A Survival Status Classification Model for Osteosarcoma Patients Based on E-CNN-SVM and Multisource Data Fusion

Area of Science:

• Oncology
• Bioinformatics
• Genetics

Background:

• Osteosarcoma patient data presents challenges including small sample sizes, high dimensionality, and imbalanced classes.
• Traditional algorithms often fail to holistically analyze genetic and feature data, leading to poor generalization and classification of survival status.
• Existing methods struggle with interclass imbalance and limited feature extraction from complex genetic datasets.

Purpose of the Study:

• To develop an advanced survival status prediction model for osteosarcoma patients.
• To address limitations of traditional algorithms by employing multisource data fusion and advanced machine learning techniques.
• To improve the accuracy and stability of osteosarcoma patient survival classification.

Main Methods:

• A novel E-CNN-SVM model integrating multisource data fusion was designed.
• Random Forest algorithm was used for dimensionality reduction and fusion of four key gene sequencing datasets.
• Hybrid sampling (SMOTE and TomekLink) addressed data imbalance, followed by feature extraction using an enhanced CNN and classification with SVM.

Main Results:

• The proposed model effectively fuses and processes high-dimensional, imbalanced genetic data from osteosarcoma patients.
• Enhanced feature extraction via CNN and stable classification by SVM significantly improved model performance.
• The E-CNN-SVM model demonstrated superior accuracy in classifying osteosarcoma patient survival status compared to traditional methods.

Conclusions:

• The developed E-CNN-SVM model offers a robust solution for osteosarcoma survival prediction, overcoming limitations of prior approaches.
• Multisource data fusion and advanced machine learning techniques are crucial for accurate classification of complex genetic data.
• This approach holds promise for improving clinical outcomes through more precise survival status prediction in osteosarcoma.