,335 dimensional) gene expression profiles from 10,173 patients, each having one of 33 cancers, this method uses unsupervised nonnegative matrix factorization (NMF) to reexpress the gene expression data for each patient in terms of 100 learned NMF factors. It then feeds these 100 factors into the Multi-Task Logistic Regression (MTLR) learner to produce cancer-specific models for each of 20 cancers (with >50 uncensored instances); this produces "individual survival distributions" (ISD), which provide survival probabilities at each future time for each individual patient, which provides a patient's risk score and estimated survival time. Our NMF-MTLR concordance indices outperformed the VAECox benchmark by 14.9% overall. We achieved optimal survival prediction using pan-cancer NMF in combination with cancer-specific MTLR models. We provide biological interpretation of the NMF model and clinical implications of ISDs for prognosis and therapeutic response prediction. NMF-MTLR provides many benefits over other models: superior model discrimination, superior calibration, meaningful survival time estimates, and accurate probabilistic estimates of survival over time for each individual patient. We advocate for the adoption of these cancer survival models in clinical and research settings.

Deep learning model for predicting the survival of patients with primary gastrointestinal lymphoma based onthe SEER database and a multicentre external validation cohort.

Due to the rarity of primary gastrointestinal lymphoma (PGIL), the prognostic factors and optimal management of PGIL have not been clearly defined. We aimed to establish prognostic models using a deep learning algorithm for survival prediction. We collected 11,168 PGIL patients from the Surveillance, Epidemiology, and End Results (SEER) database to form the training and test cohorts. At the same time, we collected 82 PGIL patients from three medical centres to form the external validation cohort. We constructed a Cox proportional hazards (CoxPH) model, random survival forest (RSF) model, and neural multitask logistic regression (DeepSurv) model to predict PGIL patients' overall survival (OS). The 1-, 3-, 5-, and 10-year OS rates of PGIL patients in the SEER database were 77.1%, 69.4%, 63.7%, and 50.3%, respectively. The RSF model based on all variables showed that the top three most important variables for predicting OS were age, histological type, and chemotherapy. The independent risk factors for PGIL patient prognosis included sex, age, race, primary site, Ann Arbor stage, histological type, symptom, radiotherapy, and chemotherapy, according to the Lasso regression analysis. Using these factors, we built the CoxPH and DeepSurv models. The DeepSurv model's C-index values were 0.760 in the training cohort, 0.742 in the test cohort, and 0.707 in the external validation cohort, which demonstrated that the DeepSurv model performed better compared to the RSF model (0.728) and the CoxPH model (0.724). The DeepSurv model accurately predicted 1-, 3-, 5- and 10-year OS. Both calibration curves and decision curve analysis curves demonstrated the superior performance of the DeepSurv model. We developed the DeepSurv model as an online web calculator for survival prediction, which can be accessed at http://124.222.228.112:8501/ . This DeepSurv model with external validation is superior to previous studies in predicting short-term and long-term survival and can help us make better-individualized decisions for PGIL patients.

Deep Learning Approaches for Glioblastoma Prognosis in Resource-Limited Settings: A Study Using Basic Patient Demographic, Clinical, and Surgical Inputs

Glioblastoma (GBM) is the most common brain tumor in the United States, with an annual incidence rate of 3.21 per 100,000. It is the most aggressive type of diffuse glioma and has a median survival of months after treatment. This study aims to assess the accuracy of different novel deep learning models trained on a set of simple clinical, demographic, and surgical variables to assist in clinical practice, even in areas with constrained health care infrastructure. Our study included 37,095 patients with GBM from the SEER (Surveillance Epidemiology and End Results) database. All predictors were based on demographic, clinicopathologic, and treatment information of the cases. Our outcomes of interest were months of survival and vital status. Concordance index (C-index) and integrated Brier scores (IBS) were used to evaluate the performance of the models. The patient characteristics and the statistical analyses were consistent with the epidemiologic literature. The models C-index and IBS ranged from 0.6743 to 0.6918 and from 0.0934 to 0.1034, respectively. Probabilistic matrix factorization (0.6918), multitask logistic regression (0.6916), and logistic hazard (0.6916) had the highest C-index scores. The models with the lowest IBS were the probabilistic matrix factorization (0.0934), multitask logistic regression (0.0935), and logistic hazard (0.0936). These models had an accuracy (1-IBS) of 90.66%; 90.65%, and 90.64%, respectively. The deep learning algorithms were deployed on an interactive Web-based tool for practical use available via https://glioblastoma-survanalysis.herokuapp.com/. Novel deep learning algorithms can better predict GBM prognosis than do baseline methods and can lead to more personalized patient care regardless of extensive electronic health record availability.

Deep learning-extracted CT imaging phenotypes predict response to total resection in colorectal cancer.

Deep learning surpasses many traditional methods for many vision tasks, allowing the transformation of hierarchical features into more abstract, high-level features. To evaluate the prognostic value of preoperative computed tomography (CT) image texture features and deep learning self-learning high-throughput features (SHF) on postoperative overall survival in the treatment of patients with colorectal cancer (CRC). The dataset consisted of 810 enrolled patients with CRC confirmed from 10 November 2011 to 10 February 2018. In contrast, SHF extracted by deep learning with multi-task training mechanism and texture features were extracted from the CT with tumor volume region of interest, respectively, and combined with the Cox proportional hazard (CoxPH) model for initial validation to obtain a RAD score to classify patients into high- and low-risk groups. The SHF stability was further validated in combination with Neural Multi-Task Logistic Regression (N-MTLR) model. The overall recognition ability and accuracy of CoxPH and N-MTLR model were evaluated by C-index and Integrated Brier Score (IBS). SHF had a more significant degree of differentiation than texture features. The result is (SHF vs. texture features: C-index: 0.884 vs. 0.611; IBS: 0.025 vs. 0.073) in the CoxPH model, and (SHF vs. texture features: C-index: 0.861 vs. 0.630; IBS: 0.024 vs. 0.065) in N-MTLR. SHF is superior to texture features and has potential application for the preoperative prediction of the individualized treatment of CRC.

Artificial Intelligence Outperforms Kaplan–Meier Analyses Estimating Survival after Elective Treatment of Abdominal Aortic Aneurysms

Long term differences in survival after elective repair of abdominal aortic aneurysms (AAAs) between open surgical repair (OSR) and endovascular aneurysm repair (EVAR) are unclear, and hitherto artificial intelligence has not been used for this purpose. The aim was to compare the precision of survival estimates between the Kaplan-Meier (KM) method and the artificial intelligence derived method Neural Multi-Task Logistic Regression (N-MTLR), and to compare survival estimates as a function of patient age and time since surgery between OSR and EVAR using N-MTLR. All AAAs in Denmark were identified in the Danish vascular registry between 2003 and 2018. Survival was estimated using the KM and N-MTLR methods, and prediction performance was estimated with the Brier score. 7 912 patients were included in the study, n = 6 569 (83%) men, median age 72 years (range 35 - 92), with a median follow up time of 45.7 months (range 0 - 120). The two treatment cohorts, OSR n = 5 495 (69%), and EVAR n = 2 417 (31%), differed significantly in patient characteristics. The Brier score for KM increased from 0.044 to 0.244, and for N-MTLR from 0.044 to 0.206, from 90 days to 10 years. The N-MTLR method was more accurate than KM from 90 days to 10 years after surgery, p ≤ .025. N-MTLR demonstrated significant increased probability for survival for OSR in patients aged 58 - 76 years at five years, and 65 - 73 at 10 years after surgery, and the opposite was found for the benefit of EVAR in patients aged 72 - 85 years at one year, 85 - 90 years at five years, and for 85 - 90 year olds at 10 years after surgery. N-MTLR outperforms KM for the entire post-operative follow up time. This N-MTLR model has the potential to render more precise patient specific survival estimates, and establish survival differences between subgroups of patients that KM is unable to detect, demonstrated here for different age groups.

Looking Through the Crystal Ball of Data Science

Artificial Intelligence Outperforms Kaplan–Meier Analyses Estimating Survival after Elective Treatment of Abdominal Aortic AneurysmsEuropean Journal of Vascular and Endovascular SurgeryPreviewLong term differences in survival after elective repair of abdominal aortic aneurysms (AAAs) between open surgical repair (OSR) and endovascular aneurysm repair (EVAR) are unclear, and hitherto artificial intelligence has not been used for this purpose. The aim was to compare the precision of survival estimates between the Kaplan–Meier (KM) method and the artificial intelligence derived method Neural Multi-Task Logistic Regression (N-MTLR), and to compare survival estimates as a function of patient age and time since surgery between OSR and EVAR using N-MTLR. Full-Text PDF Open Access

Personalized prediction of incident hospitalization for cardiovascular disease in patients with hypertension using machine learning

BackgroundPrognostic information for patients with hypertension is largely based on population averages. The purpose of this study was to compare the performance of four machine learning approaches for personalized prediction of incident hospitalization for cardiovascular disease among newly diagnosed hypertensive patients.MethodsUsing province-wide linked administrative health data in Alberta, we analyzed a cohort of 259,873 newly-diagnosed hypertensive patients from 2009 to 2015 who collectively had 11,863 incident hospitalizations for heart failure, myocardial infarction, and stroke. Linear multi-task logistic regression, neural multi-task logistic regression, random survival forest and Cox proportional hazard models were used to determine the number of event-free survivors at each time-point and to construct individual event-free survival probability curves. The predictive performance was evaluated by root mean squared error, mean absolute error, concordance index, and the Brier score.ResultsThe random survival forest model has the lowest root mean squared error value at 33.94 and lowest mean absolute error value at 28.37. Machine learning methods provide similar discrimination and calibration in the personalized survival prediction of hospitalizations for cardiovascular events in patients with hypertension. Neural multi-task logistic regression model has the highest concordance index at 0.8149 and lowest Brier score at 0.0242 for the personalized survival prediction.ConclusionsThis is the first personalized survival prediction for cardiovascular diseases among hypertensive patients using administrative data. The four models tested in this analysis exhibited a similar discrimination and calibration ability in predicting personalized survival prediction of hypertension patients.

S26 Comparison of Machine and Deep Learning Methods With the Mayo Clinical Risk Score in Predicting 1-Year Mortality in Primary Biliary Cholangitis Patients

Introduction: To compare machine learning (ML) and deep learning (DL) approaches to predict survival outcome in patients at 1- year with Primary Biliary Cholangitis (PBC) and to compare with the Mayo Clinical risk score Methods: The data is taken from a randomized trial in PBC patients conducted between 1974 and 1984 by the Mayo Clinic. The first 312 cases in the dataset were included out of a total of 424 PBC patients referred. The rest that contained subjects with >30% missing data, were excluded. We evaluated the following models: 1) The 5-covariate Mayo model: log(bilirubin), albumin, log(prothrombin time), edema, and age, 2) 17-covariate Cox proportional hazards model by componentwise likelihood based boosting with stepnumber 10 and penalty number 100, 3) 17-covariate Deephit: DeepHit: trains a neural network to learn the estimated joint distribution of survival time and event, while capturing the right-censored nature inherent in survival data. Analysis was done with frac 0.3, relu activation, 0.1 dropout, 100L epochs and a batch size of 32L, 4) 17-covariate Multitask logistic regression model with ranking based feature selection to predict survival using a logistic regression model and the parameters from each model are estimated simultaneously in the maximization of the joint likelihood function, 5) random survival forest with 1000 trees and 6) Support vector machine. For the ML & DL models all 17-covariates were include in the analysis. Analyses were done in RStudio, and missing values were imputed using missRanger package.1 The data is split into 80 percent training and 20 precent validation. All models were 5-fold cross validated. Prediciton statistics was calculated for each model developed. Finally, the ML & DL models were compared with the 5-covariate model of the Mayo Clinic Risk score. Results: Multitask logistic regression model showed the highest AUC (1.00) with a Harrell’s C-statistic of 0.90. Cox boost and random survival forest had equal AUCs (0.87) and C-indices (0.84). But the 5-covariate model of Mayo clinic Risk score had a higher AUC when compared to cox boost, random survival forest, support vector machine or even Deep hit. Conclusion: Machine learning methods offered limited improvement over the Mayo Clinical risk score except for the Multitask logistic regression model in predicting PBC survival outcome at 1-year (Table). Table 1. - Performance of Risk Scores from the 5-Covariate Model in comparison to the machine learning and deep learning models at 1-year S.No. Model specification AUC 95% CI Harrell’s C-statistic 1. 5-covariate model of Mayo clinic Risk score 0.88 0.85-0.91 0.79 2. Cox boost 0.87 0.84-0.90 0.84 3. Deep hit 0.70 0.66-0.74 0.30 4. Multitask logistic regression 1.00 1-1 0.90 5. Random survival forest 0.87 0.84-0.90 0.84 6. Support Vector Machine 0.56 0.52-0.61 0.43

Deep learning models for predicting the survival of patients with chondrosarcoma based on a surveillance, epidemiology, and end results analysis.

BackgroundAccurate prediction of prognosis is critical for therapeutic decisions in chondrosarcoma patients. Several prognostic models have been created utilizing multivariate Cox regression or binary classification-based machine learning approaches to predict the 3- and 5-year survival of patients with chondrosarcoma, but few studies have investigated the results of combining deep learning with time-to-event prediction. Compared with simplifying the prediction as a binary classification problem, modeling the probability of an event as a function of time by combining it with deep learning can provide better accuracy and flexibility.Materials and methodsPatients with the diagnosis of chondrosarcoma between 2000 and 2018 were extracted from the Surveillance, Epidemiology, and End Results (SEER) registry. Three algorithms—two based on neural networks (DeepSurv, neural multi-task logistic regression [NMTLR]) and one on ensemble learning (random survival forest [RSF])—were selected for training. Meanwhile, a multivariate Cox proportional hazards (CoxPH) model was also constructed for comparison. The dataset was randomly divided into training and testing datasets at a ratio of 7:3. Hyperparameter tuning was conducted through a 1000-repeated random search with 5-fold cross-validation on the training dataset. The model performance was assessed using the concordance index (C-index), Brier score, and Integrated Brier Score (IBS). The accuracy of predicting 1-, 3-, 5- and 10-year survival was evaluated using receiver operating characteristic curves (ROC), calibration curves, and the area under the ROC curves (AUC).ResultsA total of 3145 patients were finally enrolled in our study. The mean age at diagnosis was 52 ± 18 years, 1662 of the 3145 patients were male (53%), and mean survival time was 83 ± 67 months. Two deep learning models outperformed the RSF and classical CoxPH models, with the C-index on test datasets achieving values of 0.832 (DeepSurv) and 0.821 (NMTLR). The DeepSurv model produced better accuracy and calibrated survival estimates in predicting 1-, 3- 5- and 10-year survival (AUC:0.895-0.937). We deployed the DeepSurv model as a web application for use in clinical practice; it can be accessed through https://share.streamlit.io/whuh-ml/chondrosarcoma/Predict/app.py.ConclusionsTime-to-event prediction models based on deep learning algorithms are successful in predicting chondrosarcoma prognosis, with DeepSurv producing the best discriminative performance and calibration.

Improve individual treatment by comparing treatment benefits: cancer artificial intelligence survival analysis system for cervical carcinoma

PurposeThe current study aimed to construct a novel cancer artificial intelligence survival analysis system for predicting the individual mortality risk curves for cervical carcinoma patients receiving different treatments.MethodsStudy dataset (n = 14,946) was downloaded from Surveillance Epidemiology and End Results database. Accelerated failure time algorithm, multi-task logistic regression algorithm, and Cox proportional hazard regression algorithm were used to develop prognostic models for cancer specific survival of cervical carcinoma patients.ResultsMultivariate Cox regression identified stage, PM, chemotherapy, Age, PT, and radiation_surgery as independent influence factors for cervical carcinoma patients. The concordance indexes of Cox model were 0.860, 0.849, and 0.848 for 12-month, 36-month, and 60-month in model dataset, whereas it were 0.881, 0.845, and 0.841 in validation dataset. The concordance indexes of accelerated failure time model were 0.861, 0.852, and 0.851 for 12-month, 36-month, and 60-month in model dataset, whereas it were 0.882, 0.847, and 0.846 in validation dataset. The concordance indexes of multi-task logistic regression model were 0.860, 0.863, and 0.861 for 12-month, 36-month, and 60-month in model dataset, whereas it were 0.880, 0.860, and 0.861 in validation dataset. Brier score indicated that these three prognostic models have good diagnostic accuracy for cervical carcinoma patients. The current research lacked independent external validation study.ConclusionThe current study developed a novel cancer artificial intelligence survival analysis system to provide individual mortality risk predictive curves for cervical carcinoma patients based on three different artificial intelligence algorithms. Cancer artificial intelligence survival analysis system could provide mortality percentage at specific time points and explore the actual treatment benefits under different treatments in four stages, which could help patient determine the best individualized treatment. Cancer artificial intelligence survival analysis system was available at: https://zhangzhiqiao15.shinyapps.io/Tumor_Artificial_Intelligence_Survival_Analysis_System/.

Machine learning for optimized individual survival prediction in resectable upper gastrointestinal cancer

PurposeSurgical oncologists are frequently confronted with the question of expected long-term prognosis. The aim of this study was to apply machine learning algorithms to optimize survival prediction after oncological resection of gastroesophageal cancers.MethodsEligible patients underwent oncological resection of gastric or distal esophageal cancer between 2001 and 2020 at Heidelberg University Hospital, Department of General Surgery. Machine learning methods such as multi-task logistic regression and survival forests were compared with usual algorithms to establish an individual estimation.ResultsThe study included 117 variables with a total of 1360 patients. The overall missingness was 1.3%. Out of eight machine learning algorithms, the random survival forest (RSF) performed best with a concordance index of 0.736 and an integrated Brier score of 0.166. The RSF demonstrated a mean area under the curve (AUC) of 0.814 over a time period of 10 years after diagnosis. The most important long-term outcome predictor was lymph node ratio with a mean AUC of 0.730. A numeric risk score was calculated by the RSF for each patient and three risk groups were defined accordingly. Median survival time was 18.8 months in the high-risk group, 44.6 months in the medium-risk group and above 10 years in the low-risk group.ConclusionThe results of this study suggest that RSF is most appropriate to accurately answer the question of long-term prognosis. Furthermore, we could establish a compact risk score model with 20 input parameters and thus provide a clinical tool to improve prediction of oncological outcome after upper gastrointestinal surgery.

Machine Learning-Based Individualized Survival Prediction Model for Total Knee Replacement in Osteoarthritis: Data From the Osteoarthritis Initiative.

By using machine learning, our study aimed to build a model to predict risk and time to total knee replacement (TKR) of an osteoarthritic knee. Features were from the Osteoarthritis Initiative (OAI) cohort at baseline. Using the lasso method for variable selection in the Cox regression model, we identified the 10 most important characteristics among 1,107 features. The prognostic power of the selected features was assessed by the Kaplan-Meier method and applied to 7 machine learning methods: Cox, DeepSurv, random forests algorithm, linear/kernel support vector machine (SVM), and linear/neural multi-task logistic regression models. As some of the 10 first-found features included similar radiographic measurements, we further looked at using the least number of features without compromising the accuracy of the model. Prediction performance was assessed by the concordance index, Brier score, and time-dependent area under the curve (AUC). Ten features were identified and included radiographs, bone marrow lesions of the medial condyle on magnetic resonance imaging, hyaluronic acid injection, performance measure, medical history, and knee-related symptoms. The methodologies Cox, DeepSurv, and linear SVM demonstrated the highest accuracy (concordance index scores of 0.85, Brier score of 0.02, and an AUC of 0.87). DeepSurv was chosen to build the prediction model to estimate the time to TKR for a given knee. Moreover, we were able to decrease the features to only 3 and maintain the high accuracy (concordance index of 0.85, Brier score of 0.02, and AUC of 0.86), which included bone marrow lesions, Kellgren/Lawrence grade, and knee-related symptoms, to predict risk and time of a TKR event. For the first time, we developed a model using the OAI cohort to predict with high accuracy if a given osteoarthritic knee would require TKR, when a TKR would be required, and who would likely progress fast toward this event.

Identification of Potential Biomarkers and Biological Pathways for Poor Clinical Outcome in Mucinous Colorectal Adenocarcinoma.

Simple SummaryPatients with mucinous adenocarcinoma (MAC) have been considered to have a faster disease progression than patients with traditional adenocarcinoma (TAC) in colorectal cancer (CRC). However, to date, the roles of MAC in long-term survival remain controversial due to a small sample size and the nature of its relatively rare occurrence, although it potentially represents entities with different aggressiveness and prognoses. Here, using large-scale population data, we found that the patients with the MAC subtype had a significantly worse overall survival rate and a tendency of worse disease-specific survival rate in stage II compared with the patients with the TAC subtype. Furthermore, key gene signatures were identified using the established predictive models for the disease-specific survival of stage II mucinous CRC.Colorectal cancer (CRC) comprises several histological subtypes, but the influences of the histological subtypes on prognosis remains unclear. We sought to evaluate the prognosis of mucinous adenocarcinoma (MAC), compared to that of traditional adenocarcinoma (TAC). This study used the data of patients diagnosed with CRC between 2004 and 2016, as obtained from the Surveillance, Epidemiology, and End Results database. We established a predictive model for disease-specific survival using conditional survival forest, model, non-linear Cox proportional hazards, and neural multi-task logistic regression model and identified the gene signatures for predicting poor prognosis based on the arrayexpress datasets. In total, 9096 (42.1%) patients with MAC and 12,490 (58.9%) patients with TAC were included. Those with the MAC subtype were more likely to have a poorer overall survival rate compared to those with the TAC subtype in stage II CRC (p = 0.002). The eight major genes including RPS18, RPL30, NME2, USP33, GAB2, RPS3A, RPS25, and CEP57 were found in the interacting network pathway. MAC was found to have a poorer prognosis compared to TAC, especially in Stage II CRC. In addition, our findings suggest that identifying potential biomarkers and biological pathways can be useful in CRC prognosis.