Novel deep-learning analysis for connective tissue disease—related interstitial lung disease extent assessment on CT: a preliminary cross-sectional study
Authors' conclusion
QZIP-ILD® demonstrates superior capability in distinguishing patients with a radiological scenario correlated to severe physiological impairment, while showing relatively good correlations in quantifying the extent on HRCT compared with conventional method in CTD-ILD.
Study context
- Population
- Patients with CTD-ILD visiting a Rheumatology Centre between December 2020 and April 2024median age 66 years
- Participants
- 80 patients
- Exposure
- Deep-learning analysis (QZIP-ILD® software, DL-QZIP) vs semiquantitative visual method for interstitial lung disease (ILD) extent on high-resolution CT (HRCT)
- Outcome
- Distinguishing forced vital capacity (FVC)% <70 by QILD scores (AUC)
Methodological detailspreliminary
- Preliminary
Main finding
As reported in the abstract
| Group | AUC |
|---|---|
| DL-QZIP | 0.833 |
| Visual method | 0.660 |
AUC
DL-QZIP0.833
Visual method0.660
- DL-QZIP vs Visual method: P < 0.01 — AUC significantly higher with DL-QZIP
Observational data: associations, not effects.
Further findings
Median quantitative scores (% lung involvement): QGG, QFIB, QILD
- QGG, DL-QZIP vs visual method
- 3.45% vs 3.25%
- QFIB, DL-QZIP vs visual method
- 2.19% vs 4.06%
- QILD, DL-QZIP vs visual method
- 5.35% vs 8.48%
- Median
Correlation (intraclass correlation coefficients (ICC)) between DL-QZIP and visual method
- QGG
- 0.75
- QFIB
- 0.61
- QILD
- 0.75
Original publication
Novel deep-learning analysis for connective tissue disease—related interstitial lung disease extent assessment on CT: a preliminary cross-sectional study. Rheumatology (Oxford). Published 13 September 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →