Automatic Tuberculosis Detection Using Chest X-ray Analysis With Position Enhanced Structural Information

Document Type

Oral Presentation

Campus where you would like to present

Ellensburg

Event Website

https://digitalcommons.cwu.edu/source

Start Date

18-5-2020

Abstract

Tuberculosis is a disease responsible for the deaths of more than one million people worldwide every year. Even though the disease is preventable and curable, it remains a major threat to the humanity that needs to be taken care of. More developed countries use advanced techniques such as culture methods or sputum smear microscopy to diagnose the disease. However, since those approaches are rather expensive, they are not commonly used in poor regions of the globe such as East Asia, Africa and Bangladesh. Instead the well know and very affordable chest x-ray interpretation by radiologists is the technique employed in those places. Some of the major issues with this approach are: i) is a tedious task that requires experienced medical personnel, and ii) is performed manually which can be very painful when done for a large population. In order to accelerate the interpretation process and reduce the dependence on qualified radiologists -which is scarce it those countries, many software solutions evolved over the last few years considering computer vision, artificial intelligence and machine learning. The issue with these solutions is that they are either not reliable enough or they are rather complicated. Therefore, we propose a fully automatic software solution that uses only machine learning and image processing to analyse and detect anomalies related to Tuberculosis in Chest x-rays images. Our system has been tested on two benchmark data collections -Montgomery and Shenzhen, and produced state-of-the-art results reaching up to 97% in accuracy. College of the Sciences Presentation Award Winner.

Faculty Mentor(s)

Szilard Vajda

Department/Program

Computer Sciences

Additional Mentoring Department

https://cwu.studentopportunitycenter.com/2020/04/automatic-tuberculosis-detection-using-chest-x-ray-analysis-with-position-enhanced-structural-information/

Share

COinS
 
May 18th, 12:00 PM

Automatic Tuberculosis Detection Using Chest X-ray Analysis With Position Enhanced Structural Information

Ellensburg

Tuberculosis is a disease responsible for the deaths of more than one million people worldwide every year. Even though the disease is preventable and curable, it remains a major threat to the humanity that needs to be taken care of. More developed countries use advanced techniques such as culture methods or sputum smear microscopy to diagnose the disease. However, since those approaches are rather expensive, they are not commonly used in poor regions of the globe such as East Asia, Africa and Bangladesh. Instead the well know and very affordable chest x-ray interpretation by radiologists is the technique employed in those places. Some of the major issues with this approach are: i) is a tedious task that requires experienced medical personnel, and ii) is performed manually which can be very painful when done for a large population. In order to accelerate the interpretation process and reduce the dependence on qualified radiologists -which is scarce it those countries, many software solutions evolved over the last few years considering computer vision, artificial intelligence and machine learning. The issue with these solutions is that they are either not reliable enough or they are rather complicated. Therefore, we propose a fully automatic software solution that uses only machine learning and image processing to analyse and detect anomalies related to Tuberculosis in Chest x-rays images. Our system has been tested on two benchmark data collections -Montgomery and Shenzhen, and produced state-of-the-art results reaching up to 97% in accuracy. College of the Sciences Presentation Award Winner.

https://digitalcommons.cwu.edu/source/2020/COTS/53