Deep Learning-Based Survival Analysis for High-Dimensional Survival Data

deepAFT: A nonlinear accelerated failure time model with artificial neural network

The Cox regression model or accelerated failure time regression models are often used for describing the relationship between survival outcomes and potential explanatory variables. These models assume the studied covariates are connected to the survival time or its distribution or their transformations through a function of a linear regression form. In this article, we propose nonparametric, nonlinear algorithms (deepAFT methods) based on deep artificial neural networks to model survival outcome data in the broad distribution family of accelerated failure time models. The proposed methods predict survival outcomes directly and tackle the problem of censoring via an imputation algorithm as well as re-weighting and transformation techniques based on the inverse probabilities of censoring. Through extensive simulation studies, we confirm that the proposed deepAFT methods achieve accurate predictions. They outperform the existing regression models in prediction accuracy, while being flexible and robust in modeling covariate effects of various nonlinear forms. Their prediction performance is comparable to other established deep learning methods such as deepSurv and random survival forest methods. Even though the direct output is the expected survival time, the proposed AFT methods also provide predictions for distributional functions such as the cumulative hazard and survival functions without additional learning efforts. For situations where the popular Cox regression model may not be appropriate, the deepAFT methods provide useful and effective alternatives, as shown in simulations, and demonstrated in applications to a lymphoma clinical trial study.

Cross-attention enables deep learning on limited omics-imaging-clinical data of 130 lung cancer patients

Deep-learning tools that extract prognostic factors derived from multi-omics data have recently contributed to individualized predictions of survival outcomes. However, the limited size of integrated omics-imaging-clinical datasets poses challenges. Here, we propose two biologically interpretable and robust deep-learning architectures for survival prediction of non-small cell lung cancer (NSCLC) patients, learning simultaneously from computed tomography (CT) scan images, gene expression data, and clinical information. The proposed models integrate patient-specific clinical, transcriptomic, and imaging data and incorporate Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome pathway information, adding biological knowledge within the learning process to extract prognostic gene biomarkers and molecular pathways. While both models accurately stratify patients in high- and low-risk groups when trained on a dataset of only 130 patients, introducing a cross-attention mechanism in a sparse autoencoder significantly improves the performance, highlighting tumor regions and NSCLC-related genes as potential biomarkers and thus offering a significant methodological advancement when learning from small imaging-omics-clinical samples.

CTIVA: Censored time interval variable analysis

Traditionally, datasets with multiple censored time-to-events have not been utilized in multivariate analysis because of their high level of complexity. In this paper, we propose the Censored Time Interval Analysis (CTIVA) method to address this issue. It estimates the joint probability distribution of actual event times in the censored dataset by implementing a statistical probability density estimation technique on the dataset. Based on the acquired event time, CTIVA investigates variables correlated with the interval time of events via statistical tests. The proposed method handles both categorical and continuous variables simultaneously-thus, it is suitable for application on real-world censored time-to-event datasets, which include both categorical and continuous variables. CTIVA outperforms traditional censored time-to-event data handling methods by 5% on simulation data. The average area under the curve (AUC) of the proposed method on the simulation dataset exceeds 0.9 under various conditions. Further, CTIVA yields novel results on National Sample Cohort Demo (NSCD) and proteasome inhibitor bortezomib dataset, a real-world censored time-to-event dataset of medical history of beneficiaries provided by the National Health Insurance Sharing Service (NHISS) and National Center for Biotechnology Information (NCBI). We believe that the development of CTIVA is a milestone in the investigation of variables correlated with interval time of events in presence of censoring.

Interpretable deep learning for improving cancer patient survival based on personal transcriptomes

Precision medicine chooses the optimal drug for a patient by considering individual differences. With the tremendous amount of data accumulated for cancers, we develop an interpretable neural network to predict cancer patient survival based on drug prescriptions and personal transcriptomes (CancerIDP). The deep learning model achieves 96% classification accuracy in distinguishing short-lived from long-lived patients. The Pearson correlation between predicted and actual months-to-death values is as high as 0.937. About 27.4% of patients may survive longer with an alternative medicine chosen by our deep learning model. The median survival time of all patients can increase by 3.9 months. Our interpretable neural network model reveals the most discriminating pathways in the decision-making process, which will further facilitate mechanistic studies of drug development for cancers.

Neural Networks for Survival Prediction in Medicine Using Prognostic Factors: A Review and Critical Appraisal

Survival analysis deals with the expected duration of time until one or more events of interest occur. Time to the event of interest may be unobserved, a phenomenon commonly known as right censoring, which renders the analysis of these data challenging. Over the years, machine learning algorithms have been developed and adapted to right-censored data. Neural networks have been repeatedly employed to build clinical prediction models in healthcare with a focus on cancer and cardiology. We present the first ever attempt at a large-scale review of survival neural networks (SNNs) with prognostic factors for clinical prediction in medicine. This work provides a comprehensive understanding of the literature (24 studies from 1990 to August 2021, global search in PubMed). Relevant manuscripts are classified as methodological/technical (novel methodology or new theoretical model; 13 studies) or applications (11 studies). We investigate how researchers have used neural networks to fit survival data for prediction. There are two methodological trends: either time is added as part of the input features and a single output node is specified, or multiple output nodes are defined for each time interval. A critical appraisal of model aspects that should be designed and reported more carefully is performed. We identify key characteristics of prediction models (i.e., number of patients/predictors, evaluation measures, calibration), and compare ANN's predictive performance to the Cox proportional hazards model. The median sample size is 920 patients, and the median number of predictors is 7. Major findings include poor reporting (e.g., regarding missing data, hyperparameters) as well as inaccurate model development/validation. Calibration is neglected in more than half of the studies. Cox models are not developed to their full potential and claims for the performance of SNNs are exaggerated. Light is shed on the current state of art of SNNs in medicine with prognostic factors. Recommendations are made for the reporting of clinical prediction models. Limitations are discussed, and future directions are proposed for researchers who seek to develop existing methodology.

Default risk prediction and feature extraction using a penalized deep neural network

Online peer-to-peer lending platforms provide loans directly from lenders to borrowers without passing through traditional financial institutions. For lenders on these platforms to avoid loss, it is crucial that they accurately assess default risk so that they can make appropriate decisions. In this study, we develop a penalized deep learning model to predict default risk based on survival data. As opposed to simply predicting whether default will occur, we focus on predicting the probability of default over time. Moreover, by adding an additional one-to-one layer in the neural network, we achieve feature selection and estimation simultaneously by incorporating an L 1 -penalty into the objective function. The minibatch gradient descent algorithm makes it possible to handle massive data. An analysis of a real-world loan data and simulations demonstrate the model's competitive practical performance, which suggests favorable potential applications in peer-to-peer lending platforms.

Prediction of survival in oropharyngeal squamous cell carcinoma using machine learning algorithms: A study based on the surveillance, epidemiology, and end results database

Background

We determined appropriate survival prediction machine learning models for patients with oropharyngeal squamous cell carcinoma (OPSCC) using the “Surveillance, Epidemiology, and End Results” (SEER) database.

Methods

In total, 4039 patients diagnosed with OPSCC between 2004 and 2016 were enrolled in this study. In particular, 13 variables were selected and analyzed: age, sex, tumor grade, tumor size, neck dissection, radiation therapy, cancer directed surgery, chemotherapy, T stage, N stage, M stage, clinical stage, and human papillomavirus (HPV) status. The T-, N-, and clinical staging were reconstructed based on the American Joint Committee on Cancer (AJCC) Staging Manual, 8th Edition. The patients were randomly assigned to a development or test dataset at a 7:3 ratio. The extremely randomized survival tree (EST), conditional survival forest (CSF), and DeepSurv models were used to predict the overall and disease-specific survival in patients with OPSCC. A 10-fold cross-validation on a development dataset was used to build the training and internal validation data for all models. We evaluated the predictive performance of each model using test datasets.

Results

A higher c-index value and lower integrated Brier score (IBS), root mean square error (RMSE), and mean absolute error (MAE) indicate a better performance from a machine learning model. The C-index was the highest for the DeepSurv model (0.77). The IBS was also the lowest in the DeepSurv model (0.08). However, the RMSE and RAE were the lowest for the CSF model.

Conclusions

We demonstrated various machine-learning-based survival prediction models. The CSF model showed a better performance in predicting the survival of patients with OPSCC in terms of the RMSE and RAE. In this context, machine learning models based on personalized survival predictions can be used to stratify various complex risk factors. This could help in designing personalized treatments and predicting prognoses for patients.

MSGWO-MKL-SVM: A Missing Link Prediction Method for UAV Swarm Network Based on Time Series

Missing link prediction technology (MLP) is always a hot research area in the field of complex networks, and it has been extensively utilized in UAV swarm network reconstruction recently. UAV swarm is an artificial network with strong randomness, in the face of which prediction methods based on network similarity often perform poorly. To solve those problems, this paper proposes a Multi Kernel Learning algorithm with a multi-strategy grey wolf optimizer based on time series (MSGWO-MKL-SVM). The Multiple Kernel Learning (MKL) method is adopted in this algorithm to extract the advanced features of time series, and the Support Vector Machine (SVM) algorithm is used to determine the hyperplane of threshold value in nonlinear high dimensional space. Besides that, we propose a new measurable indicator of Multiple Kernel Learning based on cluster, transforming a Multiple Kernel Learning problem into a multi-objective optimization problem. Some adaptive neighborhood strategies are used to enhance the global searching ability of grey wolf optimizer algorithm (GWO). Comparison experiments were conducted on the standard UCI datasets and the professional UAV swarm datasets. The classification accuracy of MSGWO-MKL-SVM on UCI datasets is improved by 6.2% on average, and the link prediction accuracy of MSGWO-MKL-SVM on professional UAV swarm datasets is improved by 25.9% on average.

Artificial Intelligence-Based Prognostic Model for Urologic Cancers: A SEER-Based Study

Simple Summary

We describe a risk profile reconstruction model for cancer-specific survival estimation for continuous time points after urologic cancer diagnosis. We used artificial intelligence (AI)-based algorithms, a national cancer registry data, and accessible clinical parameters for the risk-profile reconstruction. We derived a risk stratification model and estimated the minimum follow-up duration and the likelihood for risk stability in prostate, kidney, and testicular cancers. The estimated follow-up duration was in alignment with recognized clinical guidelines for these cancers. Moreover, the estimated follow-up duration was differed by the cancer origin and the disease dissemination status. Overall, the reconstruction of the population’s risk profile for the cancer-specific prognostic score estimation is feasible using AI and has potential application in clinical settings to improve risk stratification and surveillance management.

Abstract

Background: Prognostication is essential to determine the risk profile of patients with urologic cancers. Methods: We utilized the SEER national cancer registry database with approximately 2 million patients diagnosed with urologic cancers (penile, testicular, prostate, bladder, ureter, and kidney). The cohort was randomly divided into the development set (90%) and the out-held test set (10%). Modeling algorithms and clinically relevant parameters were utilized for cancer-specific mortality prognosis. The model fitness for the survival estimation was assessed using the differences between the predicted and observed Kaplan–Meier estimates on the out-held test set. The overall concordance index (c-index) score estimated the discriminative accuracy of the survival model on the test set. A simulation study assessed the estimated minimum follow-up duration and time points with the risk stability. Results: We achieved a well-calibrated prognostic model with an overall c-index score of 0.800 (95% CI: 0.795–0.805) on the representative out-held test set. The simulation study revealed that the suggestions for the follow-up duration covered the minimum duration and differed by the tumor dissemination stages and affected organs. Time points with a high likelihood for risk stability were identifiable. Conclusions: A personalized temporal survival estimation is feasible using artificial intelligence and has potential application in clinical settings, including surveillance management.

Deep Learning for Radiotherapy Outcome Prediction Using Dose Data - A Review

Artificial intelligence, and in particular deep learning using convolutional neural networks, has been used extensively for image classification and segmentation, including on medical images for diagnosis and prognosis prediction. Use in radiotherapy prognostic modelling is still limited, however, especially as applied to toxicity and tumour response prediction from radiation dose distributions. We review and summarise studies that applied deep learning to radiotherapy dose data, in particular studies that utilised full three-dimensional dose distributions. Ten papers have reported on deep learning models for outcome prediction utilising spatial dose information, whereas four studies used reduced dimensionality (dose volume histogram) information for prediction. Many of these studies suffer from the same issues that plagued early normal tissue complication probability modelling, including small, single-institutional patient cohorts, lack of external validation, poor data and model reporting, use of late toxicity data without taking time-to-event into account, and nearly exclusive focus on clinician-reported complications. They demonstrate, however, how radiation dose, imaging and clinical data may be technically integrated in convolutional neural networks-based models; and some studies explore how deep learning may help better understand spatial variation in radiosensitivity. In general, there are a number of issues specific to the intersection of radiotherapy outcome modelling and deep learning, for example translation of model developments into treatment plan optimisation, which will require further combined effort from the radiation oncology and artificial intelligence communities.