Multi-Model System for Cancer Detection and Classification from Histopathological and Scan Images

Abhishek Korwate

Abstract: This paper presents a comprehensive web-based cancer detection system utilizing multiple artificial intelligence models for automated medical image analysis. The system integrates Convolutional Neural Networks (CNN), traditional machine learning algorithms, and ensemble methods to classify ten distinct cancer types from histopathological images. Implemented as a Flask web application, the system achieves multi-model consensus through parallel prediction pipelines, providing healthcare professionals with confidence-scored diagnoses and staging information. The architecture employs a feature extraction pipeline using pre-trained CNN layers, feeding into Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Logistic Regression, and ensemble voting mechanisms. Experimental results demonstrate that the multi-model approach improves diagnostic accuracy and provides robust predictions across diverse cancer types including blood, brain, breast, kidney, lung, and skin cancers, with both benign and malignant classifications.

Keywords: cancer detection system, artificial intelligence models, medical image analysis, multi model classification, histopathology images

How to Cite?: Abhishek Korwate, "Multi-Model System for Cancer Detection and Classification from Histopathological and Scan Images", Volume 14 Issue 11, November 2025, International Journal of Science and Research (IJSR), Pages: 1822-1825, https://www.ijsr.net/getabstract.php?paperid=SR251125122625, DOI: https://dx.doi.org/10.21275/SR251125122625