Open AccessEdge AI System for Pneumonia and Lung Cancer Detection
Author(s) -
Venkata Tulasiramu Ponnada,
Dr.S.V.Naga Srinivasu
Publication year - 2019
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.i8584.078919
Subject(s) - convolutional neural network , computer science , artificial intelligence , lung cancer , convolution (computer science) , pneumonia , edge detection , deep learning , image (mathematics) , computer vision , enhanced data rates for gsm evolution , artificial neural network , image processing , medicine , pathology
The fabulous success of machine learning algorithms for medical image analysis lead the computer aided disease detection systems for medical diagnosis. This paper presents a system “Edge AI System for Pneumonia and Lung Cancer Detection (EASPLD)". The EASPLD is a unique and one stop solution to detect Pneumonia and Lung cancer. The system is used as a clinician decision supporting system or user system to detect the pneumonia and lung cancer. EASPLD uses deep learning techniques such as convolution neural network (CNN). The proposed solution uses medical image analysis techniques and or methods to develop the system. EASPLD proposed a CNN (EASPLD-CNN) EASPLD-CNN uses seven convolution layers and one max pool layer with 3x3 and 5x5 convolutions, whereas other proposed solutions uses either 3X3 or 5X5 convolutions. In our paper, we used the lung X-Ray and CT scan images from LIDC-IDRI[1] and Mendeley[2]. EASPLD consists of Input image capturing system (IICS), Image enhancement system (IES), EASPLD engine and Results reporting engine (RRE. The EASPLD system output is notified to the end user, i.e. clinician and or a patient in the form of visual, text and email notification.