267 - Inter-Individual Variability in Thoracic vs. Abdominal Breathing Patterns Limits Generic SGRT Tracking in DIBH: A Prospective Optical Motion Capture Study

12:40pm - 12:50pm ET

Room 258

Presenter(s)

Mohamed Yousuf, PhD - Thomas Jefferson University, Philadelphia, PA

M. Yousuf¹, B. Shaver², A. Peshkar³, D. Gurari³, B. Gaffney², Y. Vinogradskiy⁴, and D. Thomas⁴; ¹Thomas Jefferson University, Philadelphia, PA, ²University of Colorado Denver | Anschutz Medical Campus, Denver, CO, ³University of Colorado Boulder, Colorado, CO, ⁴Dept. of Radiation Oncology, Sidney Kimmel Medical College and Comprehensive Cancer Center, Thomas Jefferson University, Philadelphia, PA

Purpose/Objective(s):

The clinical reliability of Surface Guided Radiation Therapy (SGRT) relies on the assumption that surface motion is an accurate surrogate for internal target position. In clinical practice, regions for surface tracking are identified without a quantitative assessment of patient-specific respiratory mechanics. Patients who exhibit primarily thoracic breathing patterns may not be well represented by generic ROI templates that favor abdominal tracking, potentially compromising tracking accuracy. This variability may lead to suboptimal surface surrogacy during DIBH. We hypothesized that inter-individual variability in chest vs abdominal breathing may limit the reliability of SGRT. This study quantifies thoracoabdominal motion patterns using high-fidelity ground-truth optical motion capture to validate the sufficiency of standard ROI selection.

Materials/Methods:

In an IRB-approved prospective clinical trial, respiratory motion was recorded using a 12-camera optical motion-capture system with 72 markers, including 16 placed on the chest and abdomen. Subjects completed breathing trials under four conditions: free breathing (FB), deep breathing (DB), and instructed thoracic and abdominal-dominant breathing patterns (TDB and ADB). The TDB and ADB maneuvers served as subject-specific reference baselines to isolate maximum thoracic and abdominal expansion capacities.

To evaluate tracking suitability, we established a primary endpoint, "vertical respiratory bias" (?), which serves as a normalized, subject-specific similarity metric ranging from +1 (pure thoracic dominance) to -1 (pure abdominal dominance). Specifically, ? was calculated for both FB and DB by computing the difference in spatial deviations for all 16 markers between the unguided breathing state and the subject's own TDB versus ADB reference baselines.

Results:

Of the 20 participants (7M/13F, Age: 23-34, BMI: 15.6-35.4 kg/m²), during deep breathing (DB), ? values varied substantially between individuals (0.09 ± 0.35), with subjects’ breathing distributed across both thoracic and abdominal-dominant patterns with no significant population-level preference (p = 0.08). In contrast, free breathing (FB) demonstrated consistent abdominal dominance (? = –0.37 ± 0.35, p = 0.0002), with 80% (16/20) of subjects exhibiting abdominal-dominant motion. Notably, 90% (18/20) of subjects changed their dominant breathing pattern between free and deep breathing. No significant associations were observed between breathing pattern and BMI or gender (p > 0.05).

Conclusion:

Free-breathing exhibits reliable abdominal-dominant, while deep breathing exhibits high inter-subject variability and frequent pattern shifts. This shows that generic abdominal ROI templates may compromise SGRT accuracy during DIBH. These findings support the need for patient-specific ROI optimization or pretreatment breathing assessment to ensure the tracking surface correlates with the primary motion vector.