Copyright © 2026 Authors retain the copyright of this article. This article 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.

@article{205865,
        author = {Manthan Zunjurke and Pranit Hake and Kaustubh Jagtap and Abhijeet Gaikwad},
        title = {An Integrated Healthcare System for Early Skin Cancer Detection Using CNNs and NLP-Driven Assessment},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {8581-8587},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=205865},
        abstract = {Melanoma is globally recognized as the most dangerous and deadly form of skin cancer, carrying high mortality rates if not detected and treated in its earliest stages. Traditional diagnostic methods, primarily visual inspection followed by surgical biopsies, are highly invasive, time-consuming, and prone to the subjective biases of human dermatological expertise. Furthermore, the global shortage of specialized dermatologists often delays initial screenings, severely impacting patient prog-noses. To address these critical limitations, this paper proposes an automated, non-invasive skin cancer detection system that integrates artificial intelligence with a patient-facing Natural Language Processing (NLP) chatbot. The system provides a comprehensive end-to-end clinical pipeline designed to bridge the gap between initial patient concern and clinical diagnosis. Initially, the NLP agent handles patient interaction, conducts a preliminary symptom assessment inspired by clinical triage questionnaires, and safely guides the user to upload dermoscopic images of concerning lesions. The subsequent image processing pipeline applies adaptive preprocessing for noise removal and contrast enhancement, followed by thresholding segmentation to isolate the region of interest. Statistical and morphological feature extraction is then performed utilizing the Gray Level Co-occurrence Matrix (GLCM) for texture analysis and the ABCD (Asymmetry, Border, Color, Diameter) rules for structural anomalies. Principal Component Analysis (PCA) is applied for optimal feature selection and dimensionality reduction to eliminate redundant data. Finally, a specialized Convolutional Neural Network (CNN) algorithm classifies the lesion as benign or malignant. Experimental results demonstrate that this proposed integrated hybrid methodology achieves a high classification accuracy of 92.1%, proving its viability as a robust pre-screening tool for modern healthcare environments.},
        keywords = {Melanoma, Skin Cancer, Convolutional Neural Networks, GLCM, PCA, Natural Language Processing, Medical Chatbot, Image Segmentation, Teledermatology.},
        month = {June},
        }