Prediction of Lungs Cancer Using Machine Learning

EasyChair Preprint 3076, version 2

Abstract

There is lot of progress made in the field of treatment of lung cancer in the last years (adjuvant chemotherapy, radio therapy, individualized therapy). Nonetheless, lung cancer is still remained the threat of society and cause of death of thousands of people in all over the world. This paper is all about detection of lungs cancer. Here Computer Tomography (CT) images are used to detect lungs cancer. There are several algorithms are used to detect Lungs cancer accurately. Here unsharp masking filter is use to filtering the image. Adaptive Canny edge detection algorithm is used to detect the edges and cancer affected areas. Neural network is used to classify the features and predict the probability of lung cancer. K-Nearest Neighbors is used to segment the cancer from lungs. Finally achieved the classification accuracy near about 99.5% by using Bayesian Regularization Neural Network (BRNN) and performance measure by Mean square error (MSE) noted as 0.0166.

Keyphrases: Lungs Cancer, Segmentation, k-nearest neighbors

@booklet{EasyChair:3076,
  author    = {Prasanta Das and Biplab Das and Himadri Sekhar Dutta},
  title     = {Prediction of Lungs Cancer Using Machine Learning},
  howpublished = {EasyChair Preprint 3076},
  year      = {EasyChair, 2020}}