Oral cancer detection model in distributed cloud environment via optimized ensemble technique. (March 2023)

• Record Type: Journal Article
• Authors: Shetty, Savita; Patil, Annapurna P.
• Is Part Of: Biomedical signal processing and control. Volume 81(2023)
• Publication Date: 2023-03
• Subjects: Cloud computing -- Distributed cloud -- Oral cancer detection -- Ensemble classifier -- AEULM; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.104311
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: The highlights of the article are given below for your kind perusal. Kindly, consider and forward my article for further processes. To select the most reliable features among the extracted features with the suggested Improved Linear Discriminant Analysis (ILDA). To construct an ensemble of classifier with Support Vector Machine (SVM), Convolutional Neural Network (CNN), and Multi-layer Perceptron (MLP) for precise oral cancer classification. The weight of CNN is fine-tuned using a new hybrid optimization model- Aquila Exploration Updated with Local Movement (AEULM). Abstract: A new Two-tier m-healthcare oral cancer detection framework is constructed in a distributed cloud environment in this study. The following is the detection model: The supplied data is pre-processed to reduce noise as well as any undesired artifacts. The region of interest is separated from the backdrop via image segmentation. The segmentation in this research is done utilizing the Region Growing Technique. Textural features, Graph features, and Morphological features are also retrieved. A suggested Improved Linear Discriminant Analysis (ILDA) is used to pick the features in this study. For cancer classification, the selected features are exposed to ensemble classifiers (EC). In EC, stage 1 includes the Support Vector Machine (SVM) and Multi-layer Perceptron (MLP)) modeled for disease classification. The stage 2 phase includes the optimized CNN, which makes the final decisions regarding the presence/ absence of oral cancer. The optimized CNN is trained with outcomes acquired from SVM and the input for optimal CNN is MLP, which will provide the final detected outcomes. Since, the CNN is the final decision makes, its weight of it is fine-tuned using a new hybrid optimization model- Aquila Exploration Updated with Local Movement (AEULM), this ensures enhanced detection accuracy. The traditional Aquila Optimizer (AO) and the proposed hybrid optimization method are conceptually combined to form Wildebeest Herd Optimization (WHO). To confirm the effectiveness of the predicted model, a comparative assessment is completed.