Dataset Information

A Multimodal Affinity Fusion Network for Predicting the Survival of Breast Cancer Patients.

ABSTRACT: Accurate survival prediction of breast cancer holds significant meaning for improving patient care. Approaches using multiple heterogeneous modalities such as gene expression, copy number alteration, and clinical data have showed significant advantages over those with only one modality for patient survival prediction. However, existing survival prediction methods tend to ignore the structured information between patients and multimodal data. We propose a multimodal data fusion model based on a novel multimodal affinity fusion network (MAFN) for survival prediction of breast cancer by integrating gene expression, copy number alteration, and clinical data. First, a stack-based shallow self-attention network is utilized to guide the amplification of tiny lesion regions on the original data, which locates and enhances the survival-related features. Then, an affinity fusion module is proposed to map the structured information between patients and multimodal data. The module endows the network with a stronger fusion feature representation and discrimination capability. Finally, the fusion feature embedding and a specific feature embedding from a triple modal network are fused to make the classification of long-term survival or short-term survival for each patient. As expected, the evaluation results on comprehensive performance indicate that MAFN achieves better predictive performance than existing methods. Additionally, our method can be extended to the survival prediction of other cancer diseases, providing a new strategy for other diseases prognosis.

SUBMITTER: Guo W

PROVIDER: S-EPMC8417828 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Male breast cancer (MBC) is rare, and most patients are diagnosed at an advanced stage. We aimed to develop a reliable nomogram to predict breast cancer-specific survival (BCSS) for MBC patients, thus helping clinical diagnosis and treatment. Based on data from the Surveillance, Epidemiology, and End Results (SEER) database, 2,451 patients diagnosed with MBC from 2010 to 2015 were selected for this study. They were randomly assigned to either a training cohort (n = 1715) or a validation cohort (n = 736). The Multivariate Cox proportional hazards regression analysis was used to determine the independent prognostic factors, which were then utilized to build a nomogram for predicting 3- and 5-year BCSS. The discrimination and calibration of the new model was evaluated using the Concordance index (C-index) and calibration curves, while its accuracy and benefits were assessed by comparing it to the traditional AJCC staging system using the net reclassification improvement (NRI), the integrated discrimination improvement (IDI), and the decision curve analysis (DCA). Multivariate models revealed that age, AJCC stage, ER status, PR status, and surgery all showed a significant association with BCSS. A nomogram based on these variables was constructed to predict survival in MBC patients. Compared to the AJCC stage, the C-index (training group: 0.840 vs. 0.775, validation group: 0.818 vs. 0.768), the areas under the receiver operating characteristic curve of the training set (3-year AUC: 0.852 vs. 0.778, 5-year AUC: 0.841 vs. 0.774) and the validation set (3-year AUC: 0.778 vs. 0.752, 5-year AUC: 0.852 vs. 0.794), and the calibration plots of this model all exhibited better performance. Additionally, the NRI and IDI confirmed that the nomogram was a great prognosis tool. Finally, the 3- and 5-year DCA curves yielded larger net benefits than the traditional AJCC stage. In conclusion, we have successfully established an effective nomogram to predict BCSS in MBC patients, which can assist clinicians in determining the appropriate therapy strategies for individual male patients.

Project description:BackgroundTriple negative breast cancer (TNBC) is an aggressive and heterogeneous disease. Nomograms predicting outcomes of TNBC are needed for risk management.MethodsNomograms were based on an analysis of 296 non-metastatic TNBC patients treated at Sun Yat-sen Memorial Hospital from 2002 to 2014. The end points were disease-free survival (DFS) and overall survival (OS). Predictive accuracy and discriminative ability were evaluated by concordance index (C-index), area under the curve (AUC) and calibration curve, and compared with the American Joint Committee on Cancer (AJCC) staging system, PREDICT and CancerMath. Models were subjected to bootstrap internal validation and external validation using independent cohorts of 191 patients from the second Xiangya Hospital and Peking University Shenzhen Hospital between 2007 and 2012.ResultsOn multivariable analysis of training cohort, independent prognostic factors were stromal tumor-infiltrating lymphocytes (TILs), tumor size, node status, and Ki67 index, which were then selected into the nomograms. The calibration curves for probability of DFS and OS showed optimal agreement between nomogram prediction and actual observation. The C-index of nomograms was significantly higher than that of the seventh and eighth AJCC staging system for predicting DFS (training: 0.743 vs 0.666 (P = 0.003) and 0.664 (P = 0.024); validation: 0.784 vs 0.632 (P = 0.02) and 0.607 (P = 0.002)) and OS (training: 0.791 vs 0.683 (P = 0.004) and 0.677 (P < 0.001); validation: 0.783 vs 0.656 (P = 0.006) and 0.606 (P = 0.001)). Our nomograms had larger AUCs compared with PREDICT and CancerMath. In addition, the nomograms showed good performance in stratifying different risk groups of patients both in the training and validation cohorts.ConclusionWe have developed novel and practical nomograms that can provide individual prediction of DFS and OS for TNBC based on stromal TILs, tumor size, node status, and Ki67 index. Our nomograms may help clinicians in risk consulting and selection of long term survivors.

Project description:ObjectiveDespite existing prognostic markers, breast cancer prognosis remains a difficult subject due to the complex relationships between many contributing factors and survival. This study seeks to integrate multiple clinicopathological and genomic factors with dimensional reduction across machine learning algorithms to compare survival predictions.MethodsThis is a secondary analysis of the data from a prospective cohort study of female patients with breast cancer enrolled in the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC). We constructed a series of predictive models: ensemble models (Gradient Boosting and Random Forest), support vector machine (SVM), and artificial neural networks (ANN) for 5-year survival based on clinicopathological and gene expression data after K-means clustering with K-nearest-neighbor (KNN) classification. Model performance was evaluated by receiver operating characteristic (ROC) curve, accuracy, and calibration slope (CS). Model stability was assessed over 10 random runs in terms of ROC, accuracy, CS, and variable importance.ResultsThe analytic cohort is composed of 1874 patients with breast cancer. Overall, the median age was 62?years; the 5-year survival rate was 75%. ROC and accuracy were not significantly different between models (ROC and accuracy around 0.67 and 0.72 across models, respectively). However, ensemble methods resulted in better fit (CS) with stable measures of variable importance across 10 random training/validation splits. K-means clustering of gene expression profiles on training data points along with KNN classification of validation data points was a robust method of dimensional reduction. Furthermore, the gene expression cluster with the highest mortality risk was an influential factor in model prediction.ConclusionsUsing machine learning methods to construct predictive models for 5-year survival in patients with breast cancer, we demonstrated discrimination ability across models with new insight into the stability and utility of dimensional reduction on genomic features in breast cancer survival prediction.

Dataset Information

A Multimodal Affinity Fusion Network for Predicting the Survival of Breast Cancer Patients.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets