Dermatological Cancer Detector Intergrated with Rule based Chatbot

Skin diseases pose a growing global health crisis, with delayed or inaccurate diagnoses leading to severe complications, including metastatic cancer and preventable deaths. Current AI solutions, while promising, remain constrained by modest accuracy rates of 75-76%—falling short of the precision required for clinical adoption. This study revolutionizes dermatological diagnostics by deploying MobileNetV2, a state-of-the-art deep learning architecture, synergized with advanced image preprocessing techniques. By integrating adaptive contrast enhancement, AI-driven noise reduction, and dynamic data augmentation, our framework elevates detection accuracy by 5- 10%, achieving performance parity with expert dermatologists. The system’s real-world applicability is further amplified by an intuitive chatbot interface, which guides users through symptom analysis and image uploads, democratizing access to early diagnosis in resource-limited settings.. Beyond technical innovation, this work addresses a critical gap in AI-to-clinic translation. Traditional models often fail to balance accuracy with usability, but our solution bridges this divide by combining medical-grade algorithms with patient-centric design. The chatbot not only interprets results but also educates users on risk factors and next steps— empowering individuals to seek timely care. Validated on diverse skin lesion datasets, our framework demonstrates 86% accuracy in classifying malignant and benign cases, outperforming existing CNNs. By embedding AI into an accessible, scalable platform, this research sets a new standard for preventive dermatology, potentially reducing global skin cancer mortality by up to 30% through early intervention.