and Correlation Analysis of Computed Methods to Convert Conventional 2D Radiological Hindfoot Measurements to 3D Equivalents Using Weight Bearing CT

, Francois Lintz2, Cesar de Cesar Netto3, Alexej Barg4, Arne Burssens5 and Scott Ellis6



(1)
Department for Foot and Ankle Surgery, Hospital Rummelsberg, Schwarzenbruck, Germany

(2)
Foot and Ankle Surgery Centre, Clinique de l’Union, Toulouse, France

(3)
Department of Orthopedics and Rehab, University of Iowa, Iowa City, IA, USA

(4)
University Orthopedic Center, University of Utah, Salt Lake City, UT, USA

(5)
Department of Orthopedics and Trauma, University Hospital of Ghent, Ghent, OVL, Belgium

(6)
Department of Orthopedic Surgery, Hospital for Special Surgery, New York, NY, USA

 



Keywords

Hindfoot alignmentWeight bearing CTComputed radiologyHindfoot correction


Introduction


Exact radiographic assessment of hindfoot alignment remains a challenge [1, 2]. The various measurement techniques and hindfoot views (either inclined anteroposterior (AP) or posterioranterior (PA)) reflect the lack of a standardized and accurate methodology [3]. All current methodologies try to overcome two main inaccuracies: the superposition caused by the osseous structures in the midfoot and the rotational errors created during the positioning of the foot, as demonstrated by several recent studies [46]. Weight bearing CT (WBCT) of the foot and ankle has been shown to be more accurate in hindfoot measurements [7]. This recent imaging technique offers the advantage of a standing position as with weight bearing radiographs but overcomes the disadvantages of the osseous superposition caused by the complex anatomy of the foot and ankle [810]. This allows for complete visualization of the hindfoot [11]. Additionally, WBCT software settings can rotate the foot and ankle after the imaging process to acquire a standardized positioning of the hindfoot [7, 8].


Although computed tomography was introduced to orthopedic surgery in the mid-1970s [12], its routine clinical and 3D use only started in the early 1990s with the introduction of the spiral CT, which allowed better insight into complex fracture patterns [13]. Further applications were lacking, which made some authors question the added value of a 3D CT [14]. Reluctance to adopt 3D CTs was evident in foot and ankle literature where most available measurements and reference angels were still performed in 2D [8, 9, 15]. Nevertheless, the orthopedic field’s interest in 3D printing and computer-assisted surgery (CAS) has grown in recent years [16, 17]. These tools allow for more precise preoperative planning and intraoperative surgical procedures [18]. However, in order to successfully apply them, a better understanding of 3D technology is required. Although the application of these techniques on the skeletal system is generally well understood, their potential use on and subsequent insights from the hindfoot remain unclear; most weight-bearing research of the lower limb has been focused on hip and knee joints [1921].


The advantage of these methods is that they incorporate each plane according to the region of interest with a high measurement accuracy [19].


Using WBCT, the previously described hindfoot measurements allow for correct foot positioning in the coronal, sagittal, and axial plane, but the actual angles are only obtained from one CT slice in one plane [7, 10, 22, 23].


Although interobserver reliability is high, important spatial data is not used, and the manually drawn angles and foot positioning steps impose additional measurement errors [8].


The aim of this paper is therefore to use computed methods to convert these conventional 2D measurements to a 3D environment. This analytic process will be assessed by rater reliability and regression analysis.


Materials and Methods


Study Population, Design, and Measurement Protocol


Forty-eight patients with clinical and radiological absence of hindfoot pathology were included [24]. The mean age was 39.6 years (SD = 3.2, age range: 19–72 years). The indications for imaging using WBCT were one of the following: minor foot and ankle trauma (e.g., foot and ankle sprain or contusion) with persistent complaints that were negative or nonsignificant for an occult fracture (n = 31), the suspicion of osteoarthritis that was undetectable on CT slices (n = 11), or a MTP I fusion to assess consolidation (n = 4) as shown in Table 10.1.


Table 10.1

Patient characteristics


























Characteristic


Total (N = 48)


Age (±) SD


39.6 ± 13.2 years


Sex (M/F)


28/20


Minor trauma


31


Absence osteoarthritis


11


MTP I fusion


4


The contralateral unaffected foot was used for each analysis. The measurements were performed on the images retrieved from the weight bearing pedCAT® cone beam CT, using the incorporated Cubevue® software for the 2D analysis (CurveBeam, Warrington, PA, USA). The 3D analysis was obtained after segmentation of the images using Mimics® 19.0 and analysis using 3-matic® software (Materialise, Leuven, Belgium). The patient records were anonymized and deidentified prior to processing in accordance with the standard data release procedures of the hospital involved in the study. 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. The Institutional Review Board of AZ Monica approved this study (OG10601102015), and formal consent was not required for this type of study. The following imaging protocol was used: radiation source was set at 4 mAs and 50 kV, with a focus distance of 100 cm, and the beam pointed at the ankle joint. PedCAT used the following settings: tube voltage, 96 kV; tube current, 7.5 mAs; CTDIvol 4.3 mGy; matrix, 160,160,130; pixel size, 0.4 mm; and slice interval, 0.4 mm.


At the department of radiology, patients were asked to stand naturally with both feet parallel to each other, shoulder width apart. Hindfoot measurements were performed in 2D by authors AB and MP. Each measurement was repeated three times. After the set of measurements was complete, the average of these three measurements was used for further analysis. A similar test/retest methodology was performed in other studies concerning hindfoot measurements [3, 5, 8, 2527]. The hindfoot angle was determined based on the inferior point of the calcaneus (HA2D), as described previously [8]. In brief, the foot was first positioned in line with the collinear axis of the shaft of the second metatarsal, which is considered as the longitudinal axis of the foot in the axial plane (Fig. 10.1). The hindfoot angle was defined as the intersection between the anatomical tibia axis (TA2D) and the talocalcaneal axis (TC2D), which connects the inferior point of the calcaneus and the middle of the upper surface of the talus in the coronal plane (Fig. 10.1c, d).

../images/484112_1_En_10_Chapter/484112_1_En_10_Fig1_HTML.png

Fig. 10.1

Measuring hindfoot alignment in 2D. (a, b) Positioning of the foot in line with the axis of the second metatarsal in the axial plane. (c, d) The hindfoot angle (HA2D) is composed out of the intersection between the anatomical tibia axis (TA2D, blue line) and the talocalcaneal axis (TC2D, orange line). The TC2D connects the inferior point of the calcaneus with the middle of the talar dome


The varus and valgus alignment of the hindfoot was respectively defined as when the TCA runs medial from the vertical axis and when the TCA runs lateral from the vertical axis, which is often considered as a reference axis [28, 29]. Authors RM and TL determined the 3D hindfoot angle (HA3D) by the use of computer-aided design (CAD) operations (Fig. 10.2a–d).

../images/484112_1_En_10_Chapter/484112_1_En_10_Fig2_HTML.png

Fig. 10.2

Measuring hindfoot alignment in 3D. (a) The anatomical tibia axis (TA3D) was computer calculated as an axis based on the moment of inertia (depicted in the upper right quadrant) through the distal end of the tibia marked above the fibular groove. (b) The inferior calcaneus point was calculated by an extrema analysis (a software function to determine the most outer point in the superior-inferior direction) (arrow). (c) The center of the talus was calculated as a centroid (depicted in the upper right quadrant) based on the mean position of all points in the talus. The talocalcaneal axis (TC3D) was calculated by connecting the inferior calcaneus point with the centroid of the talus. (d) The intersection of both axes became the HA3D


The anatomical tibia axis (TA3D) was calculated by a best fit centroidal axis along the diaphysis marked above the incisura fibularis (Fig. 10.2a). The talocalcaneal axis (TC3D) was computed by connecting the inferior calcaneus point (ICP) with central talus point (CTP). The ICP was obtained after the calculation of an extrema analysis of the calcaneus (function to determine the most outer point of a structure in the direction of a given axis) (Fig. 10.2b). The CTP was determined by the calculated centroid of the talus (mean position of all the points in a given structure) (Fig. 10.2c). The computed intersection of both the TA3D and the TC3D became the HA3D (Fig. 10.2d). The TA and the TC were measured separately in the hindfoot angle when comparing the 2D and the 3D angles in order to emphasize possible inconsistencies attributable to either the tibial or talocalcaneal component.


The talocrural angle (TCr) was used as a radiographic parameter to assess the ankle in the coronal plane [30]. TCr was measured in 2D (TCr2D) as the angle between the intersection of the intermalleolar axis (obtained after connecting the interior point of the medial with the most inferior point of the lateral malleolus) and the horizontal axis of tibial joint line (Fig 10.3a).

../images/484112_1_En_10_Chapter/484112_1_En_10_Fig3_HTML.png

Fig. 10.3

Common ankle and hindfoot measurements. (a, b) The talocrural angle (TCr) was measured in 2D (TCr2D) by the intersection of the malleolar axis and the tibial joint line. The 3D (TCr3D) was measured as the intersection between the malleolar axis, created by connecting the inferior medial and lateral malleolus through an extremity analysis and the tibial joint line. (c) Characteristics in the tibiotalar joint were measured as the tibial inclination (TI2D, upper line) the talar tilt (TT2D, lower line). (d) Representation of the TI3D, TT3D. (e) Characteristics in the hindfoot were measured as the SVA (SVA 2D). (f) Representation of the SVA3D


In 3D (TCr3D), this measurement is performed in the same manner by the intersecting angle of the intermalleolar axis (the most inferior points of the malleoli were computed using an extrema analysis) and the computed best fitted axis through the horizontal contour of the tibial joint line (Fig. 10.3b).


Characteristics in the tibiotalar joint were measured as the inclination of the tibial joint surface towards the vertical axis perpendicular to the floor (TI2D) and the tilt of the talus towards the vertical axis perpendicular to the floor (TT2D) as described previously (Fig. 10.3c) [8].


The TI3D and TT3D were similarly analyzed by reconstructing the joint surface respective of the tibia and the talus in the coronal plane. This reconstruction allows for the computation of the horizontal axis of both surfaces and the intersection with the vertical axis perpendicular to the floor resulted in the TI3D and TT3D.


In the subtalar joint (STJ), the middle subtalar vertical angle (SVA2D) was determined in the coronal plane according to the method described by Colin et al. [9]. This measurement required the length of the posterior facet of the STJ to be measured in the sagittal plane. In the midpoint of this distance, the inclination of the STJ surface towards the vertical axis perpendicular to the floor in the coronal plane was determined (Fig. 10.3e). The SVA3D was analyzed similarly to the SVA2D with the same methods as applied in the TI3D and TT3D, which is generalized in Fig. 10.3f and detailed in Fig. 10.4.

../images/484112_1_En_10_Chapter/484112_1_En_10_Fig4_HTML.png

Fig. 10.4

Sequence of translating commonly used 2D measurements to 3D angles. (a) Starting as an example with an AP radiograph of the talocrural angle, which was measured as the intersection between the axis connecting both malleoli and the axis parallel to articular surface of the distal tibia in 2D. (b) Same measurement in 2D is applied by the use of weight bearing CT after correct rotation. (c) Computer-calculated points (blue) to determine the axes and 3D angle. (d) Schematic representation of projecting a 3D angle in the coronal (yz-plane) through the used software by applying build-in goniometric functions. (e–f) Cartesian coordinate system with the origin defined in the inferior point of the calcaneus. The z-axis was calculated perpendicular to the floor through the origin. The x-axis runs through the origin perpendicular to the z-axis and lies in the sagittal plane, formed through the center of the second metatarsal head and the origin perpendicular to the ground floor. The y-axis goes through the origin, perpendicular to the x-axis and z-axis


By applying goniometric functions built into the software, commonly used measurements from a 2D radiograph can be translated into a 3D angle and its subsequent projection. The general sequence is depicted and explained for the talocrural angle as an example (Fig. 10.4a–d).


The coronal plane in these methods was derived from the Cartesian coordinate system with the inferior calcaneus point as the origin. The z-axis was defined as running through the origin perpendicular to the ground floor. The x-axis runs through the origin perpendicular to the z-axis and lies in the sagittal plane, formed through the center of the second metatarsal head and the origin perpendicular to the ground floor. The y-axis goes through the origin, perpendicular to the x-axis and z-axis (Fig. 10.5ef).

../images/484112_1_En_10_Chapter/484112_1_En_10_Fig5_HTML.png

Fig. 10.5

Measurement of the subtalar vertical angle in 3D (SVA3D). (a) The surface of the posterior facet of the subtalar joint was marked (red contour). The most posterior and anterior point of the marked surface was calculated in the direction of the AP (x-) axis (blue dots). This allowed to determine the length of the posterior facet by a software-operated connection of both points. The midpoint of this distance was calculated and used as an origin to fit a plane parallel to the coronal plane at a distance of −5 mm, 0 mm, and + 5 mm to mimic, respectively, the posterior, middle, and anterior SVA as described by Colin et al. [9]. (b) The contour of the posterior facet running in the middle subtalar plane was used to determine the inclination (dashed line) by connecting the calculated most medial with the most lateral point. (c) The intersection of this subtalar axis with the vertical (z-) axis became the middle SVA. (d) Depiction of the middle SVA in a 3D hindfoot configuration

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Apr 25, 2020 | Posted by in MUSCULOSKELETAL MEDICINE | Comments Off on and Correlation Analysis of Computed Methods to Convert Conventional 2D Radiological Hindfoot Measurements to 3D Equivalents Using Weight Bearing CT

Full access? Get Clinical Tree

Get Clinical Tree app for offline access