AarogyaShree: A Smartphone-First Framework Using Lightweight CNN/DCNN Models and Symptom Questionnaires for Cervical and Breast Cancer Screening

This record documents the research work titled “AarogyaShree,” which investigates a smartphone-based, AI-assisted screening framework for cervical and breast cancer detection in low-resource environments. The study focuses on addressing practical deployment challenges observed in existing AI-based screening approaches, including dependence on laboratory-grade imaging equipment, limited generalizability across devices, and poor accessibility in rural and underserved communities. AarogyaShree proposes a mobile-first design that integrates lightweight convolutional and deep convolutional neural network (CNN/DCNN) models with robust image preprocessing pipelines and a structured, local-language symptom questionnaire. The framework emphasizes real-world usability by supporting smartphone image capture, low-bandwidth or offline operation, multilingual interfaces, and workflows tailored for community health workers. Image-based predictions are combined with questionnaire-driven clinical context to generate a more reliable and interpretable screening outcome. This repository upload establishes the academic precedence and authorship of the AarogyaShree research. The work is shared for scholarly reference, learning, and non-commercial research purposes only. Implementation code, datasets, and deployment-level details are intentionally excluded at this stage.