Quantification of Spatial Ventilation Defect Sparsity in Hyperpolarized Gas Magnetic Resonance Imaging of Lungs Utilizing a Three‐Dimensional Clustering Algorithm

ABSTRACT Hyperpolarized gas (HPG) magnetic resonance (MR) imaging allows for the quantification of pulmonary defects with the ventilation defect percentage (VDP). Although informative, VDPs lack information regarding the spatial distribution of defects. We developed a method of quantifying the focal...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:NMR in biomedicine Jg. 38; H. 3; S. e70005 - n/a
Hauptverfasser: García Delgado, Gabriela María, Shammi, Ummul Afia, Ruppel, Mia R., Altes, Talissa A., Mugler, John P., Meyer, Craig H., Qing, Kun, Lange, Eduard E., Mata, Jaime, Ruset, Iulian C., Hersman, F. W., Thomen, Robert P.
Format: Journal Article
Sprache:Englisch
Veröffentlicht: England Wiley Subscription Services, Inc 01.03.2025
Schlagworte:
Adult
Age
Aged
algorithm
Algorithms
Asthma
Asthma - diagnostic imaging
Asthma - physiopathology
Chronic obstructive pulmonary disease
Cluster Analysis
Clustering
Clustering Algorithms
Correlation coefficient
Correlation coefficients
Defects
Female
gas
Helium
Humans
hyperpolarized
Imaging, Three-Dimensional
Lung - diagnostic imaging
Lung - physiopathology
Lung diseases
lungs
Magnetic Resonance Imaging
Male
Middle Aged
MRI
Pulmonary Disease, Chronic Obstructive - diagnostic imaging
Pulmonary Disease, Chronic Obstructive - physiopathology
Pulmonary Ventilation
Regression analysis
Regression models
Spatial distribution
Statistical analysis
Variance analysis
Ventilation
Young Adult
hyperpolarized
MRI
defects
ventilation
gas
lungs
clustering
algorithm
ISSN:0952-3480, 1099-1492, 1099-1492
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:ABSTRACT Hyperpolarized gas (HPG) magnetic resonance (MR) imaging allows for the quantification of pulmonary defects with the ventilation defect percentage (VDP). Although informative, VDPs lack information regarding the spatial distribution of defects. We developed a method of quantifying the focality/sparseness of ventilation defects in hyperpolarized‐gas lung MR images. The study involved a total of 56 subjects: 14 asthmatics (age mean ± sd = 45.1 ± 18.9), 25 COPD subjects (age = 60.6 ± 10.4), and 17 CF subjects (age = 21.8 ± 8.4). The analyzed data are from four different studies: Study 1 used a 3‐T gradient echo (GRE) sequence, Study 2 used a 1.5‐T GRE sequence, Study 3 used a 1.5‐T two‐dimensional spiral sequence, and Study 4 used a 1.5‐T three‐dimensional balanced steady‐state free precession (bSSFP) sequence. We developed an algorithm that quantifies the focality/sparseness of ventilation defects as a subject's cluster index (CI). The algorithm was assessed on synthesized spherical defect clusters and 3D lung volume defects of varying sizes/distributions. CI and whole‐lung VDP were calculated for asthmatic, COPD, and CF subjects. Pearson correlation coefficients and linear regression models between CI and FEV1pp, as well as CI and VDP, were performed to evaluate CI among asthma, COPD, and CF. T tests were performed to evaluate CI/VDP ratios among previously mentioned lung diseases. p values less than 0.05 were statistically significant. Subject CI well represents defect focality by visual inspection. Pearson correlation coefficients between CI and VDP were r = 0.60 (p = 2.21 × 10−2) for asthma, r = 0.79 (p = 3.15 × 10−6) for COPD, and r = 0.84 (p = 2.80 × 10−5) for CF. Pearson correlation coefficients between CI and FEV1pp was r = −0.47 (p = 0.0002). Analysis of variance (ANOVA) and a Tukey's honestly significant difference (HSD) test revealed that the ratio of whole‐lung CI/VDP was significantly different between asthma/CF (p = 0.04) and CF/COPD (p = 0.008), but not among asthma/COPD (p = 0.95). This method of volumetric quantification of defect spatial distribution may provide information regarding defect cluster size in which VDP alone is uninformative. Hyperpolarized gas magnetic resonance imaging allows for the quantification of pulmonary defects with the ventilation defect percentage, which lacks information regarding the spatial distribution of defects. We developed an algorithm that quantifies the focality/sparseness of ventilation defects as a subject's cluster index (CI). CI and whole‐lung VDP were calculated for asthmatic, COPD, and CF subjects. ANOVA and Tukey's HSD test revealed the ratio of whole‐lung CI/VDP was significantly different between asthma/CF (p = 0.04) and CF/COPD (p = 0.008), but not among asthma/COPD (p = 0.95).
Bibliographie:This work was supported by Xemed, Vertex Pharmaceuticals, GlaxoSmithKline, and the National Institutes of Health (R01‐HL152288).
Funding
ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:0952-3480
1099-1492
1099-1492
DOI:10.1002/nbm.70005