The lungs reflect the health condition of a person, and hence it has been imaged and analysed by diagnosticians for over a century, and it requires knowledge and experience. The human observer’s time and effort could be used productively if the lung image analysis is automated. This is especially true in case of screening of the lung chest x-rays. The lung segmentation is by default the first of a series of steps to analyse and interpret the images using a computer. One of the traditional approaches to segmentation of the lungs has been the use of statistical models, and the other is the rule based approach. This paper proposes a fusion method to segment the lungs on chest x-rays, as this modality of imaging is low cost, easy to operate, and gives first-hand information required for diagnosis. The results that are obtained are fast and promising accuracy has been documented. The entire approach can be extended to any organ on a medical image, or any object of interest in a general segmentation problem.
Athavale, P. A., Kattimani, H. D., & Puttaswamy, P. S. (2018). Segmentation of the lungs from chest X-rays: A simplified computer aided approach. International Journal of Recent Technology and Engineering, 7(4), 220–223.