Three-Dimensional Assessment of Charcot Neuroarthropathy Deformities: Comparison of Standard 2D vs Patient-Specific 3D Measurements

Abstract

Background:

The purpose of this study is to compare foot and ankle deformity measurements obtained from 2-dimensional (2D) radiographs vs 3-dimensional (3D) modalities in Charcot neuroarthropathy (CN) feet by using a patient-specific coordinate system.

Methods:

This retrospective study reviewed foot and ankle imaging for 25 patients with a diagnosis of CN of the lower extremity, type 2 diabetes mellitus with diabetic neuropathy, or lower limb neuropathy. Radiographs and either computed tomography (CT) or magnetic resonance imaging (MRI) scans were obtained for each patient and used to make angle and distance measurements used clinically to describe deformity. 2D measurements were obtained using standard methods, involving annotating planar radiographs. The 3D measurement procedure began by manually placing fiducials on anatomic landmarks. Then, a custom-built code was used to automatically transform the foot into a patient-specific anatomic coordinate system and calculate all angle and distance measurements. Each scan was measured by 2 observers and intraclass correlation was calculated for each imaging type.

Results:

The average age of the patients was 61 years, with 92% being White and 88% having diabetic neuroarthropathy. Measurements for anteroposterior talocalcaneal angle and lateral column height were larger when measured on MRI (91.1 ± 16.7 degrees vs 29.1 ± 2.8 degrees, P = .004) and CT (78.6 ± 18.5 degrees vs 24.6 ± 2.7 degrees, P = .020) compared with radiographic measurements. Additionally, MRI demonstrated significantly greater interobserver reliability for the talocalcaneal angle (0.74 vs 0.19, 95% CI 0.11, 0.96), suggesting improved detection of hindfoot valgus compared with radiographs, whereas CT reliability was comparable to plain radiographs.

Conclusion:

Larger measurements and higher interobserver reliability for the talocalcaneal angle on 3D modalities suggest that a patient-specific 3D approach may improve detection of transverse-plane malalignment in Charcot neuroarthropathy.

Graphical Abstract

This is a visual representation of the abstract.

Keywords

3D characterization 3D images Charcot neuroarthropathy diabetic neuroarthropathy osteoarthropathy

Level of Evidence: Level III, retrospective comparative study.

Introduction

Charcot neuroarthropathy (CN) is a chronic condition characterized by bone and soft tissue damage in a neuropathic joint, often resulting in collapse, fracture, and joint destruction.^12,19 The foot is commonly affected because of high stresses from weightbearing, making it susceptible to progressive 3-dimensional (3D) foot and ankle deformities.¹² Radiographic measurements can be used to follow the progression and describe the severity of foot deformities, as well as assist with surgical planning.^13,19,24 Current standard practice involves using 2-dimensional (2D) radiographic measurements to evaluate the severity of 3D deformities.⁷ However, the accuracy of these measurements is hindered by factors such as foot positioning in relation to the x-ray beam, bone superimposition, whether the patient is bearing weight, and the inability to assess the coronal plane effectively.^3,5,17,26 Using 3D modalities allows for the visualization of the deformity in multiple planes and detection of the rotational component of these deformities, but quantifiable results may be difficult to obtain or interpret.^6,30,33 The inherent weaknesses of 2D imaging technology and measurements highlight the challenges associated with using radiographs to accurately depict complex 3D bone deformities, whereas 3D imaging measurements may provide a more accurate and comprehensive assessment of deformity in CN.

Studies have explored the potential benefits of using 3D imaging modalities for generating measurements in foot and ankle surgery. Broos et al⁵ found that in healthy feet, the measurements obtained from computed tomography (CT) scans were more reproducible and precise than those obtained from radiographs. Similarly, Sangoi et al²⁶ reported that automated measurements from 3D computer models were comparable to manual measurements from 2D radiographs and could be used reliably in normal feet. Furthermore, comparisons between 3D and 2D measurements have been made in other foot and ankle pathologies. In patients with adult acquired flatfoot deformity, researchers found that measurements obtained from 3D weightbearing (WB) CT have similar reliability as those obtained from 2D WB radiographs and may provide better characterization of deformity.⁸ The demonstration of clinical utility in these cases indicates the potential application for 3D-based measurements in CN.

Several approaches have been used to obtain measurements from 3D imaging modalities. Broos et al⁵ used 3D scans with segmented bone structures, from which an automated model derived anatomical long axes and calculated angles. However, this involves segmentation, which requires additional time and expertise, and introduces subjectivity, especially in scans of patients with bone pathologies like CN. Other studies have employed statistical shape modeling to characterize foot and ankle bony morphology, which is useful for assessing anatomic variation and disease-related changes in bone shape.^15,16,21 Although apparently valuable to research, statistical shape modeling may be difficult to implement in the clinic as it introduces new forms of measuring pathologies that are unfamiliar to most physicians, and also requires segmentation and software expertise.

To address these challenges, we propose a method that involves point-annotating 3D images using the same landmarks used in current 2D measurements. The approach also incorporates a patient-specific coordinate system to normalize measurements to an individual’s anatomy, thereby minimizing variability and error due to foot positioning and potentially enabling more consistent comparisons across patients. This framework also permits the integration of clinically relevant measurements already used by clinicians to monitor disease progression and inform surgical planning. To our knowledge, no prior studies have evaluated the clinical advantages of using such a method for characterizing deformities in CN, which involves complex pathologies that may necessitate improved measurement methods. This study aimed to compare foot and ankle deformity measurements obtained from 2D radiographs vs 3D imaging modalities, including nonweightbearing (NWB) CT and MRI, in CN feet by using a patient-specific coordinate system.

Methods and Materials

Patient Selection

This retrospective study, which reviewed foot and ankle imaging for 25 patients (25 feet) between 2007 and 2022 within a single health care system, was conducted following approval from the Institutional Review Board. Patients with a diagnosis of Charcot Neuroarthropathy of the lower extremity, type 2 diabetes mellitus with diabetic neuropathy, or lower limb neuropathy who underwent casting, below the knee amputation, exostectomy, or arthrodesis with internal or external fixation were included in this study. Patients were excluded if they did not have a CT or MRI of the forefoot, hindfoot, or ankle within 1 year of casting and prior to any surgical intervention. Patients with scans that did not include the distal tibia or hindfoot were also excluded. Inclusion of the distal tibia and hindfoot was required to establish the patient-specific coordinate system, which relies on anatomic landmarks in the tibia for proper orientation. Patient charts were reviewed for demographic data including gender, age, race, and living status.

2D Radiographic Measurements

Orthogonal WB and NWB images were performed using a certified radiology technologist. Digitized, calibrated images were then uploaded to the electronic radiology viewer, Phillips PACS Software (release 4.7; Philips Healthcare, Cambridge, MA). Radiographs for all patients that met the inclusion criteria (n = 25) were collected. Fifteen patient images were obtained in a WB position, whereas the remaining (n = 10) were NWB. Whether a patient had WB or NWB radiographs was based on provider preference and the patient’s ability to bear weight. Three observers, instructed by an experienced foot and ankle surgeon, obtained 7 angle and 2 distance measurements using measurement tools on Phillips PACS Software (release 4.7; Philips Healthcare, Cambridge, MA) (Table 1). These measurements were previously described for diagnoses of CN or related deformities.^7,13,32

Table 1.

Measurements Collected from Two- and Three-Dimensional Imaging Modalities for Each Patient.

Radiographic Measurement^12,26	Definition	Reference points for Fiduciary Markers
Lateral talar–first metatarsal angle (Meary angle)	Angle formed from a line bisecting the talar body and neck and a line bisecting the first metatarsal	Center of talar body, center of talar neck, center of proximal first-metatarsal shaft, center of distal first-metatarsal shaft
Anteroposterior talocalcaneal angle (Kite angle)	Angle formed from a line bisecting the talar body and neck and line drawn along the longitudinal axis of the lateral border of the calcaneus	Center of talar body, center of talar neck, lateral border of proximal calcaneus, lateral border of distal calcaneus
Calcaneal pitch	Angle between line from the plantar aspect of the calcaneal tuberosity to the plantar aspect of the fifth-metatarsal head and a line extending from the most plantar aspect of the calcaneal tuberosity to the most plantar aspect of the anterior process of the calcaneus	Plantar aspect of the calcaneal tuberosity, plantar aspect of the fifth-metatarsal head, plantar aspect of the anterior process of the calcaneus
Calcaneal–fifth metatarsal angle	Angle between line bisecting the fifth metatarsal and a line extending from the most plantar aspect of the calcaneal tuberosity to the most plantar aspect of the anterior process of the calcaneus	Center of proximal fifth-metatarsal shaft, center of distal fifth-metatarsal shaft, plantar aspect of the calcaneal tuberosity, plantar aspect of the anterior process of the calcaneus
Talar declination angle	Angle between a line bisecting the talar body and neck and the reference line from the plantar aspect of the calcaneal tuberosity to the plantar aspect of the fifth-metatarsal head	Center of talar body, center of talar neck, plantar aspect of the calcaneal tuberosity, plantar aspect of the fifth-metatarsal head
Tibiotalar angle	Angle formed from a line bisecting the distal tibia and a line bisecting the talar body and neck	Center of talar body, center of talar neck, 2 points along center of distal tibia
Talonavicular coverage angle	Angle between a line of the articular surfaces of the talar head and a line of the articular surfaces of the proximal navicular bone	Medial aspect of articular surface of talar head, lateral aspect of articular surface of talar head, medial aspect of articular surface of proximal navicular bone, lateral aspect of articular surface of proximal navicular bone
Lateral column height	Perpendicular distance from the plantar aspect of the cuboid to a line drawn from the plantar aspect of the calcaneal tuberosity to the plantar aspect of the fifth-metatarsal head	Plantar aspect of the cuboid, plantar aspect of the calcaneal tuberosity, plantar aspect of the fifth-metatarsal head
Medial column height	Perpendicular distance from the plantar aspect of the first tarsometatarsal joint to a line drawn from the plantar aspect of the calcaneal tuberosity to the plantar aspect of the fifth-metatarsal head	Plantar aspect of the first tarsometatarsal joint, plantar aspect of the calcaneal tuberosity, plantar aspect of the fifth-metatarsal head

3D Image Measurements

Either a CT or MRI was obtained for each patient and 3D volume rendering was produced using 3D Slicer (Kitware Inc, Clifton Park, NY).¹⁰ Five observers, instructed and overseen by a board-certified orthopaedic surgeon with fellowship training in foot and ankle, placed fiducial markers on anatomic landmarks in 3D space using the generated volume and the axial, sagittal, and coronal sequences. Fiducial markers were placed on 26 landmarks to (1) define a patient-specific anatomic coordinate system based on the International Society of Biomechanics (ISB) recommendations for standard reporting and (2) define landmarks for the same measurements described for 2D radiographic analysis.^11,31 Two observers, masked to the fiducial locations of other observers, were randomly assigned to each scan. The anatomic landmarks and obtained measurements are listed in Table 1 and Figure 1.

Figure 1.

This illustration depicts the anatomic landmarks used for fiducial marker placement in 3D imaging of the foot, essential for evaluating deformities in patients with Charcot neuroarthropathy. The landmarks were to calculate the measurements traditionally taken from 2D radiographs. Anatomic landmarks used for the ankle are not included in this figure.

A custom-built code (MathWorks, Notham, MA) was used to transform all fiducial coordinates into a patient-specific coordinate system based on their anatomy. In short, the origin was defined as the midpoint between the medial and lateral malleoli and the orthogonal axes were defined as the following: (1) y axis extending from the lateral to medial malleolus, (2) x axis extending from inferior to superior defined by the center of the tibial articular surface to a point 1 to 4 cm along the tibial longitudinal axis, and (3) z axis extending from anterior to posterior.

The original and transformed points for each scan were plotted and labeled based on their anatomic landmark for visual validation (Supplemental Figure 1). Once coordinates were transformed, calculations were automatically performed using the custom code. For angular measurements, only the 2 coordinates relevant to the plane in which the conventional 2D measurement is made were used. For example, measurements that are conventionally made using sagittal plane radiographs would be in the x-z plane; hence, only x and z coordinates were used. For distance measurements, values were obtained for both 3D vectors, and vectors in the plane of the conventional measurement 2D. The custom code was validated on CT images of healthy feet (Supplemental Materials, Table 1).

Statistical Analysis

The measurements were represented as means ± SEs of the mean (SEMs). These measurements were then compared using paired t test to evaluate any variation in the 2 methods. The intraclass correlation coefficient (ICC) were calculated to evaluate the consistency in measurements across raters. These values were compared between the 2D scan and 3D scan groups to compare consistency between the 2 methods. These comparisons are expressed as 95% CIs; if the CI did not include zero, then the difference in ICC values was significant. An additional analysis was performed to compare intrarater reliability between 2D and 3D measurements for each WB status subgroup (Supplemental Materials, Tables 2 and 3). Of the 12 patients who underwent NWB CT, 2D radiographic measurements were made from WB radiographs in 6 patients and NWB radiographs in the remaining 6. Among the 14 patients who had NWB MRI, WB radiographs were available for 9 patients, whereas 5 patients had NWB radiographs.

Results

There were 33 patients who met inclusion criteria, but only 25 patients (25 feet) had 3D imaging scans that included the hindfoot region for use of coordinate system generation. The average age of the patients included was 61 (range 19-88) years; 92% (23/25) of patients were White, whereas the remaining were listed as other or multiple races. There were 88% (22/25) of patients with diabetic neuroarthropathy and 12% (3/25) with Charcot secondary to other etiologies.

MRI Measurements

The 3D-MRI data showed significantly greater anteroposterior (AP) talocalcaneal angles than radiographs (91.1 ± 16.7 degrees vs 29.1 ± 2.8 degrees, P = .004). Additionally, 3D-MRI data showed significantly greater lateral column height than radiographs (22.1 ± 5.0 mm vs 6.2 ± 3.2 mm, P = .041). Conversely, the 3D-MRI data showed significantly lower talar declination angles than radiographs (23.1 ± 3.0 degrees vs 31.6 ± 3.0 degrees, P = .012). Similarly, the 3D-MRI data showed significantly lower tibiotalar angles than radiographs (92.8 ± 1.1 degrees vs 128.7 ± 3.1 degrees, P < .001). There were no statistically significant differences in measurements for lateral talar–first metatarsal angle, calcaneal pitch, calcaneal–fifth metatarsal angle, talar declination angle, talonavicular coverage angle, and medial column height (Figure 2A and B).

Figure 2.

(A) Comparison of mean angle and SE of the mean for angular measurements between MRI and radiographs. Talocalcaneal, talar declination, and tibiotalar angles are significantly different. (B) Comparison of mean distance and SE of the mean for height measurements between MRI and radiographs. Lateral column height is significantly different. (C) Comparison of mean angle and SE of the mean for angular measurements between CT and radiographs. Talocalcaneal, talar declination, and tibiotalar angles are significantly different. (D) Comparison of mean distance and SE of the mean for height measurements between CT and radiographs. Lateral column height is significantly different.

MRI Measurement Reliability

The measurement interobserver reliability was significantly higher for MRI than radiographs for AP talocalcaneal angle measurements (0.74 vs 0.19, 95% CI 0.11, 0.96). Conversely, interobserver reliability was significantly lower for MRI than radiographs for medial column height (0.01 vs 0.93, 95% CI −0.98, −0.25) and lateral column height (0.58 vs 0.92, 95% CI −0.93, −0.07). Similarly, the interobserver reliability was also significantly lower for MRI than radiographs for calcaneal–fifth metatarsal angle (0.02 vs 0.96, 95% CI −0.99, −0.19) and tibiotalar angle (0.02 vs 0.70, 95% CI −0.86, −0.33). There were no statistically significant differences in ICC values for lateral talar–first metatarsal angle, calcaneal pitch, talar declination angle, and talonavicular coverage angle when comparing MRI to radiography (Table 2).

Table 2.

Ranges for 95% Confidence Intervals for ICC values between MRI and Radiography.

Measurement	ICC CI
Talar–first metatarsal angle	−0.75, 0.21
Talocalcaneal angle	−0.98, −0.25
Calcaneal pitch angle	−0.93, −0.07
Calcaneal fifth metatarsal angle	0.11, 0.96
Talar declination angle	−0.45, 0.23
Tibiotalar angle	−0.69, 0.16
Talonavicular coverage angle	−0.75, 0.94
Medial column height (2D)	−0.86, −0.33
Cuboid lateral column (2D)	−0.59, 0.34

Abbreviations: ICC, intraclass correlation coefficient; MRI, magnetic resonance imaging.

CT Measurements

The AP talocalcaneal angles were significantly greater in 3D-CT than in radiographs (78.6 ± 18.5 degrees vs 24.6 ± 2.7 degrees, P = .020). Additionally, the lateral column heights were significantly larger in 3D-CT compared with radiographs (22.6 ± 6.7 mm vs 3.8 ± 2.2 mm, P = .046). Conversely, the talar declination angles were significantly lower in 3D-CT compared to radiographs (25.3 ± 3.3 degrees vs 36.4 ± 3.9 degrees, P = .003). Similarly, the tibiotalar angles were significantly smaller in 3D-CT than in radiographs (89.9 ± 2.1 degrees vs 131.8 ± 5.0 degrees, P < .001). There were no statistically significant differences in lateral talar–first metatarsal angle, calcaneal pitch, calcaneal–fifth metatarsal angle, talar declination angle, talonavicular coverage angle, and medial column height (Figure 2C and D).

CT Measurement Reliability

The interobserver reliability was significantly lower for CT than radiographs for medial column height (0.51 vs 0.97, 95% CI −0.98, −0.09) and calcaneal pitch angle (0.80 vs 0.95, 95% CI −0.75, −0.07). There were no statistically significant differences in interobserver reliability for lateral talar–first metatarsal angle, AP talocalcaneal angle, calcaneal pitch, calcaneal–fifth metatarsal angle, talar declination angle, talonavicular coverage angle, and medial column height when comparing CT to radiography (Table 3).

Table 3.

Ranges for 95% Confidence Intervals for ICC values between CT and Radiography.

Measurement	ICC CI
Talar–first metatarsal angle	−0.76, 0.00
Talocalcaneal angle	−0.98, −0.09
Calcaneal pitch angle	−0.85, 0.58
Calcaneal–fifth metatarsal angle	−0.68, 0.74
Talar declination angle	−0.75, −0.07
Tibiotalar angle	−0.68, 0.16
Talonavicular coverage angle	−0.45, 0.76
Medial column height (2D)	−0.79, 0.10
Cuboid lateral column (2D)	−0.85, 0.33

Abbreviations: CT, computed tomography; ICC, intraclass correlation coefficient.

In summary, 3D imaging modalities yielded larger values for the AP talocalcaneal angle and lateral column height, whereas talar declination and lateral tibiotalar angles were smaller compared with 2D radiographs. MRI demonstrated significantly higher interobserver reliability for the talocalcaneal angle but lower reliability for medial and lateral column heights. CT showed significantly lower interobserver reliability for medial column height and calcaneal pitch angle.

Weightbearing Status Stratification

MRI measurements demonstrated higher interobserver reliability for AP talocalcaneal angle and calcaneal pitch compared to both WB and NWB radiographs. In contrast, CT had higher reliability for lateral column height and talar declination compared with both WB and NWB radiographs. Complete results can be found in Supplemental Materials.

Discussion

Several measurements in our study were found to be significantly different between 2D radiographs and 3D imaging modalities, suggesting that traditional radiographs may underestimate deformities in CN that are more clearly visualized with 3D imaging using a patient-specific coordinate system. One measurement that was significantly larger on 3D images compared to radiographs was the talocalcaneal angle, suggesting more apparent hindfoot valgus on 3D modalities.⁹ Conversely, talar declination angles and lateral tibiotalar angles were smaller on 3D imaging, indicating less midfoot deformity and less collapse of the medial longitudinal arch.³² Additionally, greater lateral column heights were observed on 3D images, suggesting reduced lateral arch collapse.⁷ MRI’s higher interobserver reliability for the talocalcaneal angle underscores the value of 3D imaging for transverse-plane assessment; however, its reliability was lower for medial and lateral column heights. Furthermore, CT demonstrated significantly lower interobserver reliability for medial column height and calcaneal pitch angle.

Interestingly, this study revealed that certain angular measurements differed significantly when obtained from 2D radiographs versus 3D imaging modalities, whereas others did not. We hypothesize that this is because 2D radiographs are associated with challenges including reliance of foot position relative to x-ray beam and bone superimposition, both of which can result in flawed images.^2,13,19,24 Improved visualization over 2D provided by 3D modalities may have contributed to the observed measurement differences, as these modalities enable observers to more clearly visualize anatomic landmarks required for making these measurements. Our findings are consistent with prior studies that have also observed measurement discrepancies between 2D and 3D imaging. Shakoor et al²⁸ compared WB radiographs to WBCTs in patients with acquired flatfoot deformity and found that although some angular measurements were similar across modalities, others differed significantly. They hypothesized that some measurements likely varied because of modality-specific factors such as foot positioning in relation to the x-ray beam, whereas other measurements might be impacted by WB status. Similarly, a study by Richter et al²³ evaluated patients with flatfoot deformity and compared several angular measurements across WB radiographs, NWB CT, and WBCT. They included comparisons between WB radiographs and NWB CT and identified modality-related and weightbearing-related differences in specific measurements.

A notable finding in our study was the mean talocalcaneal angle, a measure of deformity in the transverse plane and a reliable indicator of hindfoot alignment.^7,9,20 This angle was significantly larger on 3D imaging modalities compared with 2D radiographs, indicating greater apparent hindfoot valgus on 3D. In a study by DuBois et al,⁹ patients with hindfoot valgus were found to be 27 times more likely to develop midtarsal breakdown. This relationship is hypothesized to result from the internal rotation of the talonavicular joint caused by valgus alignment of the subtalar joint, which increases stress on the midtarsal capsuloligamentous structures.^1,25,27 As such, accurate detection of hindfoot valgus may enable clinicians to identify patients at higher risk for midfoot collapse and ulceration, prompting closer monitoring or earlier intervention to help prevent further deformity.

Furthermore, the differences in ICC were also significant, with MRI demonstrating a significantly higher interobserver reliability than radiographs and CT demonstrating a comparable reliability with radiographs. This finding suggests that our method with 3D scans may be better for detecting transverse plane deformities. Similar discrepancies in talocalcaneal angle obtained from radiographs compared to 3D images have been identified in clubfoot literature. Ippolito et al¹⁴ found that the talocalcaneal angle from radiographs did not accurately describe the true relationship between the calcaneus and talus apparent on 3D reconstructions. They hypothesized that this inaccuracy can be attributed to false radiographic projection of the talus. Our method, providing a more accurate talocalcaneal angle using 3D images, could enable the detection of correlations between transverse deformity measurements and clinical outcomes. Additionally, it may facilitate surgical planning by providing a clearer understanding of the deformity and the required correction.

Another notable finding in this study was that measurements of sagittal plane deformities, including talar declination angle, lateral tibiotalar angle, and lateral column height, were consistently smaller on MRI and CT compared to 2D radiographs, indicating less midfoot collapse when assessed with 3D imaging.^4,7,32 Accurate characterization of sagittal plane deformities is clinically important, as these measurements have been linked to ulceration risk in patients with CN, and foot ulcers are reported to increase the risk of amputations in diabetics.^22,29 If 3D imaging underestimates the extent of midfoot collapse, clinicians may be less likely to initiate close monitoring, offloading strategies, or timely reconstructive interventions.

Additionally, we found that radiographs were more reproducible than 3D modalities for these sagittal plane measurements, as indicated by lower or comparable ICC values. The lower reproducibility observed in sagittal plane measurements may stem from challenges in identifying distorted anatomic landmarks in Charcot anatomy, as well as from variability introduced by the patient-specific coordinate system, which can amplify small input errors into larger deviations in derived measurements. However, we believe that the advantages of orienting each foot in a patient-specific coordinate system outweighs the reproducibility errors, while also eliminating the need for time-consuming 3D reconstructions. Another potential source of lower reproducibility was the involvement of 5 observers making the 3D imaging measurements; a reduced number of observers would likely have improved accuracy.

One strength of this study is its novel use of a patient-specific coordinate system to obtain measurements, thereby avoiding any malalignment present in the positioning of the foot during the scan. However, there are several limitations to consider. First, CN is an uncommon disease; therefore, the sample size is limited. Moreover, the retrospective nature of the study prevented us from controlling for the type of 3D imaging modality and the timing of the 3D image acquisition relative to the disease process. This resulted in 2 different subsets of patients: those that had an MRI and those that had a CT scan. Additionally, the retrospective design limited us to using existing scans, leading to the inclusion of both WB and NWB 2D radiographs, whereas all 3D imaging was NWB. This discrepancy may have contributed to some of the differences observed in our measurements, and the effects of WB status on deformity quantification warrant further study. Furthermore, the patient-specific coordinate system was validated using healthy feet only; therefore, future work will need to focus on validating its accuracy and reliability in Charcot-affected anatomy, which may present with more complex deformities. Finally, because this was a single-center, retrospective study of patients referred for advanced imaging, selection bias cannot be excluded; future multicenter, prospective cohorts that include weightbearing CT or MRI will be needed to confirm generalizability.

The current study’s findings lay the foundation for future assessments and comparisons with weightbearing models. Future prospective studies with standardized protocols for imaging modality and WB status are needed to address the limitations and validate the findings. This study provides the groundwork for comprehensive understanding of patient-specific 3D anatomy, which is crucial for informed surgical decision making.¹⁸

Conclusion

The aim of this study was to apply a novel approach to establish initial nonweightbearing 3D measurements for patients with CN and compare them to traditional 2D radiographs. The present findings highlight the unique presentation of 3D foot deformities in CN such as those that describe transverse plane alignment. Larger measurements and higher interobserver reliability for the talocalcaneal angle on 3D imaging highlight its potential clinical utility in revealing hindfoot valgus that may be missed on radiographs. This could potentially allow for earlier identification of hindfoot valgus and enable better detection of progression or timely intervention to prevent midtarsal breakdown. These findings emphasize the importance of incorporating 3D imaging modalities—particularly for evaluating hindfoot valgus and transverse plane deformities—into routine assessment of Charcot foot to guide earlier intervention and more precise surgical planning.

Supplemental Material

sj-docx-2-fai-10.1177_10711007251351316 – Supplemental material for Three-Dimensional Assessment of Charcot Neuroarthropathy Deformities: Comparison of Standard 2D vs Patient-Specific 3D Measurements

Supplemental material, sj-docx-2-fai-10.1177_10711007251351316 for Three-Dimensional Assessment of Charcot Neuroarthropathy Deformities: Comparison of Standard 2D vs Patient-Specific 3D Measurements by Avani A. Chopra, Zachary Adam Koroneos, Michaela D. Pitcher, Christian Benedict, Peter Tortora, Taylor Lan, Michael Levidy, Allen Kunselman and Michael Aynardi in Foot & Ankle International

Supplemental Material

sj-pdf-1-fai-10.1177_10711007251351316 – Supplemental material for Three-Dimensional Assessment of Charcot Neuroarthropathy Deformities: Comparison of Standard 2D vs Patient-Specific 3D Measurements

Supplemental material, sj-pdf-1-fai-10.1177_10711007251351316 for Three-Dimensional Assessment of Charcot Neuroarthropathy Deformities: Comparison of Standard 2D vs Patient-Specific 3D Measurements by Avani A. Chopra, Zachary Adam Koroneos, Michaela D. Pitcher, Christian Benedict, Peter Tortora, Taylor Lan, Michael Levidy, Allen Kunselman and Michael Aynardi in Foot & Ankle International

Footnotes

Ethical Approval

This study received ethical approval from the Penn State University IRB (approval no. 00020092).

Declaration of Conflicting Interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Michael Aynardi, MD, reports disclosures related to manuscript from small project grant from American Orthopaedic Foot & Ankle Society (grant number 2022-22209-S) and general disclosures as consultant for Arthrex, consultant for Zimmer Biomet, consultant for Stryker, AOFAS committee member. Disclosure forms for all authors are available online.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a grant from the American Orthopaedic Foot & Ankle Society with funding from the Orthopaedic Foot & Ankle Foundation (grant number 2022-22209-S).

Consent to participate

This is an IRB-approved retrospective study, all patient information was deidentified and patient consent was not required. Patient data will not be shared with third parties.

Consent for publication

Not applicable

ORCID iDs

Avani A. Chopra, MD,

Zachary Adam Koroneos,

Michael Aynardi, MD,

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Supplemental Material

Supplementary material is available online with this article.

References

Apostle

Coleman

Sangeorzan

BJ.

Subtalar joint axis in patients with symptomatic peritalar subluxation compared to normal controls. Foot Ankle Int. 2014;35(11):1153-1158. doi:10.1177/1071100714546549

Barg

Amendola

Henninger

Kapron

Saltzman

Anderson

AE.

Influence of ankle position and radiographic projection angle on measurement of supramalleolar alignment on the anteroposterior and hindfoot alignment views. Foot Ankle Int. 2015;36(11):1352-1361. doi:10.1177/1071100715591091

Baverel

Brilhault

Odri

Boissard

Lintz

Influence of lower limb rotation on hindfoot alignment using a conventional two-dimensional radiographic technique. Foot Ankle Surg. 2017;23(1):44-49. doi:10.1016/j.fas.2016.02.003

Bevan

WPC

Tomlinson

MPW.

Radiographic measures as a predictor of ulcer formation in diabetic Charcot midfoot. Foot Ankle Int. 2008;29(6):568-573. doi:10.3113/FAI.2008.0568

Broos

Berardo

Dobbe

JGG

Maas

Streekstra

Wellenberg

RHH

. Geometric 3D analyses of the foot and ankle using weight-bearing and non weight-bearing cone-beam CT images: the new standard? Eur J Radiol. 2021;138:109674. doi:10.1016/j.ejrad.2021.109674

Burssens

Peeters

Peiffer

, et al. Reliability and correlation analysis of computed methods to convert conventional 2D radiological hindfoot measurements to a 3D setting using weightbearing CT. Int J Comput Assist Radiol Surg. 2018;13(12):1999-2008. doi:10.1007/s11548-018-1727-5

Cates

Tenley

Cook

Kim

PJ.

A systematic review of angular deformities in Charcot neuroarthropathy. J Foot Ankle Surg. 2021;60(2):368-373. doi:10.1053/j.jfas.2020.10.003

de Cesar Netto

Schon

Chinanuvathana

Lintz

Da Fonseca

. Comparison between weight bearing radiographs and weight bearing ConeBeam CT examinations in the assessment of adult acquired flatfoot deformity. Foot Ankle Orthop. 2017;2(3):2473011417S000152. doi:10.1177/2473011417S000152

DuBois

Cates

O’Hara

Lamm

Wynes

Coronal hindfoot alignment in midfoot Charcot neuroarthropathy. J Foot Ankle Surg. 2022;61(5):1039-1045. doi:10.1053/j.jfas.2022.01.011

10.

Fedorov

Beichel

Kalpathy-Cramer

, et al. 3D Slicer as an image computing platform for the Quantitative Imaging Network. Magn Reson Imaging. 2012;30(9):1323-1341. doi:10.1016/j.mri.2012.05.001

11.

Grood

Suntay

WJ.

A joint coordinate system for the clinical description of three-dimensional motions: application to the knee. J Biomech Eng. 1983;105(2):136-144. doi:10.1115/1.3138397

12.

Hastings

Johnson

Strube

, et al. Progression of foot deformity in Charcot neuropathic osteoarthropathy. J Bone Joint Surg Am. 2013;95(13):1206-1213. doi:10.2106/JBJS.L.00250

13.

Hastings

Sinacore

Mercer-Bolton

, et al. Precision of foot alignment measures in Charcot arthropathy. Foot Ankle Int. 2011;32(9):867-872. doi:10.3113/FAI.2011.0867

14.

Ippolito

Fraracci

Farsetti

De Maio

Validity of the anteroposterior talocalcaneal angle to assess congenital clubfoot correction. Am J Roentgenol. 2004;182(5):1279-1282. doi:10.2214/ajr.182.5.1821279

15.

Krähenbühl

Lenz

Lisonbee

, et al. Morphologic analysis of the subtalar joint using statistical shape modeling. J Orthop Res. 2020;38(12):2625-2633. doi:10.1002/jor.24831

16.

Lenz

Lisonbee

RJ.

Biomechanical insights afforded by shape modeling in the foot and ankle. Foot Ankle Clin. 2023;28(1):63-76. doi:10.1016/j.fcl.2022.11.001

17.

Lintz

Beaudet

Richardi

Brilhault

Weight-bearing CT in foot and ankle pathology. Orthop Traumatol Surg Res. 2021;107(1S):102772. doi:10.1016/j.otsr.2020.102772

18.

Wang

Liu

, et al. Applications of mixed reality technology in orthopedics surgery: a pilot study. Front Bioeng Biotechnol. 2022;10:740507. doi:10.3389/fbioe.2022.740507

19.

Madan

Pai

DR.

Charcot neuroarthropathy of the foot and ankle. Orthop Surg. 2013;5(2):86-93. doi:10.1111/os.12032

20.

Masquijo

Tourn

Torres-Gomez

Reliability of the talocalcaneal angle for the evaluation of hindfoot alignment. Rev Esp Cir Ortop Traumatol (Engl Ed). 2019;63(1):20-23. doi:10.1016/j.recot.2018.08.003

21.

Melinska

Romaszkiewicz

Wagel

Antosik

Sasiadek

Iskander

DR.

Statistical shape models of cuboid, navicular and talus bones. J Foot Ankle Res. 2017;10(1):6. doi:10.1186/s13047-016-0178-x

22.

Pecoraro

Reiber

Burgess

EM.

Pathways to diabetic limb amputation: basis for prevention. Diabetes Care. 1990;13(5):513-521. doi:10.2337/diacare.13.5.513

23.

Richter

Seidl

Zech

Hahn

PedCAT for 3D-imaging in standing position allows for more accurate bone position (angle) measurement than radiographs or CT. Foot Ankle Surg. 2014;20(3):201-207. doi:10.1016/j.fas.2014.04.004

24.

Rogers

Bevilacqua

NJ.

Imaging of the Charcot foot. Clin Podiatr Med Surg. 2008;25(2):263-274. doi:10.1016/j.cpm.2008.01.002

25.

Sangeorzan

Subtalar joint biomechanics. Foot Ankle Clin. 2018;23(3):341-352. doi:10.1016/j.fcl.2018.04.002

26.

Sangoi

Ranjit

Bernasconi

, et al. 2D manual vs 3D automated assessment of alignment in normal and Charcot-Marie-tooth cavovarus feet using weightbearing CT. Foot Ankle Int. 2022;43(7):973-982. doi:10.1177/10711007221084308

27.

Sarrafian

SK.

Biomechanics of the subtalar joint complex. Clin Orthop Relat Res. 1993;290:17-26.

28.

Shakoor

de Cesar Netto

Thawait

, et al. Weight-bearing radiographs and cone-beam computed tomography examinations in adult acquired flatfoot deformity. Foot Ankle Surg. 2021;27(2):201-206. doi:10.1016/j.fas.2020.04.011

29.

Sohn

Stuck

Pinzur

Lee

Budiman-Mak

Lower-extremity amputation risk after Charcot arthropathy and diabetic foot ulcer. Diabetes Care. 2010;33(1):98-100. doi:10.2337/dc09-1497

30.

Wang

Wluka

Jones

Ding

Cicuttini

FM.

Use magnetic resonance imaging to assess articular cartilage. Ther Adv Musculoskelet Dis. 2012;4(2):77-97. doi:10.1177/1759720X11431005

31.

Siegler

Allard

, et al. ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion—part I: ankle, hip, and spine. J Biomesch. 2002;35(4):543-548. doi:10.1016/S0021-9290(01)00222-6

32.

Wukich

Raspovic

Hobizal

Rosario

Radiographic analysis of diabetic midfoot Charcot neuroarthropathy with and without midfoot ulceration. Foot Ankle Int. 2014;35(11):1108-1115. doi:10.1177/1071100714547218

33.

Wukich

Sung

Charcot arthropathy of the foot and ankle: modern concepts and management review. J Diabetes Complications. 2009;23(6):409-426. doi:10.1016/j.jdiacomp.2008.09.004

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB

2.68 MB