Skip to main content
NCBI home page
Journal of Orthopaedic Surgery and Research logoLink to Journal of Orthopaedic Surgery and Research
. 2026 Jan 31;21:165. doi: 10.1186/s13018-026-06670-1

A multiple joint morphometric analysis of female patients with progressive collapsing foot deformity: a cross-sectional study

Takuma Miyamoto 1,2, Rich J Lisonbee 1, Kassidy Knutson 3, Hiroaki Kurokawa 2, Akira Taniguchi 2, Yasuhito Tanaka 2, Amy L Lenz 3,✉
PMCID: PMC12947440  PMID: 41620735

Abstract

Objective

Progressive collapsing foot deformity is a complex clinical presentation of combined deformities, which affects articular joint relationships. Herein, the study aimed to characterize joint interactions of the ankle, tibiofibular, talofibular, subtalar, talonavicular, calcaneocuboid, naviculo-cuneiform, and tarso-metatarsal joints.

Materials and methods

We quantified and compared the results between 23 female patients with progressive collapsing foot deformity and 23 female asymptomatic individuals. We used a multi-domain correspondence model from statical shape modeling to compare the alignment and morphology followed by calculating the joint-level measurements including the joint space distance and the congruence index between groups.

Results

From our results we found that almost all bones and joints were affected by the progressive collapsing foot deformity. Our main results for joint space distances were narrowing in the sinus tarsi, gapping at the medial of the calcaneocuboid joint, and narrowing at the 3rd, 4th, and 5th tarso-metatarsal joints. The primary results for the congruence index were shifted to the lateral side in the talonavicular joint and a less congruent middle facet of the subtalar joint in the progressive collapsing foot deformity group. In addition, the joint space distance was mainly influenced by alignment, and the congruence index was influenced by bone morphology.

Conclusion

We believe that assessing multi-joint interactions in progressive collapsing foot deformity will lead to a better understanding of the pathophysiology and assist in surgical treatment planning.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-026-06670-1.

Keywords: Progressive collapsing foot deformity, Joint measure analysis, Statistical shape modeling, Joint space distance, Congruence index

Introduction

Progressive collapsing foot deformity (PCFD) is a common deformity affecting the foot and ankle [1]. PCFD is a complex clinical condition with complex deformities that may be due to the medial longitudinal arch [2, 3]. However, the pathophysiology is not yet fully understood, and PCFD may affect the subtalar joint (ST), Chopart joint, and Lisfranc joint [4]. Furthermore, there is a wide variety of surgical treatments for PCFD, with different surgical techniques affecting different joints [5]. Adequate correction is very important because under- or overcorrection of deformities leads to poor postoperative outcomes [6]. Therefore, it is critical that we expand our understanding of the multiple joint interactions in PCFD.

In recent years, computed tomography (CT) has become a common tool in diagnosing and evaluating PCFD [7, 8]. High-resolution volumetric data enables accurate 3D analyses of deformity and subluxation which can be quantified as joint measures within these complex joints and can be used to create statistical shape models (SSM). SSM provides a 3D evaluation of population-wise anatomical variation as well as group-wise shape differences of anatomies reconstructed from volumetric imaging data without the influence of human bias [9–12]. Multi-domain SSM analyses allow for a broader understanding of the bone-to-bone relationships within a joint.

Herein, the study aimed to characterize joint interactions of the ankle, tibiofibular, talofibular, subtalar (ST), talonavicular (TN), calcaneocuboid (CC), naviculo-cuneiform (NC), and tarso-metatarsal (TMT) joint using Joint Measurement Analysis (JMA) in female individuals presenting with PCFD as compared to asymptomatic controls. It is hypothesized that peri-talar dislocations strongly affect the interaction of the joints around the talus, and in addition, forefoot abduction deformities affect CC and TMT joint interactions.

Materials and methods

Study design

This is a retrospective comparative study and has been approved by the authors’ affiliated institutions. An opt-out statement regarding the application of medical data was published on our institute’s website. This study was performed under the principles of the World Medical Association Declarations of Helsinki. The workflow of this study is presented in Fig. 1.

Fig. 1.

Fig. 1

Open in a new tab

Computational workflow for this study

Data collection

This study utilized the same participants as previous studies [13, 14]. A retrospective study was conducted on 35 female patients who underwent surgical treatment for PCFD at Nara Medical University between 2014 and 2019. Notable ankle and hindfoot bone morphologies are reported between women and men [15], and the incidence of PCFD is three times higher in women [16]. Therefore, only women were included to focus on disease-specific differences. The exclusion criteria included previous surgeries on the same limb, pediatric/congenital deformity; inflammatory arthropathy (e.g., rheumatoid arthritis); neuropathic arthropathy, severe hindfoot and midfoot bone collapse, and unavailability of imaging data for the hind-, mid-, or forefoot bones selected for the study. Severe collapse was defined as osseous destruction/displacement or joint destruction that obscured joint boundaries on CT, making segmentation and joint-surface measurements unreliable. Eventually, a sample of 23 patients was chosen, with an average age of 65.1 ± 7.06 years, a height of 154 ± 4.91 cm, a weight of 60.4 ± 9.79 kg, and a BMI of 25.1 ± 4.29 kg/m2. The PCFD classification [4] categorized 1 patient as 1B, 1 patient as 1C, 2 patients as 2B, 6 patients as 2C, 10 patients as 2D, and 3 patients as 2E. The asymptomatic control group consisted of 23 women whose analyzed limbs had no history of ankle joint injuries, deformities, or surgical trauma. These individuals had undergone CT scans as part of clinical evaluations for trauma or tumor around hip in the contralateral limb. The group had an average age of 46.7 ± 19.2 years, a height of 155 ± 7.22 cm, a weight of 58.2 ± 14.1 kg, and a BMI of 23.9 ± 5.48 kg/m2. For sample size calculations involving statistical shape modeling, typical statistical methods of a priori sample size calculation are replaced by quantitative metrics of compactness, generalization, and specificity to assess the model outputs and optimization [17]. In this study, these metrics were assessed retrospectively after model construction to confirm the adequacy of the sample size and the quality of the model.

For the study, a simulated weight-bearing computed tomography (SW-CT) was performed using an OPTIMA 660 device (General Electric, Boston, USA) with a 0.625 mm slice thickness and a 512/512 matrix resolution. Acquisition and reconstruction settings followed an institutional protocol and were applied consistently across participants. This was combined with DynaWell L-Spine (Dyna-Well Inc., Las Vegas, USA) to simulate a weight load of 300 N at the ankle-neutral position [18]. Patellar-forward positioning was used in all scans to standardize limb rotation and improve reproducibility. Partial weight-bearing greater than 25% bodyweight was determined sufficient to reproduce full weight-bearing sagittal foot alignments [19]. Therefore, the utilized method with 300 N exceeded the 25% bodyweight force for each of the study participants. The load was applied as an axial compressive force through the plantar surface and maintained during the CT acquisition.

Bone segmentation

For each participant, SW-CT images were auto-segmented to create 3D models of distal tibia, distal fibula, talus, calcaneus, navicular, cuboid, all cuneiforms, and all metatarsal bones (Disior 2.1, Bonelogic) [8], then manually segmented to finalized models (Mimics 22.0, Materialise), and the generated bone surfaces were meshed consistently and smoothed in 3-matic (16.0, Materialise) using the smoothing tool up to three iterations, specifically to remove segmentation-related surface irregularities. All bone surfaces were rigidly aligned and centered using an iterative closest point (ICP) algorithm [20] after mirroring right-sided bones to represent left.

Statistical shape model

For each of the 46 individuals, the tibia, fibula, talus, calcaneus, navicular, cuboid, all cuneiforms, and all metatarsal bones were used to create a multi-domain statistical shape model (MD SSM) using an open-source SSM software (ShapeWorks v6.5, University of Utah). The SSM method used utilizes co-registered landmarks, or correspondence particles, mathematically optimized in the same relative anatomic position across all shapes using registration techniques [21–26]. A generalized Procrustes analysis is applied to all bones to account for size variation. The distributions of correspondence particle locations are then analyzed to determine anatomical morphometric 3D differences. The total particle count for each bone is as follows: 1,024 for the distal tibia, talus, and calcaneus, and 512 for distal fibula, navicular, cuboid, each of the cuneiforms and metatarsals. Particle correspondence optimization used 1000 iterations per split, 1000 optimization iterations, and Procrustes alignment was enabled with scaling and rotation/translation. Initial landmarks were not used.

Particle-wise statistical testing and multiple-comparison control

Particle-wise group differences in 3D correspondence locations were assessed using a parametric Hotelling’s T2 test at each particle [27, 28]. At each comparison, the multivariate dimensionality was p = 3 (x, y, z) with n = 23 per group (p < n), ensuring the pooled covariance matrix was full-rank. Hotelling’s T2 assumes multivariate normality within groups and equality of covariance matrices; because tests were performed across thousands of particles, assumptions were not formally tested at every particle. However, the small dimensionality, balanced group sizes, and absence of extreme outliers support robustness to modest deviations from normality. Particle-wise p-values were stabilized using a repeated balanced-resampling procedure, and multiple comparisons were controlled using a Benjamini–Yekutieli false-discovery-rate correction.

Joint measurement analysis

The correspondence particles from each correspondence model was used to calculate and compare the joint space measurements. The joint space distance at a correspondence particle is calculated as the Euclidean distance between its paired node and the nearest opposing surface node using the open-source JMA toolbox [29]. Distance results are mapped to their respective correspondence particle using MATLAB. For visualization, correspondence particles with a distance value greater than six millimeters (ST, talofibular, tibiofibular, and ankle joints) or three millimeters (NC, TMT, CC, and NC joints) were not included for joint space analysis. These thresholds were chosen to avoid truncating the articular region near joint margins while excluding clearly non-articulating surfaces; in larger joints, cartilage thickness can be as high as ~ 2 mm per side, and thus a more permissive threshold was used to ensure full joint coverage. Importantly, the articulating region included in the joint analysis is primarily defined by joint coverage (i.e., correspondence-based opposing surface pairing within the joint), and surfaces outside the coverage region are not included even if they are spatially close to the opposing bone. The accuracy and repeatability of the correspondence-based joint distance measurement approach have been validated previously [30].

Joint congruency was quantified using the congruence index (CI, mm−1) based on the curvature-matching formulation described by Ateshian et al. [31]. A CI of 0 indicates perfectly matched opposing surface curvatures (maximal congruence), whereas larger CI values indicate greater curvature mismatch (lower congruence).

Prior studies have reported CI values in foot/ankle joints using correspondence-based measurement frameworks, supporting the interpretability of this metric [11, 32, 33]. The congruence index results are mapped to their respective correspondence particle using MATLAB. These data were then statistically compared following a Shapiro–Wilk’s normality test using either a Student’s t-test for parametric or Wilcoxon rank sum test for nonparametric.

Results

The information regarding the subjects’ characteristics can be found in Table 1. There were no significant differences in height, weight, or BMI (all p > 0.05).

Table 1.

Patient demographics. Bold values indicate statistically significant differences between groups

Case (n) Age (years) Height (cm) Weight (kg) BMI (kg/m2)
Asymptomatic Control 23 47.8 ± 19.2 155.0 ± 7.2 58.1 ± 14.1 23.9 ± 5.5
PCFD 23 65.1 ± 7.1 154.0 ± 4.9 60.4 ± 9.8 25.1 ± 4.3
p-value (α = 0.05)  < 0.01  > 0.05  > 0.05  > 0.05
Open in a new tab

Morphometric shape comparison

Morphometric comparison of the PCFD versus asymptomatic control groups found that 75.71% of particles were statistically significant for alignment and shape, 23.39% of particles were statistically significant for alignment only, 0.65% of particles were statistically significant for shape only, and 0.25% of particles were not statistically significant at all. Although it is ultimately a combination of shape and alignment that leads to PCFD differences, in the 3rd, 4th, and 5th TMT joint, it is more dominated by alignment in the group differences (Supplementary Fig. 1).

Joint measurements analysis

For reference, congruence index values in asymptomatic controls were generally near zero within articulating regions, consistent with high curvature matching, whereas regions demonstrating reduced congruence in PCFD showed larger CI values.

Ankle joint

In a comparison of asymptomatic controls and PCFD, the values of joint space distance were significantly smaller posteriorly, mid-lateral, and medial, and significantly larger anteriorly in PCFD. The congruence index showed that the medial posterior was significantly larger and the anterior was significantly smaller in PCFD (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

This is a quantitative visual description of Joint space distance and Congruence Index at Ankle joint of PCFD and Asymptomatic control. The color of each particle indicates the scale of each value. Black mark areas indicate significant differences between groups

Tibiofibular and talofibular joint

In a comparison of asymptomatic controls and PCFD, the values of joint space distance were significantly large at the proximal tibiofibular joint and proximal talus, and significantly smaller at the lateral process of the talus in the PCFD. The congruence index was significantly smaller at the proximal tibiofibular joint and proximal talus, and significantly larger at the lateral process of the talus in the PCFD (Fig. 3).

Fig. 3.

Fig. 3

Open in a new tab

This is a quantitative visual description of Joint space distance and Congruence Index at Tibiofibular and Talofibular Joint of PCFD and Asymptomatic control. The color of each particle indicates the scale of each value. Black mark areas indicate significant differences between groups

Subtalar joint

In a comparison of asymptomatic controls and PCFD, the values of joint space distance were significantly large at the medial side of the posterior facet, and significantly smaller at the anterior facet and sinus tarsi in the PCFD. The congruence index was significantly smaller at the posterior facet and middle facet in the PCFD (Fig. 4).

Fig. 4.

Fig. 4

Open in a new tab

This is a quantitative visual description of Joint space distance and Congruence Index at Subtalar Joint of PCFD and Asymptomatic control. The color of each particle indicates the scale of each value. Black mark areas indicate significant differences between groups

Chopart joint

In a comparison of asymptomatic controls and PCFD, the values of joint space distance were significantly large at the medial inferior side of the TN joint and CC joint, and significantly smaller at the lateral superior side of the TN joint and medial-dorsal side of the CC joint in the PCFD. The congruence index was significantly smaller at the lateral-dorsal side of the TN joint and lateral side at the CC joint, and significantly larger at the medial-planter side of the TN joint and at the medial -planter side of the CC joint in the PCFD (Fig. 5).

Fig. 5.

Fig. 5

Open in a new tab

This is a quantitative visual description of Joint space distance and Congruence Index at Chopart Joint of PCFD and Asymptomatic control. The color of each particle indicates the scale of each value. Black mark areas indicate significant differences between groups

Naviculo-cuneiform joint

In a comparison of asymptomatic controls and PCFD, the values of joint space distance were significantly large at the medial and lateral NC joint in the PCFD. The congruence index was significantly smaller and the lateral NC joint, and significantly larger at the inferior side of the intermediate NC joint in the PCFD (Supplementary Fig. 2).

Tarso-metatarsal joint

In a comparison of asymptomatic controls and PCFD, the values of joint space distance were significantly large at the inferior side of the 2nd TMT joint, and significantly smaller at the 3rd, 4th, 5th TMT joint in the PCFD. The congruence index was significantly smaller at the planter side of the 2nd TMT joint and the 3rd, 4th, and 5th TMT joint, and significantly larger at the dorsal side of the 2nd TMT joint in the PCFD (Fig. 6).

Fig. 6.

Fig. 6

Open in a new tab

This is a quantitative visual description of Joint space distance and Congruence Index at Tarso-metatarsal Joint of PCFD and Asymptomatic control. The color of each particle indicates the scale of each value. Black mark areas indicate significant differences between groups

Discussion

This study utilized MD SSM and JMA to analyze morphometric differences and joint interactions from the ankle to the TMT joint in both female individuals presenting with PCFD and asymptomatic controls. Our results found regions of statistically significant differences between PCFD and controls in joint space distance and congruence index for nearly every joint. In addition, the morphometric analysis found 75.1% of particles were statistically different for alignment and shape between the two groups. The mechanical complexity of the foot is not just one simple joint but is said to result from a combination of many structures that can be altered when deformity is present [34]. Our results also found that almost every bone and joint was involved in PCFD. This finding indicates the usefulness of assessing PCFD across the entire foot instead of focusing too much on one specific structure.

Our primary results for joint space distances found narrowing in the sinus tarsi, gapping at the anterior facet of the ST joint, and narrowing at the 3rd, 4th, and 5th TMT joint. Prior studies have demonstrated the occurrence of lateral impingement to the distal fibula of the talus and/or calcaneus as the peritalar subluxation progresses [35]. Which indicates, as hypothesized, that the results of this study support previous reports and findings. In addition, our previous study found that all cuneiform and metatarsal bones demonstrated no statistical difference in shape between PCFD and asymptomatic control [13]. In this study, shape differences were more dominated by alignment in the 3rd, 4th, and 5th TMT joints. Suggesting that the relatively small joint space distance of the 3rd, 4th, and 5th TMT joints of the PCFD group were not because of metatarsal bone morphology, but rather due to differences in foot alignment.

Evans et al. noted that the lateral column appeared relatively short compared to the medial column in flatfoot [36]. Therefore, the commonly used surgical treatment for PCFD, lateral column lengthening (LCL), involved lengthening the calcaneal and CC joints. We hypothesized that the forefoot abduction deformity would affect the joint interaction between the CC and TMT joint. The results of this study support this hypothesis, with differences in joint space distance between Lisfranc and CC joints. However, these characteristics were very curious. The 3rd, 4th, and 5th TMT joints were affected by forefoot abduction due to their short joint space distance, which contributed to the shortening of the lateral column. The CC joint, on the other hand, was affected by abduction of the Chopart joint mainly due to medial gapping of the joint space. Our previous reports have shown that the Lisfranc joint has a complementary deformity progression to that of the Chopart joint [14]. Suggesting that the change in the alignment of the Chopart joints, which were further spaced on the medial side of the CC joint, may result in a compensatory reduction of the joint space distance of the 3rd, 4th, and 5th TMT joints, and then result in a shortening of the lateral column and abduction of the forefoot. Furthermore, if the condition of the Chopart joint is changed by surgery for LCL, the joint space distance of the 3rd, 4th, and 5th TMT joints also contains the possibility of change. LCL overcorrection of the deformity has been related to reduced postoperative outcomes [6]. This LCL overcorrection may be influenced by the interaction of these joints, and evaluation of multi-joint interactions in PCFD may assist in surgical treatment planning.

The main results for the congruence index are shifted to the lateral side in the talonavicular joint, a decrease in the posterior facet of the ST joint, and a decrease in the middle facet of the ST joint. Previous reports found that PCFD had the 3D bone shape of shifted of the talar head to the inferior and adductor [13, 37, 38], medio-inferior defects of the articular surface of the navicular [13], and narrow posterior facet and anterior-medial facet shifted anteriorly in the calcaneus [13, 39]. The locations of these bone morphology differences corresponded with differences of the congruence index. Perhaps the bone morphology in PCFD has a strong influence on joint interaction, which may produce joint deformity presentation, or conversely the bone morphology is adapted due to the altered joint configuration from the deformity. Future work should include a longitudinal morphometric analysis to better understand possible changes or causes of joint deformity and bone remodeling over time in patients with PCFD.

This study has some limitations that must be taken into consideration. Firstly, it is unclear whether the changes observed in joint interaction are the cause or effect of foot deformity. Consequently, the results of this study may reflect a combination of both causes and effects of PCFD. To clarify this, further research, including longitudinal studies, is necessary. Second, controls were not age-matched, which may confound some between-group differences. This imbalance reflects the limited availability of truly asymptomatic, normally aligned CT datasets in older individuals, whereas surgically treated PCFD is more prevalent in older adults. Age-matched and/or longitudinal studies are needed. Thirdly, the CT used in this study is simulated weight-bearing and not fully loaded. Therefore, these factors may have influenced the results to a varying extent. Fourth, the sample size was modest, which may limit statistical power and generalizability. Larger, multi-center studies are warranted to confirm these findings across a broader spectrum of PCFD severity. Finally, this study focused on morphometric and joint interaction mapping from SW-CT and did not include formal correlations with clinical severity measures, PROMs, or surgical outcomes. Future studies should integrate correspondence-based joint metrics with clinical and radiographic severity measures and longitudinal outcomes to further establish clinical utility.

Although this study does not correlate imaging findings with PROMs or postoperative outcomes, it provides foundational 3D descriptors of multi-joint interactions in PCFD. Future pre- vs postoperative analyses and outcome-linked studies are warranted to establish clinical associations and to determine whether morphology-driven congruence changes and alignment-driven joint space changes predict correction durability or recurrence.

In conclusion, we conducted a study using MD SSM and JMA to analyze the joint interactions of the ankle, tibiofibular, talofibular, ST, TN, CC, NC, and TMT joint in both PCFD and asymptomatic control. Based on our findings, there were statistically significant differences between PCFD and controls in joint space distance and congruence index for almost every joint. The joint space distance was mainly influenced by alignment, and the congruence index was influenced by bone morphology. We believe that assessing multi-joint interactions in PCFD will lead to a better understanding of the pathophysiology and assist in surgical treatment planning. We strongly believe that further research is needed to effectively track the development of deformity, with a focus on expanding the MD SSM model.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (9.3MB, tif)
Supplementary Material 2 (9.4MB, tif)
Supplementary Material 3 (20.1KB, docx)

Acknowledgements

Not applicable.

Abbreviations

PCFD

Progressive collapsing foot deformity

ST

Subtalar

CT

Computed tomography

SSM

Statistical shape modeling

TN

Talonavicular

CC

Calcaneocuboid

NC

Naviculo-cuneiform

TMT

Tarso-metatarsal

JMA

Joint measurement analysis

SW-CT

Simulated weight-bearing computed tomography

MD SSM

Multi-domain statistical shape modeling

LCL

Lateral column lengthening

Authors’ contributions

T.M. and A.L. designed the study, T.M., H.K., A.T., and Y.T. collected the image data, T.M. performed segmentations and statistical shape model development, R.L., K.K., and A.L. provided image processing guidance, model troubleshooting, and quality assurance, T.M. and R.L. performed the data and statistical analysis, Y.T. and A.L. performed data interpretation, H.K., A.T., and Y.T. provided clinical insight, T.M. wrote the manuscript, and all authors reviewed the manuscript.

Funding

This work was supported by the Nakatani Foundation International Fellowship (2022H1002) and the JSSF International Fellowship Foundation. The funding sources were not involved in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication. Funding provided by the National Institutes of Health supported Dr. Lenz (NIH K01AR080221), and ShapeWorks software (NIH U24EB029011 and NIH R01AR076120).

Data availability

Data may be made available by the authors upon request.

Declarations

Ethical approval and consent to participate

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.

Consent for publication

Participants provided proper consent for the results of this study to be published.

Competing interests

The authors declare no competing interests.

Footnotes

Investigation Performed at the University of Utah, Salt Lake City, USA.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Pita-Fernandez S, Gonzalez-Martin C, Alonso-Tajes F, Seoane-Pillado T, Pertega-Diaz S, Perez-Garcia S, et al. Flat foot in a random population and its impact on quality of life and functionality. J Clin Diagn Res. 2017;11(4):LC22–7. 10.7860/JCDR/2017/24362.9697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Flores DV, Mejia Gomez C, Fernandez Hernando M, Davis MA, Pathria MN. Adult acquired flatfoot deformity: anatomy, biomechanics, staging, and imaging findings. Radiographics. 2019;39(5):1437–60. 10.1148/rg.2019190046. [DOI] [PubMed] [Google Scholar]
  • 3.Tang CYK, Ng KH, Lai J. Adult flatfoot. BMJ. 2020;368:m295. 10.1136/bmj.m295. [DOI] [PubMed] [Google Scholar]
  • 4.Myerson MS, Thordarson DB, Johnson JE, Hintermann B, Sangeorzan BJ, Deland JT, et al. Classification and nomenclature: progressive collapsing foot deformity. Foot Ankle Int. 2020;41(10):1271–6. 10.1177/1071100720950722. [DOI] [PubMed] [Google Scholar]
  • 5.Lee W, Tecle N, Knapp E, Awad H, Oh I. Biomechanical effects of subtalar joint fusion and medial ligament reconstruction in simulated progressive collapsing foot deformity. Foot Ankle Int. 2024;45(3):279–90. 10.1177/10711007231220940. [DOI] [PubMed] [Google Scholar]
  • 6.Thordarson DB, Schon LC, de Cesar Netto C, Deland JT, Ellis SJ, Johnson JE, et al. Consensus for the indication of lateral column lengthening in the treatment of progressive collapsing foot deformity. Foot Ankle Int. 2020;41(10):1286–8. 10.1177/1071100720950732. [DOI] [PubMed] [Google Scholar]
  • 7.Lintz F, Bernasconi A, Li S, Lalevee M, Fernando C, Barg A, et al. Diagnostic accuracy of measurements in progressive collapsing foot deformity using weight bearing computed tomography: a matched case-control study. Foot Ankle Surg. 2022;28(7):912–8. 10.1016/j.fas.2021.12.012. [DOI] [PubMed] [Google Scholar]
  • 8.Krahenbuhl N, Kvarda P, Susdorf R, Burssens A, Ruiz R, Barg A, et al. Assessment of progressive collapsing foot deformity using semiautomated 3d measurements derived from weightbearing CT scans. Foot Ankle Int. 2022;43(3):363–70. 10.1177/10711007211049754. [DOI] [PubMed] [Google Scholar]
  • 9.Lenz AL, Lisonbee RJ. Biomechanical insights afforded by shape modeling in the foot and ankle. Foot Ankle Clin. 2023;28(1):63–76. 10.1016/j.fcl.2022.11.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Cates J, Elhabian S, Whitaker R. ShapeWorks. Statistical Shape and Deformation Analysis, 2017;257–98.
  • 11.Peterson AC, Lisonbee RJ, Krahenbuhl N, Saltzman CL, Barg A, Khan N, et al. Multi-level multi-domain statistical shape model of the subtalar, talonavicular, and calcaneocuboid joints. Front Bioeng Biotechnol. 2022;10:1056536. 10.3389/fbioe.2022.1056536. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Knutson K, Peterson AC, Lisonbee RJ, Hintermann B, Krähenbühl N, Lenz AL. Joint coverage analysis in progressive collapsing foot deformity. J Orthop Res. 2023. 10.1002/jor.25543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Miyamoto T, Lisonbee RJ, Knutson K, Kurokawa H, Taniguchi A, Tanaka Y, et al. Describe anatomical differences in 3D bone morphology between female individuals with progressive collapsing foot deformity and asymptomatic controls. J Foot Ankle Surg. 2025. 10.1053/j.jfas.2025.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Takuma Miyamoto RJL, Kassidy Knutson, Hiroaki Kurokawa, Akira Taniguchi, Andrew E. Anderson, Charly L. Saltzman, Yasuhito Tanaka, Amy L. Lenz. Differences in three-dimensional foot alignment between patients with progressive collapsing foot deformity and asymptomatic controls in females. J Foot Ankle Res. 2025;In press.
  • 15.Gabrielli AS, Gale T, Hogan M, Anderst W. Bilateral symmetry, sex differences, and primary shape factors in ankle and hindfoot bone morphology. Foot Ankle Orthop. 2020;5(1):2473011420908796. 10.1177/2473011420908796. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Godoy-Santos AL, Schmidt EL, Chaparro F. What are the updates on epidemiology of progressive collapsing foot deformity? Foot Ankle Clin. 2021;26(3):407–15. 10.1016/j.fcl.2021.05.006. [DOI] [PubMed] [Google Scholar]
  • 17.Styner MA, Rajamani KT, Nolte L-P, Zsemlye G, Székely G, Taylor CJ, et al., editors. Evaluation of 3D correspondence methods for model building. Information Processing in Medical Imaging: 18th International Conference, IPMI 2003, Ambleside, UK, July 20–25, 2003 Proceedings 18; 2003: Springer. [DOI] [PubMed]
  • 18.Tomiwa K, Tanaka Y, Kurokawa H, Kadono K, Taniguchi A, Maliwankul K. Simulated weightbearing computed tomography for verification of radiographic staging of varus ankle osteoarthritis: a cross-sectional study. BMC Musculoskelet Disord. 2021. 10.1186/s12891-021-04618-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kimura T, Thorhauer ED, Sangeorzan BJ, Ledoux WR. Foot radiographic angle variation as a function of weightbearing magnitude. J Orthop Res. 2022;40(11):2620–5. 10.1002/jor.25283. [DOI] [PubMed] [Google Scholar]
  • 20.Schenker PS, Besl PJ, McKay ND. Method for registration of 3-D shapes. Sensor Fusion IV: Control Paradigms and Data Structures1992;586–606.
  • 21.Davies RH, Twining CJ, Cootes TF, Waterton JC, Taylor CJ. A minimum description length approach to statistical shape modeling. IEEE Trans Med Imaging. 2002;21(5):525–37. 10.1109/tmi.2002.1009388. [DOI] [PubMed] [Google Scholar]
  • 22.Styner M, Gerig G, Lieberman J, Jones D, Weinberger D. Statistical shape analysis of neuroanatomical structures based on medial models. Med Image Anal. 2003;7(3):207–20. 10.1016/s1361-8415(02)00110-x. [DOI] [PubMed] [Google Scholar]
  • 23.Cates J, Fletcher PT, Styner M, Shenton M, Whitaker R. Shape modeling and analysis with entropy-based particle systems. Inf Process Med Imaging. 2007;20:333–45. 10.1007/978-3-540-73273-0_28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Datar M, Cates J, Fletcher PT, Gouttard S, Gerig G, Whitaker R. Particle based shape regression of open surfaces with applications to developmental neuroimaging. Med Image Comput Comput Assist Interv. 2009;12(Pt 2):167–74. 10.1007/978-3-642-04271-3_21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Datar M, Gur Y, Paniagua B, Styner M, Whitaker R. Geometric correspondence for ensembles of nonregular shapes. Med Image Comput Comput Assist Interv. 2011;14(Pt 2):368–75. 10.1007/978-3-642-23629-7_45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Cates J, Fletcher PT, Styner M, Hazlett HC, Whitaker R. Particle-based shape analysis of multi-object complexes. Med Image Comput Comput Assist Interv. 2008;11(Pt 1):477–85. 10.1007/978-3-540-85988-8_57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Genovese CR, Lazar NA, Nichols T. Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage. 2002;15(4):870–8. 10.1006/nimg.2001.1037. [DOI] [PubMed] [Google Scholar]
  • 28.Cates J, Nevell L, Prajapati SI, Nelon LD, Chang JY, Randolph ME, et al. Shape analysis of the basioccipital bone in Pax7-deficient mice. Sci Rep. 2017;7(1):17955. 10.1038/s41598-017-18199-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Peterson AC. Static joint measurement analysis. 2022.
  • 30.Lisonbee RJ, Peterson AC, Lenz AL. Correspondence model-based approach for evaluating static and dynamic joint distance measurements. Comput Methods Biomech Biomed Engin. 2025. 10.1080/10255842.2025.2478527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Ateshian GA, Rosenwasser MP, Mow VC. Curvature characteristics and congruence of the thumb carpometacarpal joint: differences between female and male joints. J Biomech. 1992;25(6):591–607. 10.1016/0021-9290(92)90102-7. [DOI] [PubMed] [Google Scholar]
  • 32.Lenz AL, Krahenbuhl N, Peterson AC, Lisonbee RJ, Hintermann B, Saltzman CL, et al. Statistical shape modeling of the talocrural joint using a hybrid multi-articulation joint approach. Sci Rep. 2021;11(1):7314. 10.1038/s41598-021-86567-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Krahenbuhl N, Lenz AL, Lisonbee RJ, Peterson AC, Atkins PR, Hintermann B, et al. Morphologic analysis of the subtalar joint using statistical shape modeling. J Orthop Res. 2020;38(12):2625–33. 10.1002/jor.24831. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Behling AV, Rainbow MJ, Welte L, Kelly L. Chasing footprints in time - reframing our understanding of human foot function in the context of current evidence and emerging insights. Biol Rev Camb Philos Soc. 2023;98(6):2136–51. 10.1111/brv.12999. [DOI] [PubMed] [Google Scholar]
  • 35.Auch E, Barbachan Mansur NS, Alexandre Alves T, Cychosz C, Lintz F, Godoy-Santos AL, et al. Distal tibiofibular syndesmotic widening in progressive collapsing foot deformity. Foot Ankle Int. 2021;42(6):768–75. 10.1177/1071100720982907. [DOI] [PubMed] [Google Scholar]
  • 36.Evans D. Calcaneo-valgus deformity. J Bone Joint Surg Br. 1975;57(3):270–8. [PubMed] [Google Scholar]
  • 37.Fallon Verbruggen F, Killen BA, Burssens A, Boey H, Vander Sloten J, Jonkers I. Unique shape variations of hind and midfoot bones in flatfoot subjects-a statistical shape modeling approach. Clin Anat. 2022. 10.1002/ca.23969. [DOI] [PubMed] [Google Scholar]
  • 38.Flury A, Hasler J, Beeler S, Imhoff FB, Wirth SH, Viehofer A. Talus morphology differs between flatfeet and controls, but its variety has no influence on extent of surgical deformity correction. Arch Orthop Trauma Surg. 2022;142(11):3103–10. 10.1007/s00402-021-03925-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Zhang L, Peng X, He S, Zhou X, Yi G, Tang X, et al. Association between subtalar articular surface typing and flat foot deformity: which type is more likely to cause flat foot deformity. BMC Musculoskelet Disord. 2021;22(1):979. 10.1186/s12891-021-04872-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (9.3MB, tif)
Supplementary Material 2 (9.4MB, tif)
Supplementary Material 3 (20.1KB, docx)

Data Availability Statement

Data may be made available by the authors upon request.

Articles from Journal of Orthopaedic Surgery and Research are provided here courtesy of BMC

RESOURCES