MalenoCare: Skin Cancer Detection and Prescriptive Guild lines using AI and ML

Abstract

MalenoCare is an AI-powered healthcare solution designed to assist in the early detection and stage prediction of skin cancer using deep learning techniques. The system leverages convolutional neural networks (CNNs) to analyze dermoscopic images and accurately classify types of skin cancer. It offers a user-friendly interface for both patients and doctors, enabling remote diagnosis, medical report generation, and real-time chat communication. By integrating machine learning with accessible digital tools, MalenoCare aims to support timely intervention, reduce diagnostic delays, and contribute to improved patient outcomes, especially in underserved or remote areas. The project utilizes publicly available datasets such as HAM10000 and ISIC 2019 to train and validate the model, ensuring high accuracy and generalization. Advanced techniques like data augmentation, normalization, and transfer learning are applied to enhance performance. MalenoCare also features a medical history form, secure storage of patient data, and intelligent report generation. Its vision extends beyond diagnosis—aiming to spread awareness, promote preventive care, and align with global health goals by making skin cancer screening accessible and affordable to all

Introduction

A. Background:
Skin cancer is rising globally due to factors like sun exposure and pollution. Early diagnosis is crucial but often inaccessible, especially in rural areas. There’s a pressing need for an AI-powered diagnostic tool to improve early detection.

B. Motivation:
MalenoCare was inspired by the role of AI in healthcare. It aims to close the gap between patients and dermatologists using mobile technology and machine learning, enabling timely diagnosis and preventive care.

C. Problem Statement:
Lack of accessible, affordable diagnostic tools for early skin cancer detection delays treatment and worsens outcomes, particularly in low-resource regions.

D. Objectives:

• Develop a deep learning system for classifying skin cancer types and stages.

• Provide a user-friendly platform for patient-doctor interaction.

• Enable digital health record management and raise awareness.

E. Scope & Limitations:
Focuses on classifying skin cancer using dermoscopic images. Usable via web/mobile but depends on dataset quality. Intended as a support tool—not a replacement for medical professionals.

Literature Review Highlights:

System Design Overview

A. Architecture:
Client-server model with AI-based backend for image analysis, accessible via web/mobile.

B. Modules:

1. Frontend: Built with React (web/mobile); supports image upload, chat, report viewing.

2. Backend: Node.js + Express; manages logic and API calls.

3. ML Model: CNN (e.g., DenseNet, ResNet) trained on ISIC/HAM10000 datasets.

4. Database: MongoDB for user and report data.

5. Chat Module: Real-time communication via WebSocket.

Algorithms & Tools

• Languages/Frameworks: React, Node.js, MongoDB, Python (TensorFlow/Keras)

• Techniques Used:

  • CNN for classification

  • Transfer Learning (DenseNet121, MobileNetV2)

  • Data augmentation & normalization

  • Loss functions: Binary/Categorical Cross Entropy

Results & Discussion

A. Key Findings:

• Achieved 90–92% accuracy in classification.

• Classifies 7 skin cancer types (e.g., Melanoma, BCC).

• Fast prediction: ~1.5s on CPU, <800ms on mobile.

B. Insights:

• High performance: Transfer learning enhanced accuracy with moderate datasets.

• Usability: Fully functional platform with chat and reports.

• Scalability: Easily extendable to include more conditions or integrate with wearables.

Conclusion

MalenoCare presents a promising AI-based solution for the early detection and classification of skin cancer. By integrating machine learning with a user-friendly web and mobile interface, the system enhances accessibility to diagnostic tools, particularly in areas where dermatological services are limited. The project successfully leverages convolutional neural networks and transfer learning techniques to achieve high accuracy, with a prediction speed that supports real-time feedback. The inclusion of supportive features such as patient-doctor chat, medical form submission, and automated report generation transforms MalenoCare from just a diagnostic model into a complete digital healthcare assistant. Though some limitations exist—such as dependence on image quality and data imbalance—the system lays the foundation for scalable and impactful teledermatology solutions. In the future, MalenoCare can be expanded to detect other skin conditions, integrate wearable health devices, and offer multilingual support to reach broader populations. Overall, the project demonstrates how artificial intelligence can contribute meaningfully to public health by providing timely, affordable, and reliable skin cancer screening.

References

[1] S. Kaur and A. Sharma, “Survey of various machine learning approaches for skin disease detection,” in Proc. 7th International Conference on Computing for Sustainable Global Development (INDIACom), New Delhi, India, 2020, pp. 1453–1458. [2] M. Goyal and P. Ramesh, “Development of a CNN-based framework for skin lesion detection,” IEEE Transactions on Medical Imaging, vol. 39, no. 9, pp. 2343-2350, 2020. [3] S. Wu and D. Mo, “Evaluation of skin cancer image classification techniques using deep learning,” International Journal of Computer Applications, vol. 179, no. 27, pp. 1-6, 2018. [4] A. Y. Wang and M. J. Xu, “Automated melanoma classification using deep learning algorithms,” Journal of Dermatological Science, vol. 87, no. 2, pp. 132-138, 2018, doi: 10.1016/j.jds.2018.01.010. [5] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet classification with deep convolutional neural networks,” in Advances in Neural Information Processing Systems 25 (NIPS 2012), Lake Tahoe, NV, USA, 2012, pp. 1097–1105. [6] A. Parikh, L. Wang, and D. Yu, “Applying transfer learning for skin disease identification,” Biomedical Signal Processing and Control, vol. 60, pp. 101979, 2020. [7] P. C. Pandya and R. P. Patel, “Analysis of melanoma detection techniques using convolutional neural networks,” International Journal of Computer Vision, vol. 128, no. 3, pp. 485-495, 2020 [8] Puthal, D., Sahoo, B. P., & Rodrigues, J. J. P. C. (2018). Blockchain as a decentralized security framework for cloud data storage. IEEE Communications Magazine. [9] https://ieeexplore.ieee.org/document/9176321 [10] Chen, L., et al. (2021). \"Compliance and Trust in Data Management: The Role of Blockchain.\" International Journal of Information Management.

Copyright

Copyright © 2025 More Nikhil Sharad, Zinjad Mahesh Sharad, Shaha Shakur Anwar, Prof. Chaudhari N. J.. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.