Abstract
Background: Pleural plaques (PP) are morphological manifestations of long-term asbestos exposure. The relationship between PP and lung function is not well-understood. The time-consuming nature of PP delineation to obtain volume impedes research.
Aims: We aimed to explore the relationship between the pleural plaque volume and pulmonary function tests (PFT). To automatize the laborious task of delineation, we aimed to develop automatic Artificial Intelligence (AI)-driven segmentation of PP
Methods: Radiologists manually delineated pleural plaques in n=422 CT images of patients with occupational exposure to asbestos. The Pearson correlation coefficient (r) was used for the correlation between PP volume and PFT metrics. When recorded, these were VC, FVC, DLCO, and KCO.
Results: We found moderate VC (n=138, r=-0.41), FVC (n=152, r=-0.44) to no correlations DLCO (n=137, r=-0.16) KCO (n=119, r=0.15). For DLCO and KCO, no significant differences were found (p>0.05). Significant differences were found for VC (p=0.02) and FVC (p=0.006). We trained the AI system on subjects in the training set (n=322). On the independent test set (n=100), the correlation between the predicted volume and the ground truth was r=0.89, the median overlap was 0.70 Dice Similarity Coefficient.
Conclusion: We have observed that PP volume is associated with loss in VC and FVC. We successfully developed an AI algorithm to automatically segment PP in CT images to enable fast volume extraction. We envision that this could be used to non-invasively gain insight into lung morphology and lung function. Furthermore, it could make investigating the impact of PP volume more accessible to other researchers.
Footnotes
Cite this article as: European Respiratory Journal 2021; 58: Suppl. 65, PA2544.
This abstract was presented at the 2021 ERS International Congress, in session “Prediction of exacerbations in patients with COPD”.
This is an ERS International Congress abstract. No full-text version is available. Further material to accompany this abstract may be available at www.ers-education.org (ERS member access only).
- Copyright ©the authors 2021