Artificial Immune Based Medical Image Classification By Optimized Features And Ensemble Tree Learning Model

Skin cancer poses a serious threat to human health worldwide, affecting a large number of individuals. Therefore, early detection and accurate diagnosis are crucial, and dermoscopic imaging plays a vital role in identifying these conditions at an initial stage. This paper has proposed a model that classify the skin medical image into healthy and non-healthy image. Whole model is divide into two module first improve the input image data quality by removing noise and identifying the effecting portion of image that impact more. Further second module extract the histogram and CCM features from the image to train the Ensemble Tree model. Experiment is done on real dataset of skin cancer images. Result shows that proposed MICAIML (Medical Image Classification using Artificial Immune Machine Learning Model) has increase the detection accuracy by % as compared to existing model.