An Ensembled Framework for Human Breast Cancer Survivability Prediction Using Deep Learning.

Area of Science:

• Oncology
• Computational Biology
• Bioinformatics

Background:

• Breast cancer is a leading cause of death, necessitating accurate survival predictions for effective treatment planning.
• Timely prognosis aids physicians in making critical treatment decisions for breast cancer patients.
• There is a significant need for efficient and rapid computational models for breast cancer prognosis.

Purpose of the Study:

• To propose an ensemble model for breast cancer survivability prediction (EBCSP) utilizing multi-modal data.
• To develop an efficient and rapid computational tool for breast cancer prognosis.
• To improve the accuracy of breast cancer survival predictions.

Main Methods:

• An ensemble model (EBCSP) was developed by stacking the outputs of multiple neural networks.
• A convolutional neural network (CNN) was used for clinical data, a deep neural network (DNN) for copy number variations (CNV), and a long short-term memory (LSTM) architecture for gene expression data.
• A random forest method was employed for binary classification of survivability (long-term > 5 years vs. short-term < 5 years).

Main Results:

• The EBCSP model integrates multi-modal data, including clinical information, CNV, and gene expression, for enhanced prediction.
• The ensemble approach effectively handles complex, multi-dimensional biological data.
• The EBCSP model demonstrated superior performance compared to models using single data modalities and existing benchmarks.

Conclusions:

• The proposed EBCSP model offers an effective computational approach for breast cancer survivability prediction.
• Utilizing multi-modal data within an ensemble framework significantly improves prediction accuracy.
• The EBCSP model provides a valuable tool for aiding clinical decision-making in breast cancer treatment.