“
TL;DR
The paper proposes a 6G IoT-enabled framework using MobileNetV3 and a novel optimization algorithm (AOAHG) for medical image classification.
Diagnostics · 5 authors, 7 centres
This summary was generated by AI from a single paper. It has not been reviewed by a clinician and is not clinical advice. Verify against the source before acting on it.
The paper proposes a 6G IoT-enabled framework using MobileNetV3 and a novel optimization algorithm (AOAHG) for medical image classification.
The paper addresses the need for efficient real-time medical image classification in 6G-enabled Internet of Medical Things (IoMT), where massive data generation requires accurate processing for early disease detection, such as skin cancer and retinal disorders. Methods: The framework integrates deep learning with MobileNetV3 for feature extraction and a novel optimization algorithm (AOAHG) that combines Arithmetic Optimization Algorithm (AOA) and Hunger Games Search (HGS) for feature selection. It was evaluated on four datasets: ISIC-2016 and PH2 for skin cancer, white blood cell (WBC) detection, and optical coherence tomography (OCT) classification, using metrics like accuracy, precision, recall, and F1-score. Key Results: The AOAHG method with SVM classifier achieved high accuracy: 87.30% for ISIC, 96.40% for PH2, 88.60% for WBC, and 99.69% for OCT, outperforming other optimization methods. Limitations: The framework is time- and memory-intensive, which could hinder practical deployment. Implications: The approach demonstrates potential for improving medical image classification in IoT environments, but future work is needed to simplify the framework and enhance efficiency, such as through hyperparameter optimization.