Hybrid Rider Optimization with Deep Learning Driven Biomedical Liver Cancer Detection and Classification
Mesfer Al Duhayyim1, Hanan Abdullah Mengash2, Radwa Marzouk2
1Department of Computer Science, College of Sciences and Humanities-Aflaj, Prince Sattam Bin Abdulaziz University, Al-Kharj, Saudi Arabia.
Computational Intelligence and Neuroscience
|July 11, 2022
Summary
This study introduces a novel Hybrid Rider Optimization with Deep Learning Driven Biomedical Liver Cancer Detection and Classification (HRO-DLBLCC) model for accurate liver cancer diagnosis. The HRO-DLBLCC model significantly improves early detection and classification of liver cancer from medical images.
Area of Science:
- Biomedical Engineering
- Computational Intelligence
- Medical Image Analysis
Background:
- Liver cancer is a leading cause of mortality worldwide, necessitating improved diagnostic methods.
- Manual identification of cancerous tissue is labor-intensive and prone to errors.
- Computer-aided diagnosis (CAD) systems, particularly those using deep learning, offer potential for accurate and efficient liver cancer detection.
Purpose of the Study:
- To introduce and evaluate a novel Hybrid Rider Optimization with Deep Learning Driven Biomedical Liver Cancer Detection and Classification (HRO-DLBLCC) model.
- To enhance the accuracy and efficiency of liver cancer identification and classification in medical images.
Main Methods:
- The HRO-DLBLCC model utilizes a two-stage preprocessing pipeline: Gabor filtering (GF) for noise reduction and watershed transform for image segmentation.
- A DenseNet-201 architecture with a NAdam optimizer serves as the feature extractor.
- The Hybrid Rider Optimization (HRO) algorithm is employed to optimize hyperparameters for a recurrent neural network-long short-term memory (RNN-LSTM) model for final classification.
Main Results:
- The proposed HRO-DLBLCC model demonstrated promising performance in identifying and classifying liver cancer.
- Experimental validation showed superior results compared to existing state-of-the-art models.
- The model effectively extracts relevant features and achieves accurate classification.
Conclusions:
- The HRO-DLBLCC model offers a robust and effective approach for computer-aided liver cancer diagnosis.
- The integration of deep learning and optimization algorithms significantly advances the field of biomedical image analysis for cancer detection.
- This model holds potential for improving early diagnosis rates and patient survival in liver cancer cases.
