An Exploratory Radiomics Approach to Quantifying Pulmonary Function in CT Images

Abstract

Contemporary medical imaging is becoming increasingly more quantitative. The emerging field of radiomics is a leading example. By translating unstructured data (i.e., images) into structured data (i.e., imaging features), radiomics can potentially characterize clinically useful imaging phenotypes. In this paper, an exploratory radiomics approach is used to investigate the potential association between quantitative imaging features and pulmonary function in CT images. Thirty-nine radiomic features were extracted from the lungs of 64 patients as potential imaging biomarkers for pulmonary function. Collectively, these features capture the morphology of the lungs, as well as intensity variations, fine-texture, and coarse-texture of the pulmonary tissue. The extracted lung radiomics data was compared to conventional pulmonary function tests. In general, patients with larger lungs of homogeneous, low attenuating pulmonary tissue (as measured via radiomics) were found to be associated with poor spirometry performance and a lower diffusing capacity for carbon monoxide. Unsupervised dynamic data clustering revealed subsets of patients with similar lung radiomic patterns that were found to be associated with similar forced expiratory volume in one second (FEV 1 ) measurements. This implies that patients with similar radiomic feature vectors also presented with comparable spirometry performance, and were separable by varying degrees of pulmonary function as measured by imaging.

Age-associated changes in aging lungs: A study with auto-segmentation and radiomics based on CT images

“…The present study revealed that two coarse texture features, LowGreyLevelRunEmphasis_AllDirection_offset2_SD (ρ = −0.63) and ShortRunLowGreyLevelEmphasis_AllDirection_offset9_SD (ρ = −0.70), were found to be negatively correlated with DLCO. Similar results were reported by Barjaktarevic et al [29]. Low Gray Level Run Emphasis (LGLRE) is only a measure of CT attenuation, while Short Run Low Gray Level Emphasis (SRLGLE) is a joint-measure of CT attenuation and HU heterogeneity.…”

“…A high HGLRE value indicates a dense and homogenous tissue, while a low value indicates a low attenuating and heterogeneous tissue. The relationship between HGLRE and FEV1 observed in the present study means that: (1) patients with normal spirometry may have a dense and homogenous lung; and (2) patients with abnormal spirometry may also have a low attenuating and heterogeneous lung [29,32].…”

“…It is noteworthy that these features could detect the invisible micro-structural changes of the lung. For example: histogram parameters could be used to assess the attenuation of the lungs [26]; GLCM could reflect the degree of heterogeneity of the whole lung, and the micro-structural changes may give uneven signals in the CT scan, which is irreversible of any anatomy change in the lung [27]; form factor features could re ect the changes in the region of interest, in terms of volume, area and shape, and these may be displayed more visually, allowing the changes in lung volume and shape to be easily observed and compared [28]; in addition to those mentioned above, the study also revealed that the RLM features, which reflect the roughness and directionality, were also associated with the changes in pulmonary structure [29].…”

“…Second, the risk factors of COPD included age, suggesting that COPD has an association with lung aging. Third, declining lung function is a signi cant feature for the development of COPD, which increases with age, especially in smoking individuals [29]. Furthermore, an injured lung function cannot recover after smoking cessation [1].…”

Digital pathology and computational image analysis in nephropathology

“…Although the implementation of these approaches and integration of new knowledge arising from these findings into clinical practice will be complex and require extensive validation, future ML pipelines will be an integral component of the pathologist’s workflow as decision support tools. Indeed, some of these algorithms are already making their way into routine practice in the field of radiation oncology 42 , 124 , 125 . The ultimate challenge will be the integration and implementation of computational image analysis-generated data with other knowledge domains to establish fused datasets and associated morpho-omic classifications, prognostications and predictions for kidney diseases at the individual patient level (that is, fused morpho-omic signatures ).…”