Related Experiment Video
Updated: Nov 23, 2025
07:15
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
7.2K
Optimal Deep Belief Network with Opposition based Pity Beetle Algorithm for Lung Cancer Classification: A DBNOPBA
Mrs M Mary Adline Priya1, Dr S Joseph Jawhar2, Dr J Merry Geisa3
1Department of Information and Communication Engineering, Anna University, Chennai, Tamil Nadu, India.
Computer Methods and Programs in Biomedicine
|December 31, 2020
Summary
This study introduces an Opposition-Based Pity Beetle Algorithm (OPBA) for classifying metastatic cancer cells using CT images. The novel method achieves high accuracy, improving cancer detection and diagnosis.
Area of Science:
- Medical imaging analysis
- Computational biology
- Machine learning for healthcare
Background:
- Classifying metastatic cancer cells is crucial for effective treatment.
- Current methods often struggle with subtle textural differences in medical images.
- Automated analysis of Computed Tomography (CT) images can improve diagnostic accuracy.
Purpose of the Study:
- To develop and evaluate a novel meta-heuristic algorithm for classifying low- and high-metastatic cancer cells.
- To enhance the accuracy of cancer cell classification using textural features from CT images.
- To improve the efficiency of cancer detection by optimizing feature extraction and classification models.
Main Methods:
- Extraction of Gray-Level Co-occurrence Matrix (GLCM) and Local Binary Patterns (LBP) for textural feature analysis.
- Application of wavelet filters and a four-layer deep belief network (DBN) for feature extraction and dimensionality reduction.
- Development of the Opposition-Based Pity Beetle Algorithm (OPBA), incorporating Opposition Based Teaching (OBL), Position Clamping (PC), and Cauchy Mutation (CM) to optimize classification.
- Utilizing Local Tangent Space Alignment (LTSA) for feature compression.
Main Results:
- The proposed methodology achieved a maximum sensitivity of 96.86%, precision of 97.24%, and accuracy of 97.92% on the LIDC-IDRI CT dataset.
- The OPBA algorithm demonstrated superior performance compared to state-of-the-art methods.
- OBL, PC, and CM effectively helped the OPBA escape local optima and accelerated convergence.
Conclusions:
- The developed OPBA method offers a highly accurate and efficient approach for classifying metastatic cancer cells from CT images.
- The integration of advanced feature extraction techniques and meta-heuristic optimization significantly enhances diagnostic capabilities.
- This research provides a promising tool for improving early cancer detection and patient outcomes.