Cancer Gene Detector Service

Abstract

In medical Science, the well-known technique for cancer detection is biopsy imaging where photographs of the affected area are taken through a CT scanner these images help the doctor to understand where the exact position of the needle is for taking the image, and then using those images the doctor can investigate cancer in patients. The issue with biopsy imaging is that the images are not always the same they can vary from one expert or doctor as compared to the results given by another expert, also there is a lack of quantitative measures while classifying these images as normal or cancerous ones. The method which is used for cancer detection is by using the diagnostic data available on the UCI machine learning repository. Machine Learning models are applied for the prediction of cancer. XGBoost is used for the final cancer prediction which gives an accuracy of 95.61% with normal data and with scaled data. In the future, more improvements are possible if there is more data set there will be increased prediction accuracy and also the possibility to make it available to users, particularly for those in the medical field.