Detecting Lung Cancer Diseases by Using Federated and Machine Learning Model | ||
| IJCI. International Journal of Computers and Information | ||
| Articles in Press, Accepted Manuscript, Available Online from 23 February 2025 | ||
| Document Type: Original Article | ||
| DOI: 10.21608/ijci.2025.349287.1187 | ||
| Authors | ||
| karim shamekh1; Arabi Keshk* 2; Mohamed Sakr3; Gamal Farouk Elhady4 | ||
| 1Computer science , High Technology Institute, 10 of Ramadan | ||
| 2President of Delta Technological University | ||
| 3Department of Computer Science, Faculty of Computers and Information, Menoufia University, Menoufia 32511, Egypt | ||
| 4Faculty of Computers and Information, Menoufia University,Shebin Elkom 32511, Egypt | ||
| Abstract | ||
| Detecting and categorizing lung cancer at an early stage is vital for enhancing a patient's prognosis. Lung cancer continues to be a major public health concern, contributing significantly to the global tally of cancer-related fatalities. While advancements in imaging and sequencing technologies have led to considerable progress in lung cancer research, the sheer volume of data produced by these techniques presents challenges in terms of efficient processing and analysis. However, the human capacity to efficiently process and make full use of the vast volumes of data generated by these developments remains limited. Most researchers applied Machine Learning (ML) algorithms to detect and classify lung cancer from medical images. However, traditional classification methods depend on single machine learning models, which are limited by the quantity and quality of data stored on a central computing server. In this research, we present a novel approach for the detection of lung cancer using an ensemble Federated Learning (FL) methodology. This method works with the K-Nearest Neighbors (KNN), a Decision Tree model and Support Vector Machine (SVM) machine learning classifier, which was trained on a variety of datasets to improve precision and generality. Furthermore, the "federated learning" technique enables the use of remote data while maintaining data confidentiality and security. The effectiveness of the proposed approach using the Kaggle lung cancer dataset of computerized tomography (CT) scans. Our findings indicate an impressive 97.84% accuracy in lung cancer classification. | ||
| Keywords | ||
| Machine Learning (ML); Federated Learning (FL); k-Nearest Neighbors (KNN); Support Vector Machine (SVM); Computerized Tomography (CT) | ||
|
Statistics Article View: 54 |
||
