Machine Learning in Healthcare : Identifying Pneumonia with Artificial Intelligence
Syed, Maria (2018)
Syed, Maria
Metropolia Ammattikorkeakoulu
2018
All rights reserved
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2018101315963
https://urn.fi/URN:NBN:fi:amk-2018101315963
Tiivistelmä
Artificial Intelligence has become popular in a multitude of everyday tasks, from self driving cars to predicting stock market fluctuations, however, it has most notably impacted the healthcare field where its usefulness is especially significant due to the nature of healthcare centers recording large volumes of health data.
Therefore, the applications of Artificial Intelligence in healthcare are endless, for instance, future health problems can be predicted and treatment plans can be devised instantly. However, this thesis aims to explore one of these applications, i.e. disease diagnosis, particularly the diagnosis of pneumonia.
Pneumonia is the infection of the lungs that affects about 400 million worldwide causing various health problems such as difficulty in breathing and uncontrollable coughing. This disease is typically diagnosed with the help of a chest X-ray along with other physical examinations. This thesis aims to build a machine learning model that is able to leverage computer vision to analyse chest X-ray images and identify features that indicate the presence of pneumonia. It aims to speed up the time taken by doctors to diagnose pneumonia and start a suitable treatment plan immediately, helping save lives.
The thesis will uncover the fundamentals of deep learning and computer vision to aid the understanding of how the practical work was done and at the same time explores different methods of developing a functional convolutional neural network. The outcomes of each model developed will be analysed and a conclusion drafted based on the analysis.
Therefore, the applications of Artificial Intelligence in healthcare are endless, for instance, future health problems can be predicted and treatment plans can be devised instantly. However, this thesis aims to explore one of these applications, i.e. disease diagnosis, particularly the diagnosis of pneumonia.
Pneumonia is the infection of the lungs that affects about 400 million worldwide causing various health problems such as difficulty in breathing and uncontrollable coughing. This disease is typically diagnosed with the help of a chest X-ray along with other physical examinations. This thesis aims to build a machine learning model that is able to leverage computer vision to analyse chest X-ray images and identify features that indicate the presence of pneumonia. It aims to speed up the time taken by doctors to diagnose pneumonia and start a suitable treatment plan immediately, helping save lives.
The thesis will uncover the fundamentals of deep learning and computer vision to aid the understanding of how the practical work was done and at the same time explores different methods of developing a functional convolutional neural network. The outcomes of each model developed will be analysed and a conclusion drafted based on the analysis.