Algorithm for quantification of pulmonary sequelae in chest X-ray

• Published in: Physica medica 2016-09, Vol.32, p.331-331
• Main Authors: de Oliveira, Marcela, Giacomini, Guilherme, Alvarez, Matheus, Câmara Pereira, Paulo Marques, Ribeiro, Sergio Marrone, de Pina, Diana Rodrigues
• Format: Article
• Language: English
• Subjects: Radiology
• ISSN: 1120-1797; 1724-191X
• DOI: 10.1016/j.ejmp.2016.07.239

Introduction Tuberculosis (TB) is one of the oldest infectious diseases in the world. Even after effective treatment, TB leaves pulmonary sequela that compromises patients’ life quality. Sequelae evaluations are usually performed subjectively through chest X-ray radiographs. While new treatments for TB are being developed, the necessary tools for monitoring patients in treatment and quantification of the sequelae remain limited. Purpose The main purpose was to objectively quantify the pulmonary impairment pre- and post-treatment of patients with pulmonary TB through an computational algorithm. Materials and methods We used 20 X-ray exams, pre- and post-treatment of 10 patients with TB. Lung area and affected regions were manually segmented in both postero-anterior (PA) and profile projections. We selected regions-of-interest in both affected and normal regions and obtained by the Signal Difference to Noise Ratio (SDNR). Values of SDNR were related to the relative thickness of lung affected in CT scans. Thus the algorithm used this relationship to estimate the relative thickness of pulmonary impairment from X-ray exams PA projection. Results We observed a mean pulmonary impairment of 5.21% (±3.35) before treatment and 1.26% (±0.72) after treatment. This shows a reduction of 72.54% between pre- and post-treatment. Conclusion The computational algorithm allows the quantification of pulmonary impairment through chest X-ray radiographs. Detection and quantification aided by computer systems is of great importance for reliable assessment of pulmonary involvement, assisting radiologists in the diagnosis. Future studies will help the choice of the correct treatment for TB patients. Disclosure The authors declare that there is no conflict of interest.