Surgical Robots in Smart Hospitals

16Improving Lung Cancer Detection Using Hybrid Features and Optimized through VGG-16

A. Ritu^* and A. Eshaan

Department of Computer Science and Engineering, Maharishi Markandeshwar Engineering College, Mullana, India

Abstract

Across the world, cancer is main reason of death in the human life. The medical practioners and the other researchers are fighting against the lung disease cancer. This report is presented in 2019 as per the American Cancer Society. Due to the change in environment, pollution, and other unwanted habits, such as drinking and smoking, which lead to lung disease, early prediction of the disease could be done thru clinical records of a patient. This paper proposes a result that take the syndromes of the patient and identifies or indicates how severe or moderate a disease level is. The user could upload the image or sample of lung disease as an input data to the model or machine learning for predicting and diagnosing the disease in its early stages. In this paper, the current study is focused on the tuberculosis, pneumonia, and at the time of COVID-19–affected samples. Various machine learning algorithms and image samples were computed on the VGG-16 deep learning model, efficient_net, Mobile_net, Dense_net, etc. This current study achieved results as in terms of classification accuracy as efficient net, Dense net, Mobile-net deep learning algorithms.

Keywords: Lung disease, efficient net, dense net, mobile net, VGG-16, deep learning, machine learning

16.1 ...