Detection of Lung Tumor using an efficient Quadratic Discriminant Analysis Model

Abstract

Deep Learning has paved a way for finding solutions to numerous problems across the globe. It is one of the techniques being used widely in all the application being used worldwide. Healthcare is a field where various new innovations need to be done in order to diagnosis various new health related issues. In today's modern era, the increase in new medications is required at the most. One such application is the diagnosis of lung tumor. Tumors are the cells that do not grow as normally as other cells in the body and leads to many other side effects and if diagnosed lately could also lead to death. Hence, to diagnosis it very early need to new medications and techniques are required. Numerous types are tumors are present and they tend to occur in various parts of a human body. Sometimes they are very easily identified whereas in some cases they get detected only after a long time. Detection of tumor in very early stages is curable and hence it is very important that it needs to be identified very early. In this research, we present a model that uses the Deep Learning approach in order to identify tumors in the lungs at extremely early stages. Through there are numerous researchers working on the same domain, we have enhanced our model with the use of the Quadratic Discriminant Analysis. To detect the tumor in lung, the model initially detects and segments the tumor cells. Once the cells are segmented, the model also classifies if the tumor cells are malignant or benign. The findings illustrate how the suggested model differs from the conventional approaches in a number of ways. The comparative research was carried out, and the data reveal that the model that was presented is better to the other in terms of the amount of time it needs, clocking in at an average of 2.16 seconds.