Conclusion
To the best of our knowledge this was the first study to use software
aided methods for a deeper analysis of HRCT data which is often
routinely gathered in SSc patients.
It was demonstrated that lung
density decreased in patients with any signs of PF compared to those
without PF and healthy controls. It was also shown that, even in the
absence of signs of pulmonary fibrosis in HRCT, lung density was
decreased in patients with dcSSc compared to both lcSSc patients and
healthy controls. Decreased baseline lung density may be a predictive
parameter for future PF in SSc patients although this proposal requires
validation in larger patient groups. The methods described herein may
contribute to the use of software-assisted machine learning techniques
in assessing lung parenchyma in SSc.
Acknowledgments The authors thank all investigators for their
contribution to the study. The corresponding author certifies that all
authors approved the entirety of the submitted material and contributed
actively to the study. We would like to thank Mr. Jeremy Jones, of the
Kocaeli University Academic Writing Department, for the revision of the
English in this paper.
Funding No funding.
Conflict of interest None of the authors has financial or
non-financial conflicts of interest to disclose.
Ethical approval This article does not contain any studies with
animals performed by any of the authors. Informed consent was obtained
from all individual participants included in the study.
Author Contributions : DTK acquired the clinical data,
contributed to experimental plan design, performed all the statistical
analysis and drafted the manuscript. OC contributed to the acquisition
of HRCT data. AK contributed to the acquisition of clinical data. AY and
AC contributed to critical revision of the manuscript.