Low-dose vs standard-dose photon-counting CT for myopathy-associated interstitial lung disease: a prospective feasibility study.

OBJECTIVES: To evaluate the clinical utility of photon counting computed tomography (PCCT)-derived quantitative parameters for assessing interstitial lung disease (ILD) burden in patients with inflammatory idiopathic myopathy (IIM), comparing standard-dose (SD) and tin-filtered low-dose (LD) protocols, and to assess the added value of ultrathin reconstructions. MATERIALS AND METHODS: In this single-centre prospective study (September 2023-November 2024), outpatients with IIM-ILD underwent two consecutive PCCT scans. Quantitative ILD features were extracted using a commercial deep-learning software (ContextFlow ADVANCE Chest CT). Primary analyses compared LD 1.0 mm with SD 1.0 mm reconstructions using Bland-Altman analysis, with exploratory analyses of ultrathin (0.4 mm) reconstructions. Correlation analyses between deep learning-derived quantitative parameters, visual assessment, and pulmonary function tests were performed. RESULTS: Forty-eight adults with IIM-ILD (40 females; mean age, 53.88 ± 11.13 years) were included. The LD protocol achieved a substantial dose reduction (CTDIvol: 0.77 vs 5.40 mGy) while maintaining good agreement with SD (mean net differences [MNDs] < 1% for all patterns). LD-1.0 mm protocol yielded higher volume of ground-glass opacity (MND, 0.98%) and lower volume of fibrosis extent (MND, -0.64%). Reticular pattern showed strong association with FVC% (rhoSD1.0mm = -0.62, p < 0.001) and DLCO% (rhoSD1.0mm = -0.73, p < 0.001), with similar correlations observed under LD acquisition (Δrho ≤ 0.02). Ultrathin reconstructions yielded greater measurement of bronchiectasis (MND, 0.13% on SD scans). CONCLUSION: LD PCCT serves as a radiation-sparing alternative for quantitative disease assessment. Ultrathin reconstructions provide limited additional value. Future longitudinal studies are warranted to validate the performance of LD PCCT for disease monitoring. KEY POINTS: Question Can LD photon-counting CT provide robust quantitative assessment of ILD burden in patients with inflammatory myopathy compared with SD imaging? Findings Deep learning segmentation yielded LD ILD patterns that closely matched SD results, with all MNDs below 1%. Clinical relevance LD PCCT maintains quantitative robustness and physiological validity for IIM-ILD assessment, with total disease burden showing good agreement with the SD protocol, offering a radiation-sparing alternative without compromising CT-physiology correlation.