Related Experiment Video
Updated: Jul 19, 2025
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
14.2K
Automatic detection and classification of lung cancer CT scans based on deep learning and ebola optimization search
Tehnan I A Mohamed1,2, Olaide N Oyelade3, Absalom E Ezugwu4
1Department of Computer Science, Faculty of Mathematical and Computer Sciences, University of Gezira, Wad Madani, Sudan.
Plos One
|August 17, 2023
Summary
This study introduces a hybrid deep learning model combining Convolutional Neural Networks (CNN) with the Ebola Optimization Search Algorithm (EOSA) for accurate lung cancer classification from CT scans. The novel EOSA-CNN approach significantly improves early lung cancer detection and survival rates.
Area of Science:
- Medical Imaging Analysis
- Artificial Intelligence in Oncology
- Computational Biology
Background:
- Non-communicable diseases, particularly cancer, are a growing global health concern.
- Accurate and early lung cancer diagnosis is critical for improving patient survival rates and treatment efficacy.
- Existing Convolutional Neural Network (CNN) models face challenges in optimal parameter selection for lung cancer classification from CT scans.
Purpose of the Study:
- To develop a hybrid metaheuristic and CNN algorithm for precise lung cancer classification using CT images.
- To address the limitations of traditional CNNs in selecting optimal weights and biases for classification tasks.
- To enhance the accuracy and reliability of lung cancer detection through an optimized deep learning approach.
Main Methods:
- A CNN architecture was designed and its solution vector computed.
- The Ebola Optimization Search Algorithm (EOSA) was employed to optimize the CNN's weights and biases.
- The hybrid EOSA-CNN model was trained and validated on the IQ-OTH/NCCD lung cancer dataset.
Main Results:
- The EOSA-CNN model achieved a high classification accuracy of 0.9321 on the lung cancer dataset.
- The hybrid model demonstrated superior performance compared to other methods like GA-CNN, LCBO-CNN, and classical CNN.
- EOSA-CNN reported specificities of 0.7941, 0.97951, 0.9328 and sensitivities of 0.9038, 0.13333, 0.9071 for normal, benign, and malignant cases, respectively.
Conclusions:
- The proposed hybrid EOSA-CNN algorithm offers a robust and effective solution for lung cancer classification from CT images.
- This approach significantly improves early detection capabilities, potentially leading to better patient outcomes.
- The study highlights the potential of metaheuristic optimization in enhancing deep learning models for medical diagnostics.
