International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 04 | Apr 2023
p-ISSN: 2395-0072
www.irjet.net
X-Ray Disease Identifier Rohan Darji1, Siddhant Chavan2, Harsh Chauhan3, Parth Khanolkar4, Amruta Sankhe5 1Rohan Darji, Dept of Information Technology, Atharva College of Engineering
2Siddhant Chavan, Dept of Information Technology, Atharva College of Engineering 3Harsh Chauhan, Dept of Information Technology, Atharva College of Engineering
4Parth Khanolkar, Dept of Information Technology, Atharva College of Engineering 5Prof. Amruta Sankhe, Dept of Information Technology, Atharva College of Engineering, Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------In general, for the prediction of diseases, we try to use either Abstract - Numerous lung diseases are frequently been
X-ray, CT, or MRI scan techniques for taking decisions on the appropriate disease but with the help of deep learning methodologies there has been ease for all the doctors, radiologists as well as other researches by giving them a direction for the detection of lung diseases. With this advancement in technology and the use of AI, successful research and viable results help to save countless lives by estimating diseases in remote areas without the use of heavy machinery. Thus, a system capable of predicting lung diseases and diagnosing them with good accuracy will reduce the load on all doctors by helping them to work more effectively and smoothly.
diagnosed globally and the global pandemic COVID-19 acted in addition to affecting the lifestyle of the people. It is essential to provide a well-timed diagnosis for diseases like Emphysema, Effusion, Pneumonia, Edema, etc., for this various image-processing models are developed. One of the promising research areas in the medicinal field is Medical Image analysis which delivers quick and accurate results along with providing decisions with their appropriate diagnosis.
Inspired by recent research on image analysis that correlates the findings in chest x-ray images, we have developed an approach that uses the existing deep learning model – the VGG19 classification model to process the X-ray images and diagnose them according to the respective disease and provide basic knowledge about them. As the implementation tool, Jupyter notebook is used and this model has the NIH (National Institute of Health) X-ray image dataset. Experiments have shown that the classification method applied in this system is able to detect the findings in the diseases more effectively and with an accuracy of above 60% for most of the diseases.
2. LITERATURE REVIEW 2.1 CNN-based Deep Learning Model for Chest X-ray Health Classification Using TensorFlow (2020)[1] The article discusses the use of machine vision, image processing techniques, and deep learning algorithms as diagnostic tools for respiratory ailments, specifically pneumonia. These tools are more accurate, portable, and cost-effective, making them efficient for physicians to use. Artificial intelligence and machine learning are considered the most accurate methodologies for identifying and classifying health issues, including pneumonia. The study focuses on training a system to distinguish between healthy and diseased lungs based on a set of parameters such as the data set's size and the model and neural network attributes. The MobileNetV2 pre-trained neural network model is used as a backbone for feature extraction, enabling accurate results in object detection and semantic segmentation without prior features. The convolutional neural network was trained and analysed to classify lungs based on the output labels: NORMAL and PNEUMONIA, achieving accurate results of over 90% during testing. The study concludes that the MobileNetV2 convolutional neural network model offers accurate results and several advantages, including high accuracy, even without prior features.
Key Words: VGG19, Deep Learning, X-Ray images, Classification Methods.
1. INTRODUCTION Alterations in the environment, lifestyle, and other factors are causing a rapid increase in the effect of diseases on human health. The countries where millions of people are facing poverty and air pollution are especially endangered with the risk of getting several lung diseases. According to the estimation of WHO, over 4 million premature deaths have occurred annually from lung diseases including asthma, pneumonia, and Emphysema. Therefore, it is essential to implement diagnostic systems that will help in detecting lung diseases and providing a respective diagnosis. During the global pandemic, pneumonia i.e. - breathing and lung problem were at their peak. It became crucial for all to early detect the findings in the lungs and for this purpose machine learning and deep learning played a vital role which helped millions of people worldwide.
© 2023, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 1214