The Computer-Aided Lung Informatics for Pathology Evaluation and Rating (CALIPER) software performed well in evaluating pulmonary fibrosis for patients with idiopathic pulmonary fibrosis (IPF) undergoing antifibrotic treatment.
The CALIPER software was developed at the Mayo Clinic in the US and presented with a decrease in pulmonary function as the CALIPER-derived abnormalities increased. Also, the CALIPER performance was not altered by computed tomography (CT) doses.
“These findings bolster the applicability of CALIPER as a routine adjunct to conventional pulmonary function surrogates for evaluation of fibrotic changes in patients undergoing anti-fibrotic treatment,” the authors said.
Koo et al found an inverse correlation between pulmonary function testing (PFT) parameters, including percent predicted forced vital capacity and forced expiratory volume in 1 second. Percent CALIPER features, such as ground-glass, reticular, interstitial lung disease (ILD), vessel-related structures (VRS), and combined ILD and VRS correlations were also found.
Read more about IPF testing
These CALIPER features were also associated with oxygen usage. Moreover, the extent of honeycomb (HC), another CALIPER feature, was inversely correlated with the 6-Minute Walk Test (6MWT) distance and total lung capacity percentage.
HC was the CALIPER feature that was most negatively affected by CT dosage. In contrast, CALIPER ILD and VRS demonstrated moderate concordance between standard dose (SD) and ultra-low dose (ULD), as well as almost perfect or substantial concordance, respectively, between SD and limited dose (LD).
In addition, CALIPER-derived total lung volume showed substantial concordance across all CT dosages. Regional CALIPER segmented lung volumes showed substantial concordance between SD and LD and moderate to substantial concordance between SD and ULD.
“In the absence of confounders affecting PFT, CALIPER may serve as a robust, objective adjunct to conventional surrogates for pulmonary function in assessing pulmonary fibrosis in patients undergoing treatment,” the authors concluded.
Koo CW, Larson NB, Parris-Skeete CT, et al. Prospective machine learning CT quantitative evaluation of idiopathic pulmonary fibrosis in patients undergoing anti-fibrotic treatment using low- and ultra-low-dose CT. Clin Radiol. Published online December 7, 2021. doi:10.1016/j.crad.2021.11.006