Automated estimation of progression of interstitial lung disease in CT images

• Published in: Medical physics (Lancaster) Vol. 37; no. 1; pp. 63 - 73
• Main Authors: Arzhaeva, Yulia, Prokop, Mathias, Murphy, Keelin, van Rikxoort, Eva M., de Jong, Pim A., Gietema, Hester A., Viergever, Max A., van Ginneken, Bram
• Format: Journal Article
• Language: English
• Published: United States American Association of Physicists in Medicine 01-01-2010
• Subjects: Adult; Algorithms; Artificial Intelligence; Computed tomography; Computer aided diagnosis; Computer systems; computerised tomography; diagnostic radiography; Diseases; dissimilarity; feature extraction; Female; filtering theory; Humans; image classification; image registration; image texture; interstitial lung disease; Laser Doppler velocimetry; lung; Lung Diseases, Interstitial - diagnostic imaging; Lungs; Male; medical image processing; Medical image segmentation; Medical imaging; Middle Aged; Opacity; Pattern Recognition, Automated - methods; Probability theory, stochastic processes, and statistics; Radiographic Image Enhancement - methods; Radiographic Image Interpretation, Computer-Assisted - methods; Radiologists; Registration; Reproducibility of Results; Sensitivity and Specificity; serial CT evaluation; statistical analysis; Subtraction Technique; texture analysis; Tomography, X-Ray Computed - methods; computed tomography; serial CT evaluation; computer-aided diagnosis; dissimilarity; interstitial lung disease; texture analysis
• ISSN: 0094-2405; 2473-4209
• DOI: 10.1118/1.3264662

Purpose:
A system is presented for automated estimation of progression of interstitial lung disease in serial thoracic CT scans.
Methods:
The system compares corresponding 2D axial sections from baseline and follow-up scans and concludes whether this pair of sections represents regression, progression, or unchanged disease status. The correspondence between serial CT scans is achieved by intrapatient volumetric image registration. The system classification function is trained with two different feature sets. Features in the first set represent the intensity distribution of a difference image between the baseline and follow-up CT sections. Features in the second set represent dissimilarities computed between the baseline and follow-up images filtered with a bank of general purpose texture filters.
Results:
In an experiment on 74 scan pairs, the system classification accuracies were 76.1% and 79.5% for the two feature sets, respectively, while the accuracies of two observer radiologist were 78.5% and 82%, respectively. The agreements of the system with the reference standard, measured by weighted kappa statistics, were 0.611 and 0.683 for the two feature sets, respectively.
Conclusions:
The system employing the second feature set showed good agreement with the reference standard, and its accuracy approached that of two radiologists.