Three-dimensional features of the tarsal sinus and canal complex in normal feet and flexible flatfeet of children

Abstract

Background

Understanding dimensions of both sinus tarsi (ST) and tarsal canal (TC) in normal and flatfeet may influence the design, usage and possible control of postoperative complications of arthroereisis. This study was to assess 3D features of the tarsal sinus and canal complex (TSCC) in both normal control and flatfeet in children under non-weightbearing and weightbearing conditions.

Methods

Three-dimensional (3D) computer-aided design modeling from WBCT scans of 22 children with flexible flatfeet (age 9–14) and 14 with normal control feet (age 9–15) were used to evaluate volume and dimensions of TSCC. Correlations between Meary’s angle, hindfoot alignment, and volume of TSCC were calculated on non-weightbearing and weightbearing CT scans.

Results

The volume of TSCC was smaller in the flatfoot group compared with the control group both under weightbearing and non-weightbearing conditions. From non-weightbearing to weightbearing the volume of TSCC decreased by 19.1% in the flatfoot group and 13.1% in the control group while the long axis of TSCC moved superiorly and anteriorly with internal rotation in both groups. When bearing weight, TSCC volume was negatively linearly correlated with the hindfoot alignment and Meary’s angle.

Conclusions

This study demonstrates that weightbearing has different influences in 3D features of the tarsal sinus and canal complex in children with control and flatfeet. These findings may help understanding biomechanics of pediatric flatfoot and guiding treatment in particular when arthroereisis is used.

Level of evidence

Level III, Retrospective Comparative Study.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12891-026-10147-x.

Introduction

Pediatric flatfoot (pes planus) is described as a dimension or absence of longitudinal medial arch and the deformity may include an abnormality in the hindfoot or ankle. The prevalence of pediatric flatfeet is 4% of 10-year-olds and around 20–24% of the total population has a deficient medial longitudinal arch in at least one foot [1, 2]. Children with no arch development by the age of 10 may have limited potential for natural resolution of flatfeet. The main factor that distinguishes physiological flatfeet from pathological flatfeet is the flexibility of the foot and the presence of pain [3].

Subtalar arthroereisis is typically used for flexible flatfoot, while arthrodesis is more common for rigid deformities or end-stage arthritis in adults. The purpose of subtalar arthroereisis, a simple, minimally invasive procedure, involves the placement of an implant within the tarsal sinus and canal to restrict excessive motion at the subtalar joint [4]. Theoretically, this implant limits the valgus of the subtalar joint and prevents the excessive pronation of the hindfoot, subsequently helping to maintain the longitudinal arch [5]. Pavone et al. reported a 96.83% good-to-excellent level of patient satisfaction treated with arthroereisis [6]. Many studies have shown similar postoperative functional and radiographic results as well [7–10]. However, complications are common, including pain, screw loosening, peroneal muscle contracture, infection and talar fracture [11–13].

An implant of the correct size that matches the TSCC can significantly reduce various postoperative complications. Although multiple implant designs and materials have been proposed, no consensus has been reached regarding the design and choice of implant, because the changes in the TSCC with different conditions remain poorly understood. The conical interosseous cavity between the talus and calcaneus could be divided into two parts: the narrow posteromedial part called the tarsal canal and the widening anterolateral part called the tarsal sinus [14]. It lies beyond the calcaneus with its axis orienting anterolaterally. In literature, the tarsal sinus was a cone-shaped anatomical space bounded by the calcaneus and talus [15]. But its shape is unknown under WB condition.

A reference for dimensions in the TSCC could help in designing the implants and understanding the arthroereisis. Bali et al. measured the dimensions in the narrowest point in the tarsal canal using CT under NWB condition [16]. Guan et al. utilized MRI to measure the anatomical morphology of the sinus tarsi. It found that the sinus tarsi in patients with pes planus is significantly longer, wider, and has different coronal and sagittal axis angles compared to healthy volunteers [17]. Another study employed WBCT to measure morphometric parameters of the sinus tarsi. It discovered that a smaller sinus tarsi angle and a greater ratio of sinus tarsi length to tibia width are significantly correlated with the radiographic severity of flatfoot [18]. Most studies have focused on the anatomy or radiological visualization of the tarsal sinus [19, 20]. Recently, Kleipool et al. conducted a comprehensive 3D assessment of the TSCC dimensions in healthy adults under simulated weightbearing using positional CT, highlighting the influence of foot position on its morphology [21]. However, their study focused on an adult population without foot pathology.

To the authors’ knowledge, no studies have specifically documented of the TSCC in pediatric flatfeet using WBCT. While preliminary 3D data on the tarsal sinus and canal in pediatric flatfoot have been reported in a conference abstract [22], the dynamic morphological changes of the entire TSCC under WB conditions, along with its comprehensive correlation with conventional radiographic parameters, remain underexplored. Therefore, the purpose of this study was to systematically measure and compare the 3D morphological changes of the TSCC in pediatric flatfeet under both WB and NWB conditions, and to further analyze the correlation between these novel 3D parameters and established clinical indices. These fundamental morphological data, particularly the dynamic changes under load, are expected to provide a key reference for future designs of subtalar arthroereisis implants that better conform to the anatomical morphology and biomechanical environment of the TSCC.

Materials and methods

The study was approved by the institutional review board of the hospital (TRECKY2020-146) and was performed in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from the parents or legal guardians of all participants under the age of 16, and assent was obtained from the participants themselves where appropriate. All data were anonymized to protect patient privacy. Twenty-two Flexible flatfeet children (flatfeet group) and 14 non-flatfeet children (control group) were included in this study from June 2018 to June 2021.

Control group children were recruited from those presenting to our clinic during the same period for minor ankle sprains or other non-flatfoot issues. Clinical and radiographic examinations ruled out flatfoot, skeletal pathology, and post-traumatic deformity. After explanation of the study purpose and radiation risks, their guardians voluntarily consented to participate in the WBCT scanning for this study. Both feet from these participants were included only after verifying that the injured foot showed no osseous abnormality on WBCT. Patients were diagnosed with bilateral flexible flatfeet through weightbearing X-ray and physical examination by a specialized foot and ankle surgeon. The patients’ inclusion criteria were 1) age between eight to fifteen years old and 2) confirmed flexible flatfeet. The exclusion criteria were 1) age younger than eight or older than 15 years old 2) rigid flatfeet 3) preexisting other foot pathology, and 4) a history of foot and ankle surgery.

All WBCT scans were obtained from a PedCAT unit (CurveBeam, USA). The acquisition parameters were as follows: tube voltage 100 kV, tube current 10 mA, slice thickness 0.4 mm, pixel spacing 0.4 mm, image matrix 512 × 512, with a field of view (FOV) covering the bilateral foot and ankle region. Data collected included age, gender, height, weight and WBCT scans.

Three-dimensional model reconstruction and coordinate system definition

The DICOM data from both WB and NWB scans were imported into medical image processing software (Mimics, version 20.0, Materialise, Leuven, Belgium) for 3D reconstruction. Semiautomated segmentation was performed using the CT Bone tool with a Hounsfield Unit (HU) threshold range of 529 to 1535 to create a mask of the entire foot bones. A second mask encompassing both bone and the encapsulated soft tissue of the sinus tarsi region was created using a broader threshold range (-425 to 1535 HU). The mask of the TSCC was then generated by a logical subtraction of the bone mask from the second, cropped mask. The resulting TSCC mask was manually refined in 3D view to ensure anatomical accuracy. Finally, the 3D surface model of the TSCC was created, wrapped (smallest detail: 0.5 mm, gap closing: 0.5 mm), and smoothed (three iterations, smooth factor: 0.5) to obtain a model representative of the true morphology.

Considering the inconsistent positioning of the patients’ feet during the WBCT scan, an idealized coordinate system was reconstructed based on the anatomy of the foot to accurately assess the TSCC deviation between WB and NWB conditions. The process consisted of two main steps: model alignment and coordinate system definition. A foot axis was first defined on the WB model as the line connecting the most plantar point of the calcaneus and the center of the second metatarsal head. This foot axis of the WB model was aligned with the anteroposterior axis of the global coordinate system. Subsequently, the NWB model was rigidly registered to the aligned WB model by performing an iterative closest point (ICP) algorithm, with the objective of minimizing the distance between the surfaces of the calcaneus and talus. This registration ensured that the WB and NWB models were placed in the same spatial orientation, allowing for direct comparison of the TSCC independent of initial scanning posture.

Following alignment, a subject-specific anatomic coordinate system was defined based on the calcaneus of the WB model, with its origin at the most plantar point of the calcaneus. The axes were defined as follows: 1) Vertical Axis (Z): A line perpendicular to the floor (simulating the direction of gravity during weightbearing). 2) Anteroposterior Axis (Y): A line perpendicular to the Z-axis and parallel to the line of foot axis. This defined the sagittal plane (Y-Z plane). 3) Mediolateral Axis (X): A line orthogonal to both the Y and Z axes, defined by the cross product of the Z and Y axes (X = Z × Y). This defined the coronal plane (X-Z plane) and the axial (horizontal) plane (X-Y plane).

All subsequent linear and angular measurements of the TSCC were performed within this standardized anatomic coordinate system. A schematic diagram illustrating the coordinate system and alignment process is provided in the supplementary material (Supplementary Fig. 1). All reconstruction, alignment, and measurement steps were performed by the same experienced operator following a consistent protocol.

Measurements of TSCC morphological parameters

Following the establishment of the standardized coordinate system, a series of 3D morphological parameters of the TSCC were measured. These parameters were designed to comprehensively describe the size, shape, and spatial orientation of the TSCC. For clarity, they are categorized and defined as follows (illustrated in Fig. 1).

Fig. 1.

Schematic diagram of measurements of various parameters in the TSCC. A: The measurement diagram of the WDPTC, WDATS, NDPATC, NSDTC, LTSC and the LTC. B: The measurement diagram of the angle between TSCC and coronal plane. Superior (axial) view of the calcaneus. C: The measurement diagram of the angle between TSCC and the horizontal plane. Posterior view of the calcaneus. Abbreviations: A = anterior, P = posterior, M = medial, L = lateral. S = Superior, P = Posterior. TSCC: Tarsal sinus and canal complex; WDPTC: widest diameter of the posteromedial tarsal canal; WDATS: widest diameter of the anterolateral tarsal sinus; NDPATC: narrowest diameter perpendicular to the axis of the tarsal canal; NSDTC: narrowest superoinferior diameter of the tarsal canal; LTSC: length of the tarsal sinus and canal; LTC: length of the tarsal canal; HATS: the angle between TSCC and the horizontal plane; CATS: the angle between TSCC and coronal plane

Volumetric parameters include tarsal sinus and canal complex volume (TSCV). It is the total enclosed volume of the 3D reconstructed TSCC model and calculated automatically by the 3-matic software (v. 12.0, Materialise) as a property of the surface model. Linear dimensional parameters include the followings and were described in prior anatomical studies [14, 16–18] and were quantified in 3D in this study. Widest diameter of the posteromedial tarsal canal (WDPTC) was the maximum distance between the most lateral and most medial points at the wide, posteromedial opening of the tarsal canal. Widest diameter of the anterolateral tarsal sinus (WDATS) was the maximum distance between the most lateral and most medial points at the wide, anterolateral opening of the tarsal sinus. Narrowest diameter perpendicular to the axis of the tarsal canal (NDPATC) means the minimal cross-sectional diameter measured perpendicular to the long axis of the tarsal canal. This typically represents the constricted mid-portion of the canal. Narrowest superoinferior diameter of the tarsal canal (NSDTC) was the minimal vertical (superior-to-inferior) distance within the tarsal canal. Length of the tarsal sinus and canal (LTSC) was measured from the most proximal (posteromedial) point to the most distal (anterolateral) point along the central axis of the entire TSCC. And the length of the tarsal canal (LTC) was measured from the proximal opening to the distal end (transition to the sinus) along the axis of the tarsal canal portion.

Angular parameters include the angle between TSCC and the horizontal plane (HATS) and the angle between TSCC and coronal plane (CATS). These are novel parameters defined in this study to quantify the spatial orientation of the TSCC within the established anatomic coordinate system. The former was the projection on the coronal plane of the angle between TSCC and the horizontal plane. It represents the orientation of the TSCC from superior to inferior. CATS was the projection on the horizontal plane of the angle between TSCC and coronal plane which represents the orientation of the TSCC from posteromedial to anterolateral. All of the measurements were performed for both WB and NWB conditions for comparison.

Furthermore, Simulated X-ray images were extracted from the WB and NWB CT images using the Insta X function in the software CubeVue. These X-ray images were used for geometrical measurements. The measurements include Meary’s angle, Calcaneal pitch, Talonavicular coverage angle, and hindfoot alignment angle (HAA).

Statistical analysis

All statistical analysis was performed by SPSS (version 24.0, IBM Corp). The data were presented as mean ± standard deviation. Comparison between the WB and NWB model measurements was performed using the paired samples Student’s t-test. Between-group differences under specific loading conditions were assessed using independent samples t-tests. The intraclass correlation coefficient (ICC) was performed for interobserver reliability of measurement parameters among three trained surgeons. Categorical variables were presented as frequency or percentage and the comparison was calculated using the Chi-square test or Fisher’s exact test. For correlation analyses, Pearson correlation coefficients (r) were derived from linear mixed-effects models. P < 0.05 was regarded as statistically significant.

As the study included bilateral foot data from some subjects, to account for potential within-subject correlation and to rigorously test whether the effect of weightbearing differed significantly between the flatfoot and control groups, linear mixed-effects models were employed, including an interaction term (Group × Weightbearing) with Subject ID as a random effect. Model fit was evaluated using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC).

Results

A total of 36 subjects’ WBCT scans were obtained from the foot and ankle center. Of the 36 subjects, 22 were flexible flatfeet children (16 male and 6 females, mean age, 11.2 years, range, 9–14) and 14 were in control group (3 male and 11 females, mean age, 12.4 years, range, 9–15) (Table 1).

Table 1.

Demographic clinical data

	Flatfeet group (n = 22)	Control group (n = 14)	P-value
Age, y	11.2 ± 1.3 (9–14)	12.4 ± 1.8 (9–15)	0.026*
Sex (male/female)	16/6	3/11	0.003*
BMI(Kg/m2)	18.7 ± 2.8(14.0–26.0)	18.9 ± 3.0(13.7–23.0)	0.843

* Statistically significant: P < 0.05

In this study, a CAD model of the TSCC was successfully established and the white photosensitive resin model using 3D printing technology was printed (Fig. 2A, B, I and J). The model showed that the tarsal canal was semi-cylindrical, the floor of the tarsal canal was shallow and semi-cylindrical and the tarsal sinus was semi-funnel-shaped. Furthermore, the split of the canal between the talus and calcaneum was asymmetric. The superior two-thirds of the canal was formed by an anatomic arch within the neck of the talus. The narrowest part of the tarsal canal was formed by the undersurface of the neck of the talus. The shape and volume of the tarsal sinus were mainly affected by a change from NWB to a WB position, with little effect on the tarsal canal, especially the narrowest part (Fig. 2C-H). This site actually appeared to be a relatively stable and immobile surface for the interference fit of the implant.

Fig. 2.

Graphical representation of a 3D CAD foot model (right) based on a 12-year-old man with flexible flatfeet. A: The 3D CAD intact foot model. B: Partial enlarged view of tarsal sinus. C-E: 3D CAD model of the TSCC from different perspectives in a NWB condition. F-H: 3D CAD model of the TSCC from different perspectives in a WB condition. I-J: 3D printer model of the TSCC in top and bottom view of the NWB and WB conditions, respectively. Abbreviations: CAD: Computer-aided design; NWB: nonweightbearing; WB: weightbearing; TSCC: tarsal sinus and canal complex

The 3D measurements of the established CAD model were compared under WB and NWB conditions. The measurements showed excellent concordance in the model. The ICC was 0.952 for the 3D measurements of WDPTC and 0.912 for that of the WDATS. Both ICCs were significant at the P < 0.001. In the flatfeet group, the TSCV decreased significantly by nearly 19.1% from 3554.4 ± 694.1 to 2875.6 ± 588.4 mm³ after WB (P < 0.001) (Table 2). Interestingly, there was no significant difference in the WDPTC (P = 0.390), NDPATC (P = 0.982) and LTC (P = 0.568) in the two conditions. Furthermore, the WDATS (P < 0.001), NSDTC (P = 0.006) and the LTSC (P = 0.001) decreased with significant difference as compared with NWB condition. In addition, the HATS decreased statistically significantly from 13.7 ± 4.6 degrees to 12.4 ± 5.4 degrees after WB (P = 0.017). Inversely, the CATS increased by WB significantly from 36.3 ± 6.3 degrees to 41.3 ± 6.8 degrees (P < 0.001). These results demonstrated that the WB condition could mainly affect the morphology and volume of the tarsal sinus rather than that of the tarsal canal (Fig. 3A-C). Compared with control group, the TSCV, WDATS and LTC decreased with statistical difference in flatfeet group both under NWB and WB condition, while the CATS increased (Fig. 4).

Table 2.

Comparison of 3D measurements of the TSCC between NWB and WB conditions in the two groups

	Flatfeet group (44 feet)		P-value	Control group (28 feet)		P-value
	NWB	WB		NWB	WB
TSCV, mm3	3554.4 ± 694.1	2875.6 ± 588.4	< 0.001*	4003.4 ± 731.4	3478.7 ± 694.2	< 0.001*
WDPTC, mm	10.3 ± 1.9	10.1 ± 1.8	0.390	10.1 ± 1.7	10.7 ± 1.7	0.024*
WDATS, mm	22.2 ± 3.9	19.9 ± 3.9	< 0.001*	27.0 ± 2.2	24.8 ± 2.4	< 0.001*
NDPATC, mm	8.2 ± 1.3	8.2 ± 1.3	0.982	8.3 ± 1.0	8.8 ± 1.1	< 0.001*
NSDTC, mm	8.0 ± 1.2	7.7 ± 1.2	0.006*	8.6 ± 1.2	8.3 ± 1.4	0.062
LTSC, mm	33.1 ± 3.1	34.0 ± 3.2	0.001*	32.8 ± 3.8	33.9 ± 3.9	0.001*
LTC, mm	15.4 ± 1.8	15.2 ± 2.0	0.568	18.5 ± 2.1	18.4 ± 2.9	0.834
HATS, deg	13.7 ± 4.6	12.4 ± 5.4	0.017*	14.6 ± 4.3	11.0 ± 3.9	< 0.001*
CATS, deg	36.3 ± 6.3	41.3 ± 6.8	< 0.001*	30.9 ± 7.0	36.8 ± 6.40	< 0.001*

*Statistically significant: P < 0.05

Fig. 3.

Comparison of tarsal sinus and canal volume with postural changes. A and B: the 3D CAD model in WB and NWB posture, respectively. C: 3D printing model of the TSCC in WB and NWB positions, respectively. Abbreviations: CAD: Computer-aided design; WB: weightbearing; NWB: nonweightbearing

Fig. 4.

Comparison of 3D measurement parameters in the control group and the flatfeet group

To formally test whether the morphological response to weightbearing differed between pediatric flatfoot and control groups, linear mixed‑effects models were constructed with Group and Weightbearing as fixed effects, their interaction (Group × Weightbearing), and Subject ID as a random effect. As summarized in Supplementary Table 1, a significant interaction effect was observed only for HATS (β = −2.265, P = 0.027), indicating that weightbearing induces distinct changes in the horizontal orientation of the tarsal sinus in flatfeet compared to controls. No significant interaction effects were detected for TSCV (P = 0.171) or other morphometric parameters.

Meary’s angle, Calcaneal pitch, Talonavicular coverage angle and HAA in both groups changed significantly after WB. Furthermore, the parameters measured except for the Calcaneal pitch, were larger in flatfeet group than in control group both in NWB and WB conditions as was shown in Table 3.

Table 3.

Comparison of Meary’s angle, Calcaneal pitch, Talonavicular coverage angle and HAA between NWB and WB conditions in the two groups

	Flatfeet group (44 feet)		P-value	Control group (28 feet)		P-value
	NWB	WB		NWB	WB
Meary’s angle, deg	8.0 ± 5.2	14.0 ± 5.4	< 0.001*	-0.9 ± 6.4	4.6 ± 6.1	< 0.001*
Calcaneal pitch, deg	17.9 ± 4.5	16.1 ± 4.4	< 0.001*	22.9 ± 5.5	21.2 ± 5.0	< 0.001*
Talonavicular coverage angle, deg	10.1 ± 7.3	19.3 ± 8.8	< 0.001*	3.7 ± 1.3	7.8 ± 5.5	0.001
HAA, deg	1.0 ± 3.9	7.7 ± 3.3	< 0.001*	-1.0 ± 4.1	2.0 ± 2.1	< 0.001*

*Statistically significant: P < 0.05

The correlation between HAA and tarsal sinus parameters was further investigated. In flatfeet group, negative linear correlations were found between HAA and TSCV (r = -0.4756, P < 0.0001) and WDATS(r = -0.2389, P = 0.0250) (Fig. 5A, C). A positive linear correlation was found between the HAA and the CATS (r = 0.5832, P < 0.0001). There was no correlation found between the HAA and the WDPTC (r = -0.1566, P = 0.1451), NDPATC (r = 0.0554, P = 0.6081) and the NSDTC (r = 0.0249, P = 0.8183). This indicated that little had altered in the tarsal canal when the hindfoot changed from varus to valgus. An almost similar performance could be seen in control group (Supplementary Fig. 2).

Fig. 5.

Correlation between HAA and 3D measurement parameters in flatfeet group under WB conditions. A-I: Linear regression analysis of HAA and TSCV, WDPTC, WDATS, NDPATC, NSDTC, LTSC, LTC, HATS, CATS. Correlations were derived from linear mixed-effects models that accounted for bilateral measurements within subjects. The Pearson r values presented represent marginal correlations estimated from the mixed-effects framework

In flatfeet group, Meary’s angle was negatively correlated with the TSCV (r= -0.4115, P < 0.0001) and the WDATS (r = -0.2688, P = 0.0120), while a significantly positive correlation was found between Meary’s angle and the CATS (r = 0.3663, P = 0.0004) (Fig. 6). In control group, a negative correlation was also found between the Meary’s angle and the TSCV (r = -0.2906, P = 0.0298) (Supplementary Fig. 3A).

Fig. 6.

Correlation between Meary’s angle and 3D measurement parameters in flatfeet group under WB conditions. A-I: Linear regression analysis of Meary’s angle and TSCV, WDPTC, WDATS, NDPATC, NSDTC, LTSC, LTC, HATS, CATS. Correlations were derived from linear mixed-effects models that accounted for bilateral measurements within subjects. The Pearson r values presented represent marginal correlations estimated from the mixed-effects framework

Discussion

In this study, a 3D CAD model of the TSCC in pediatric flatfeet and non-flatfeet was established and measured using WBCT scans. The morphological change of the TSCC was also described and compared between NWB and WB conditions. We found that the TSCC was something like a combination of a semi-cylinder and semicircular cone. A significant decrease of the TSCV was found in both WB condition and pediatric flatfeet. Besides, the TSCC became more parallel to the horizontal plane and the foot axis after WB. Significantly negative correlation was found between HAA and TSCV. Moreover, TSCV decreased with a greater Meary’s angle.

The CAD model of the TSCC in this study showed that the TSCC was a lateral continuation of the medial tarsal canal and formed by the talar and calcaneal sulci; the stem-like tarsal canal extended medially, whereas the larger tarsal sinus opened laterally. The sinus on the lateral side was conical and was bounded by the lateral process and neck of the talus and by the anterosuperior aspect of the calcaneus. Furthermore, it continued in the posteromedial direction as a funnel-shaped canal that opened medially posterior to the sustentaculum tali, as found in cadaver studies [20, 23, 24]. In addition, the tarsal canal was not a regular columnar shape. The upper surface of the middle of the tarsal canal was obviously protruding, while its bottom was relatively flat. These morphological characteristics have not been considered so that most implants were designed as a cylindrical soft-tissue screw, constructed of titanium and threaded. Therefore, we hope that these designs could be improved with better confirmation to the TSCC by taking into account these morphological characteristics and dynamic changes in the TSCC.

The purpose of subtalar arthroereisis is to restrict abnormal excessive pronation and restore normal alignment of the hindfoot. However, some patients showed no significant functional improvement, resulting in a high rate of sinus tarsi pain and the primary reason may be the mismatch of the implants [25]. The design of the implant is very difficult because of the lack of 3D measurements of the tarsal sinus of pediatric flatfeet. For example, the Maxwell-Brancheau Arthroereisis (MBA™) implant prevents the talus from touching the bottom of tarsi sinus via siting beneath the neck of talus, while the Sgarlato implant and the STA-peg implant are fixed on the calcaneus and alter the tarsi sinus axis [26]. The implant of HyProCure was used in an extra-osseous talotarsal stabilization via reinforcing the talotarsal joint axis [27]. The shape of the implant in tarsal canal is cylinder while our study found the canal was semi-cylinder.

The size of most implants is six to 12 mm and the most frequently used size is eight mm [13]. In our study, the NDPATC was about 8.2 mm and the NSDTC was about 8.0 mm in flatfeet group during NWB. Bali et al. reported that the narrowest mean AP diameter and superoinferior diameter of the canal were about 7.3 mm and 9.2 mm in sagittal computed tomography, respectively [16]. In a more recent study, Kleipool et al. found that the diameter did not change much in the middle of the tarsal canal in different foot WB conditions [21]. Our results were consistent favorably with these studies. Guan et al. reported a longer and wider sinus tarsi in flatfoot patients on non-weightbearing MRI [17]. Our study extends this by showing that, despite a potentially larger inherent volume in the unloaded state, this space is highly compressible and undergoes significant reduction under physiological load. Chen et al. found on WBCT that a smaller 2D sinus tarsi angle correlated with more severe flatfoot [18]. Our HATS and CATS parameters extend this by providing a 3D vector analysis of the entire canal’s axis. The correlation of HATS with talonavicular coverage angle reinforces the link between TSCC morphology and deformity severity. Taken together, the current study showed that the TSCV in pediatric flatfeet decreased about 19.1% from NWB to WB and decreased about 13.1% compared to the control group. However, there was no significant change between the WDPTC and NDPATC. These measurement results are not only helpful in designing implants but also could guide surgeons in choosing the size-matched implants.

For insertion of the implant, a guide pin needs to be inserted from lateral to medial across the tarsal sinus perpendicular to the longitudinal axis of calcaneus. Then the implant should be placed onto the guide pin and advanced from lateral to medial until the tip abuts the medial margin of the tarsal sinus [28]. The operative procedure is mainly based on the experience of the surgeon. Our results indicated that in pediatric flatfeet, the CATS was larger. It meant the TSCC was much more parallel to the foot axis. The HATS varied from 8.46 degrees to 18.14 degrees and the CATS was from 28.73 degrees to 42.15 degrees in flatfeet group during NWB condition. Furthermore, the HATS decreased by 9% while the CATS increased by 12% in WB condition. Theoretically, the talus in pediatric flatfeet becomes more plantarflexed so that the HATS should be larger in WB condition. Conversely, it decreased in our study. We inferred that the distal calcaneus could be more dorsiflexed. The larger CATS implied a larger inversion angle of the tarsal canal, indicating a subtalar joint eversion angle in WB condition. Additionally, the horizontal and coronal angles of the TSCC are of great significance in assisting the surgeons to place the guide pin and implants. Importantly, mixed‑effects modeling revealed a significant Group × Weightbearing interaction for HATS (P = 0.027), suggesting that the adaptive reorientation of the tarsal sinus in the horizontal plane is a distinctive feature of pediatric flatfoot under load. In contrast, TSCV did not show a significant interaction effect (P = 0.171), indicating that although flatfeet exhibit greater absolute volume reduction, this difference in loading response is not statistically distinguishable from controls after accounting for within‑subject correlation.

In this study, a significantly negative linear correlation was found between the HAA and the TSCV, WDATS. It could be inferred that the volume of the TSCC decreased when the hindfoot alignment changed to valgus. And the positive linear correlation between the HAA and the CATS indicated that the TSCC adducted when hindfoot alignment became more valgus. On the other hand, increased Meary’s angle and Talonavicular coverage angle, as well as the decreased Calcaneal Pitch, were found in pediatric flatfeet. Not surprisingly, the Meary’s angle was negatively correlated to the TSCV, indicating that the available TSCC space decreased with increasing sagittal-plane flatfoot severity. This may be explained by the plantarflexion of the talus. No correlation was found between Meary’s angle and the HAA. It could be infered that the TSCV may assist surgeons in diagnosing pediatric flatfeet.

The findings of this study extend beyond the initial dimensional descriptions of the TSCC [22] and carry direct implications for understanding flatfoot biomechanics and refining treatment. Compared to our team’s prior preliminary 3D descriptive study of the TSCC in pediatric flatfoot, the key advancements of the present study are as follows: Firstly, it has systematically quantified the dynamic impact of weightbearing on the 3D morphology of the TSCC, revealing significant changes in its volume and orientation under load. Secondly, it introduces angular parameters such as HATS and CATS to enrich the morphological description of the TSCC from a spatial orientation perspective. Lastly, it establishes, for the first time, linear correlations between TSCC volume and key clinical radiographic parameters including Hindfoot Alignment Angle and Meary’s angle, providing deeper insights into the biomechanical underpinnings of flatfoot deformity and the foundational data for implant design. In all, this study not only provides novel 3D parameters of the TSCC but also, through correlation analysis, links these parameters to the 2D radiographic indices that are widely used in clinical practice to assess flatfoot severity and surgical outcomes. This lays the groundwork for the potential future use of TSCC parameters as an auxiliary, 3D assessment metric, which could be combined with existing parameters to provide a more comprehensive preoperative evaluation.

The clinical relevance of the above 3D morphological parameters of the TSCC lies in their direct applicability to preoperative planning and implant selection for subtalar arthroereisis. First, the volume and narrowest diameters (NDPATC and NSDTC) reported in this study provide a quantitative reference for selecting implant size, potentially reducing complications related to under- or oversized implants. Second, the angular parameters (HATS and CATS) describe the spatial orientation of the TSCC under weightbearing conditions, which can guide the trajectory of guide pin placement during surgery—an aspect currently dependent largely on surgeon experience. Third, the observed negative correlation between TSCC volume and both HAA and Meary’s angle suggests that preoperative measurement of TSCC volume on WBCT may serve as an auxiliary parameter to assess flatfoot severity and predict the degree of space available for implant placement. Collectively, these data provide an anatomical and biomechanical foundation for the future development of patient-specific or size-matched TSCC implants that better conform to the dynamic morphology of the TSCC.

In our preliminary institutional experience after completion of the present morphometric study, these WBCT-derived parameters have been used to assist preoperative planning for subtalar arthroereisis in a small number of pediatric patients. In particular, NSDTC was used as a reference for implant diameter selection, whereas CATS was used to help determine the guide-pin trajectory. Although these cases were not included in the present study and formal outcome analysis is ongoing, no sinus tarsi pain or implant-related complications have been observed to date in approximately 20 children treated with this 3D-parameter-assisted planning strategy. By contrast, among 22 earlier cases treated without 3D planning, 5 children developed postoperative sinus tarsi pain and required implant removal. Retrospective application of the current measurement protocol suggested that these implants were oversized relative to the patient-specific NSDTC, resulting in medial cortical impingement. These preliminary observations should be interpreted cautiously, but they support the potential clinical relevance of patient-specific TSCC measurements for implant sizing and trajectory planning. A dedicated prospective clinical study with larger sample size and longer follow-up is underway to validate whether these parameters can reduce implant-related complications and improve clinical outcomes.

Despite the positive 3D measurements and correlation results, we recognized that there existed limitations in this study. This was a single-center retrospective study with a relatively small sample size. More cases could be examined in a multicenter investigation and further prospective controlled clinical studies on measurements of the TSCC are needed. The simulated radiographs for Meary’s angle were extracted from WBCT scans. The quality and resolution were lower compared to conventional radiographs. This may be another limitation but it gave much reliability to the correlation analysis on the Meary’s angle and the 3D measurements since both of them were obtained from the same acquisition. Additionally, participants with unilateral soft-tissue ankle injuries were included in the control group. Although the absence of bony abnormality was confirmed, future studies with truly asymptomatic healthy volunteers would further strengthen the validity of the control data.

Conclusion

This study demonstrates that weightbearing has different influences in 3D features of the tarsal sinus and canal complex in children with control and flatfeet. The tarsal sinus and canal volume is negatively correlated with both the HAA and Meary’s angle. These findings may be very relevant in future study of the subtalar joint in both children and adults, and clearly have an implication for treatment of the pediatric flatfoot with arthroereisis.

Authors’ contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Chengyi Sun, Shenglong Ding and Yingqi Zhang. The first draft of the manuscript was written by Chengyi Sun. Linfeng Ji participated in its design and helped to draft the manuscript. Jianzhong Zhang participated in the data collection and data analysis. Mingzhu Zhang participated in the design and coordination of the study and revised the manuscript. All authors read and approved the final manuscript.

Funding

This research was supported by Beijing Natural Science Foundation (L2609028), the Sailing Plan for Clinical Medicine Development (YGLX202505) and Beijing High-level Public Health Technical Talents Construction Project (Academic Leader-02-23).

Data availability

All data generated or analysed during this study are included in this published article and its supplementary information files.

Declarations

Ethics approval and consent to participate

Informed consent to participate was obtained from the parents or legal guardians of any participant under the age of 16. This retrospective study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Institutional Review Board of Beijing Tongren hospital (TRECKY2020-146).

Consent for publication

Written informed consent for publication of their clinical details and/or clinical images was obtained from the parents or legal guardians. A copy of the consent form is available for review by the Editor of this journal. All of the data were made anonymous to protect patients’ privacy.

Competing interests

The authors declare no competing interests.