Summary: Localized bone loss, known as void spaces, is not captured using standard measures in clinical computed tomography (CT). This study validated the use of CT to systemically assess void spaces throughout the skeleton. We found that void spaces were localized to skeleton regions (e.g., L3 vs L4), providing new insights into bone fragility. Introduction: Localized trabecular bone microarchitectural loss, referred to as void spaces, can be measured using HR-pQCT and is related to bone fragility. Due to gantry size constraints, HR-pQCT is limited to peripheral sites; therefore, it is unknown whether these features occur at axial skeletal sites. This study aimed to validate the use of quantitative computed tomography (QCT) to assess void spaces at the hip and spine and combine those measures with HR-pQCT to determine its systemic skeletal prevalence. Methods: The original methodology developed to identify void spaces on HR-pQCT was adapted for QCT to accommodate its lower resolution and was validated in 101 participants scanned using both modalities. Validation was performed by comparing void space volume (VS), void space volume ratio (VS/TV), and the Euclidean distance between void space centroids across both modalities. Subsequently, void spaces were measured on a separate cohort of 209 participants who were scanned with HR-pQCT (radius and tibia) and QCT (lumbar spine and hip). Pearson’s correlation assessed VS and VS/TV relationships across skeletal sites. Results: Validation at the knee was confirmed by a strong Pearson correlation between HR-pQCT and QCT for VS (r = 0.88, <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="198_2025_7558_Article_IEq1.gif" Format="GIF" Height="14" Rendition="HTML" Resolution="72" Type="Linedraw" Width="31" /> </InlineMediaObject> <EquationSource Format="TEX">$ p < $</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mi>p</mi> <mo><</mo> </mrow> </math></EquationSource> </InlineEquation> 0.01) and VS/TV (r = 0.83, <InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="198_2025_7558_Article_IEq1.gif" Format="GIF" Height="14" Rendition="HTML" Resolution="72" Type="Linedraw" Width="31" /> </InlineMediaObject> <EquationSource Format="TEX">$ p < $</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mi>p</mi> <mo><</mo> </mrow> </math></EquationSource> </InlineEquation> 0.01), and good spatial similarity (median Euclidean distance = 3.1 mm, IQR = 1.6–5.3 mm). The systemic skeletal site investigation found that the VS/TV correlations were highest between spatially similar sites (e.g., L1–L4, r = 0.50–0.83) and weaker across differing skeletal sites (e.g., radius and L1, r = 0.10–0.49). Conclusion: This study validated the use of QCT to identify void spaces by comparison with HR-pQCT. Across skeletal sites, void spaces tended to be localized to skeletal regions (e.g., L3 vs L4) rather than systemic throughout the skeleton (e.g., L1 vs radius). Tracking void spaces offers insights into bone fragility and may help direct orthopedic treatment options such as hip arthroplasty or bone hydrogels in the future.

Quantifying localized bone loss in clinical computed tomography using void spaces

期刊论文
发表日期：

2025年6月10日

查看全文

Annabel R. Bugbird,
Bryn E. Matheson,
Matthias Walle,
Nathan J. Neeteson,
Nina Pavlovic,
Callie E. Stirling,
Danielle E. Whittier,
Steven K. Boyd

摘要

Summary:

Localized bone loss, known as void spaces, is not captured using standard measures in clinical computed tomography (CT). This study validated the use of CT to systemically assess void spaces throughout the skeleton. We found that void spaces were localized to skeleton regions (e.g., L3 vs L4), providing new insights into bone fragility.

Introduction:

Localized trabecular bone microarchitectural loss, referred to as void spaces, can be measured using HR-pQCT and is related to bone fragility. Due to gantry size constraints, HR-pQCT is limited to peripheral sites; therefore, it is unknown whether these features occur at axial skeletal sites. This study aimed to validate the use of quantitative computed tomography (QCT) to assess void spaces at the hip and spine and combine those measures with HR-pQCT to determine its systemic skeletal prevalence.

Methods:

The original methodology developed to identify void spaces on HR-pQCT was adapted for QCT to accommodate its lower resolution and was validated in 101 participants scanned using both modalities. Validation was performed by comparing void space volume (VS), void space volume ratio (VS/TV), and the Euclidean distance between void space centroids across both modalities. Subsequently, void spaces were measured on a separate cohort of 209 participants who were scanned with HR-pQCT (radius and tibia) and QCT (lumbar spine and hip). Pearson’s correlation assessed VS and VS/TV relationships across skeletal sites.

Results:

Validation at the knee was confirmed by a strong Pearson correlation between HR-pQCT and QCT for VS (r = 0.88, $ p < $ $p <$ 0.01) and VS/TV (r = 0.83, $ p < $ $p <$ 0.01), and good spatial similarity (median Euclidean distance = 3.1 mm, IQR = 1.6–5.3 mm). The systemic skeletal site investigation found that the VS/TV correlations were highest between spatially similar sites (e.g., L1–L4, r = 0.50–0.83) and weaker across differing skeletal sites (e.g., radius and L1, r = 0.10–0.49).

Conclusion:

This study validated the use of QCT to identify void spaces by comparison with HR-pQCT. Across skeletal sites, void spaces tended to be localized to skeletal regions (e.g., L3 vs L4) rather than systemic throughout the skeleton (e.g., L1 vs radius). Tracking void spaces offers insights into bone fragility and may help direct orthopedic treatment options such as hip arthroplasty or bone hydrogels in the future.