Recent Papers

The project titled Deep Learning-Based Nail Disease ClassificationUsing MobileNetV3-Capsule Architecture presents an intelligentsystem for identifying nail diseases through image analysis. Nailsoften reflect underlying health conditions, and early diagnosis of naildisorders like onychomycosis and psoriasis can aid timely treatment.This work proposes a hybrid deep learning model that combinesMobileNetV3 for efficient feature extraction and a Capsule Networkfor robust classification. A curated dataset of nail images, categorizedinto healthy, onychomycosis, and psoriasis classes, was used. Imageswere pre-processed using standard techniques such as resizing,normalization, and augmentation to ensure consistency and improvegeneralization. MobileNetV3, known for its lightweight design andspeed, was utilized to extract meaningful features from the images.The features were passed to the Capsule Network, which capturedspatial relationships more effectively than traditional CNNs. Theproposed model achieved an impressive classification accuracy of99.4%, demonstrating high precision in distinguishing between nailconditions. This accuracy highlights the effectiveness of the integratedarchitecture, particularly for medical image-based diagnosis. Thesystem is designed to be efficient and scalable making it suitable fordeployment on mobile or low-resource devices, especially in remotehealthcare settings. This project underscores the potential ofcombining lightweight deep learning models with advanced classifiersto create accessible, non- invasive diagnostic tools that support earlydisease detection and improve healthcare outcomes. Invasivediagnostic tools that support early disease detection and improvehealthcare outcomes.

Copyright © 2025 Babin Nivya P. This is an open access article distributed under the Creative Commons Attribution License.