Multi-label classification and description generation of Pulmonary diseases in Chest X-rays using Deep Learning techniques

Lavania, Madhav Kant (2019) Multi-label classification and description generation of Pulmonary diseases in Chest X-rays using Deep Learning techniques. Masters thesis, Dublin, National College of Ireland.

Preview

PDF (Master of Science) Download (2MB) | Preview

Abstract

The applications of deep learning have broadened their spectrum in the field of medical research. One such field is medical radiography which uses several imaging techniques like CT-Scan, X-Ray to view the internal abnormalities of a human body. Traditional image classifiers could not process these noisy, blurred or unclear X-Ray images, which leads to incorrect results. These image classifiers lack in mimicking the exact understanding of a professional, which comes from rigorous training and hands-on experience. Hence, a novel
solution to x-ray image classification and description generation has been presented. This approach works towards detection and classification of pulmonary diseases like fibrosis as well as generates a medical interpretation of the detected abnormalities. This model is prepared in two parts. The first part works as a multi-label classifier and is based on a pre-trained MobileNet Convolutional Neural Network (CNN), while the second part of the model is designed by combining a pre-trained CNN VGG16 and a Long-short term memory-recurrent neural network (LSTM-RNN) to generate a medical description of the diagnosis. Chest x-ray image dataset from OpenI and NIHCC with their medical labels have been used for training and testing the model. This multi-label classifier and the captioning system performs well with an overall prediction accuracy of about 87% and BLEU score of 0.58.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Computer software T Technology > T Technology (General) > Information Technology > Computer software R Medicine > Healthcare Industry
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Caoimhe Ní Mhaicín
Date Deposited:	14 Oct 2019 10:59
Last Modified:	14 Oct 2019 10:59
URI:	https://norma.ncirl.ie/id/eprint/3867

Actions (login required)

View Item