Dissertation
From fissure integrity to regional mechanics: quantitative CT evidence in COPD
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Autumn 2025
DOI: 10.25820/etd.008244
Abstract
Chronic obstructive pulmonary disease (COPD) is heterogeneous in both structure and function, and interlobar fissures may help organize how deformation and sliding are distributed during breathing. In this work, we combined deep learning-based fissure integrity assessment, lobe-aware registration, and ensemble quality control to quantify how fissure integrity relates to regional lung biomechanics in a large COPD cohort.
Paired TLC/RV chest CT scans from 3,081 SPIROMICS participants were processed through a pipeline that included deep learning-based fissure, lobe, and integrity segmentation, reconstruction- and uncertainty-based quality control, and tissue- and vessel-preserving lobewise registration. After triage, 2,532 subjects were included for analysis. From the resulting displacement fields, we computed voxel-wise surrogates of lung mechanics and summarized their means and heterogeneities in fissure-proximal regions, lobes, and whole lungs.
Fissure integrity proved to be highly variable across subjects and fissures and declined only modestly with COPD severity. In fissure-adjacent parenchyma, higher FI was associated with slightly greater mean expansion, weaker deformation anisotropy, and reduced deformation heterogeneity, indicating more organized but not dramatically larger deformation near intact fissures. At fissure surfaces, sliding magnitude showed a strong, positive association with fissure integrity, even after adjusting for COPD severity. Classical upper-lower gradients in deformation were reproduced and were largely independent of fissure integrity but strongly modulated by lobar sliding magnitude. Higher fissure integrity was also linked to more symmetric emphysema burden between adjacent lobes.
These findings support a view of fissure integrity as a biomechanical organizer that shapes where and how sliding and deformation are expressed, superimposed on the broader effects of lung inflation, gravity, and COPD-related tissue destruction. The methodological framework developed here – deep learning fissure integrity assessment, lobe-wise registration, and scalable quality control – provides a foundation for future studies that seek to integrate anatomy, mechanics, and clinical outcomes to refine COPD phenotyping and treatment selection.
Details
- Title: Subtitle
- From fissure integrity to regional mechanics: quantitative CT evidence in COPD
- Creators
- Zachary W. Althof
- Contributors
- Joseph M. Reinhardt (Advisor)Eric A. Hoffman (Committee Member)Sarah E. Gerard (Committee Member)David W. Kaczka (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Biomedical Engineering
- Date degree season
- Autumn 2025
- DOI
- 10.25820/etd.008244
- Publisher
- University of Iowa
- Number of pages
- xv, 114 pages
- Copyright
- Copyright 2025 Zachary Althof
- Grant note
This work is supported in part by NIH training grant T32 HL144461 and NIH grant HL142625, and the Roy J. Carver Charitable Trust.
- Language
- English
- Date submitted
- 12/08/2025
- Description illustrations
- Illustrations, graphs, charts, tables
- Description bibliographic
- Includes bibliographical references (pages 78-94).
- Public Abstract (ETD)
Chronic obstructive pulmonary disease (COPD) is a common lung disease that makes it hard to breathe and limits daily activities. In COPD, damage does not occur evenly throughout the lungs. Instead, different regions can be affected in different ways. The lungs are divided into five lobes by thin internal walls called fissures. These fissures are sometimes incomplete, allowing air to move between lobes through shortcuts known as collateral ventilation. Fissure anatomy is already used to decide which patients may benefit from valve-based lung volume reduction procedures, but we know much less about how fissures influence the way the lungs move during breathing.
of people with and without COPD. We built an automated pipeline that uses modern deep learning tools to (1) map fissure completeness, (2) check for imaging and algorithm errors, and (3) estimate how each part of the lung expands, stretches, and slides as people move from exhalation to inhalation.
The results show that fissure completeness varies widely between people and changes only slightly with disease severity. More intact fissures were linked to stronger and more sharply localized sliding between lobes and to more balanced damage (emphysema) on either side of a fissure. Advanced COPD was marked by generally weaker and more uniform motion, regardless of fissure anatomy.
These findings suggest that fissure integrity is not just a structural detail or a screening tool for valve therapy. It also helps organize how the lungs move, which may ultimately influence how disease develops and which treatments work best for individual patients.
- Academic Unit
- Roy J. Carver Department of Biomedical Engineering
- Record Identifier
- 9985135149002771
Metrics
1 Record Views