Machine Learning approach to classify and predict different Osteosarcoma types
Sanket Mahore, Kalyani Bhole, Shashikant Rathod
Abstract
Family physicians rarely see a malignant bone cancer because it is hard to find, and most of the time, bone cancer is benign. It is very time-consuming and complicated for the pathologist to classify Osteosarcoma histopathological images. Typically Osteosarcoma classifies into viable, Non-viable, and Non-tumor classes, but intra-class variation and inter-class similarity are complex tasks. This paper used the Random Forest(RF) machine learning algorithm, which efficiently and accurately classifies Osteosarcoma into Viable, Non-viable, and Non-tumor classes. The Random Forest method gives a classification accuracy of 92.40%, a sensitivity of 85.44%, and specificity 93.38% with AUC=0.95.
Topics & Concepts
OsteosarcomaRandom forestArtificial intelligenceComputer scienceBone cancerClass (philosophy)Machine learningCancerSimilarity (geometry)Pattern recognition (psychology)MedicinePathologyInternal medicineImage (mathematics)AI in cancer detectionDigital Imaging for Blood DiseasesRadiomics and Machine Learning in Medical Imaging