Abstract Rationale Pulmonary structural changes in patients with idiopathic pulmonary fibrosis (IPF) are conventionally assessed using computed tomography (CT), which exposes patients to ionizing radiation. Ultra-short time-to-echo (UTE) MRI is a non-ionizing radiation alternative that allows for the sensitive evaluation of lung density changes over time but lacks a consistent signal reference unlike CT. We aimed to measure lung density in healthy controls and participants with IPF over one-year using an air/blood normalized signal intensity. Methods Participants with IPF underwent free-breathing UTE MRI at baseline and at 1-3, 6-9, and 12-15-month follow-up; controls were assessed at baseline and at 1-3 months. IPF participants were classified as stable or progressive based on 10% decline in forced vital capacity (FVC) at 1-year. Signal intensity (Sᵢnt) was normalized using bias-field correction for RF-coil regional sensitivity using the chest wall. The median Sᵢnt within the trachea and ascending aorta served as air and blood reference values, respectively, for density normalization. Whole-lung Sᵢnt statistics were computed after normalization. One-way ANOVA evaluated group differences; paired t-tests evaluated longitudinal changes. Results Fourteen stable (12M/2F, 70±9y), nine progressive (8M/1F, 72±7y) IPF patients and sixteen healthy controls (6M/10F, 57±14y) were analyzed. At baseline, groups differed significantly for mean (p 0. 001), SD (p 0. 001) and median (p = 0. 002) Sᵢnt, and for FVC%pred (p = 0. 001), FEV1%pred (p = 0. 001), and DLCO%pred (p 0. 001). Sᵢnt mean, SD, and median were significantly greater in progressive vs. stable IPF and healthy controls. All three metrics were greater in stable vs. controls. Longitudinally, DLCO%pred significantly worsened in progressive IPF between baseline and 1-3 months, but not over any other period. FVC%pred and FEV1%pred were not significantly different over time for any group. Sᵢnt mean (p = 0. 004), SD (p = 0. 008) and median (p = 0. 03) were significantly increased in stable IPF between baseline and 1-3 months, and 6-9 months (mean: p = 0. 002; SD: p 0. 001; median: p = 0. 02), but not 12-15 months. Sᵢnt mean (p = 0. 002) and median (p = 0. 001) were significantly increased for progressive IPF between baseline and 6-9 months. Conclusion Anatomical intensity-corrected UTE MRI can detect differences in lung density between healthy, stable, and progressive IPF and shows increases in mean, SD, and median Sᵢnt over time despite stable lung function. As such, anatomical intensity-corrected UTE MRI may offer a quantitative, radiation-free approach to differentiate disease progression in IPF and monitor disease severity over time. Future work will include regional and texture analyses to determine which features of IPF contribute to the observed increases in Sᵢnt metrics. This abstract is funded by: NIH/NHLBI R01HL169765, NIH/NHLBI R01 HL126771, NIH NCATS S10OD026960, NIH NCATS S10OD025025, and the NIH CTSA program grant UM1TR004403
Gritters et al. (2026) studied this question.