Abstract:Background: For the purpose to examine lower limb melanoma (LLM) and its long-term survival rate, we used data from the Surveillance, Epidemiology and End Results (SEER) database. To estimate the prognosis of LLM patients and assess its efficacy, we used a powerful deep learning and neural network approach called DeepSurv. Methods: We gathered data on those who had an LLM diagnosis between 2000 and 2019 from the SEER database. We divided the people into training and testing cohorts at a 7:3 ratio using a random selection technique. To assess the likelihood that LLM patients would survive, we compared the results of the DeepSurv model with those of the Cox proportional-hazards (CoxPH) model. Calibration curves, the time-dependent area under the receiver operating characteristic curve (AUC), and the concordance index (C-index) were all used to assess how accurate the predictions were. Results: In this study, a total of 26,243 LLM patients were enrolled, with 7873 serving as the testing cohort and 18,370 as the training cohort. Significant correlations with age, gender, AJCC stage, chemotherapy status, surgery status, regional lymph node removal and the survival outcomes of LLM patients were found by the CoxPH model. The CoxPH model's C-index was 0.766, which signifies a good degree of predicted accuracy. Additionally, we created the DeepSurv model using the training cohort data, which had a higher C-index of 0.852. In addition to calculating the 3-, 5-, and 8-year AUC values, the predictive performance of both models was evaluated. The equivalent AUC values for the CoxPH model were 0.795, 0.767, and 0.847, respectively. The DeepSurv model, in comparison, had better AUC values of 0.872, 0.858, and 0.847. In comparison to the CoxPH model, the DeepSurv model demonstrated greater prediction performance for LLM patients, as shown by the AUC values and the calibration curve. Conclusion: We created the DeepSurv model using LLM patient data from the SEER database, which performed better than the CoxPH model in predicting the survival time of LLM patients.

Regularizing the Deepsurv Network Using Projection Loss for Medical Risk Assessment

Deep Semisupervised Multitask Learning Model and Its Interpretability for Survival Analysis.

Deep Correlational Learning for Survival Prediction from Multi-modality Data.

ResDeepSurv: A Survival Model for Deep Neural Networks Based on Residual Blocks and Self-attention Mechanism

Multiview Deep Learning-based Efficient Medical Data Management for Survival Time Forecasting

Attention-Based Deep Recurrent Model for Survival Prediction

Spatio-temporally smoothed deep survival neural network

DeepSurv landmarking: a deep learning approach for dynamic survival analysis with longitudinal data

ResSurv: Cancer Survival Analysis Prediction Model Based on Residual Networks

Joint analysis of clinical risk factors and 4D cardiac motion for survival prediction using a hybrid deep learning network

Deep Learning of Semi-Competing Risk Data via a New Neural Expectation-Maximization Algorithm

Survival analysis of localized prostate cancer with deep learning

SurvLatent ODE : A Neural ODE based time-to-event model with competing risks for longitudinal data improves cancer-associated Deep Vein Thrombosis (DVT) prediction

DySurv: dynamic deep learning model for survival analysis with conditional variational inference

DNNSurv: Deep Neural Networks for Survival Analysis Using Pseudo Values

SurvRNC: Learning Ordered Representations for Survival Prediction using Rank-N-Contrast

Deep-CR MTLR: a Multi-Modal Approach for Cancer Survival Prediction with Competing Risks

Deep-CSA: Deep Contrastive Learning for Dynamic Survival Analysis With Competing Risks

Deep-learning-based survival prediction of patients with lower limb melanoma

Stabilized Sparse Ordinal Regression for Medical Risk Stratification

MS-ResNet: disease-specific survival prediction using longitudinal CT images and clinical data