INTRAOPERATIVE PEDOGRAPHY – A NEW VALIDATED METHOD FOR INTRAOPERATIVE BIOMECHANICAL ASSESSMENT

SUMMARY
Background
Pedography (P) is known to be one of the most efficient methods for biomechanical assessment of the foot. A new device was developed to perform intraoperative P. The purpose of this study was to validate the introduced method by a comparison with the standard method for dynamic and static P.

Methods
A device named Kraftsimulator Intraoperative Pedographie® (KIOP®) was developed for an intraoperative introduction of standardized forces to the sole of the foot. The measurements were performed with a custom-made mat with capacitive sensors (PLIANCE®, Novel Inc., St. Paul, MN, USA). The validation was performed in two steps:
Step 1. Comparison of standard dynamic P, static P in standing position and P with KIOP® in supine position in 30 healthy subjects.
Step 2. Comparison of P in standing position, P with KIOP® in supine position, and P with KIOP® in anaesthetized (general or spinal anaesthesia ) subjects in supine position. Subjects who had operative procedures performed at the knee or distal to the knee were excluded. 30 Subjects were included. The different measurements were compared (ONEWAY ANOVA, t-test; significance level 0.05).

Results
No significant differences between the all measurements of step 1 and 2 were found for the hindfoot versus midfoot/forefoot force distribution. For the medial versus lateral force distribution and the 10-region-mapping, significant differences were found when comparing all measurements (step 1 & 2), and when comparing the measurements of step 1 only. No differences were found for these distributions when comparing the measurements of step 2 alone, and when comparing the measurements of step 1 and 2 without the platform measurements of step 1 (dynamic walking pedography and static standing pedography). No significant differences in the force distributions were found in step 2 when comparing subjects without, with general anaesthesia, and with spinal anaesthesia.

Conclusion
The new method allows a valid intraoperative pedography since no statistical significant force distribution differences were found between standing subjects and anaesthetized subjects in supine position.

INTRODUCTION
For any kind of reduction or correction procedures at the foot and ankle an immediate biomechanical assessment after the reduction or correction would be desirable 1-11. This is especially true for Computer Assisted Surgery (CAS) guided reductions or corrections, that these are more accurate than a conventional reduction 12. The assessment of the accuracy of the reduction or correction is performed with a conventional C-arm or a with intraoperative threedimensional imaging with ISO-C-3D if available 11. Analyzing the position of the bones radiographically allows conclusions regarding the biomechanics of the foot1;13. However, pedography is more effective for the analysis of the biomechanics of the foot 14-16. So far, pedography for biomechanical assessment was only available during clinical follow-up 17. An intraoperative pedography may be useful for immediate intraoperative biomechanical assessment 17. A new device was developed to perform intraoperative pedography.
The purpose of this study was to validate the introduced method by a comparison with the standard method for dynamic and static pedography.

METHODS
KIOP® – Kraftsimulator Intraoperative Pedographie®
For an intraoperative introduction of standardized forces to the sole of the foot, a device named Kraftsimulator Intraoperative Pedographie® (KIOP®, manufactured by the Workshop of the Hannover Medical School, Hannover, Germany; Registered Design No. 20 2004 007 755.8 by the German Patent Office, Munich, Germany & St. Paul, MN, USA) was developed (Fig. 1). This device allows an introduction of force to the sole of the foot by applying force to the knee in flexed position (Fig. 2). The pedographic measurement is performed with a custom-made mat with capacitive sensors (model Pliance®, Novel Inc., Munich, Germany & St. Paul, MN, USA) (Fig 1). The mat is connected to a standard IBM compatible laptop computer with the standard adaptor (model Pliance-X, Novel Inc., Munich, Germany & St. Paul, MN, USA). The standard software was used for the measurements (model Pliance Expert®, version 10.2.20, Novel Inc., Munich, Germany & St. Paul, MN, USA). The system allows real-time pedography and comparison to the contralateral side intraoperatively (Fig 3). The introduced total force and force distribution is displayed in real-time to control the amount and distribution of the force.
The total force in standing position of the tested subject determined the introduced total force with KIOP® for the validation process. The aimed force distribution hindfoot : midfoot/forefoot was 60 : 40, the medial : lateral was 50 : 50 as described for a standing position 16;18. The force distribution was controlled by positioning of foot and tibia for the medial : lateral distribution and flexion/extension of the knee and ankle for the force distribution hindfoot : midfoot/forefoot. A more flexed knee and more dorsally extended ankle resulted in a higher percentage of force at the midfoot/forefoot and a lower percentage of force at the hindfoot. A less flexed knee and less dorsally extended ankle resulted in a higher percentage of force at the hindfoot and a lower percentage of force at the midfoot/forefoot. The achieved distributions were analyzed during the validation process. The flexion/extension angles at the knee and ankle during the measurements were not registered.

Fig. 1. The newly developed device for intraoperative force introduction (Kraftsimulator Intraoperative Pedographie® (KIOP®), registered design no. 202004007755.8, German Patent Institute, Munich, Germany & St. Paul, MN, USA). The custom made mat for force registration (pliance®, Novel, Munich, Germany & St. Paul, MN, USA) is covered intraoperatively with a sterile plastic bag and is placed on the KIOP® as also shown in figure 6. The size of the mat is 16 x 32 cm. The mat includes 32 x 32 sensors with a sensor size of 0.5 x 1 cm.

Fig. 2. Scheme of the modus for Intraoperative Pedography (IP)

Fig. 3. Intraoperative Pedography (IP) during a correction arthrodesis at the talonavicular joint (performed during a feasibility prestudy. 300 N, 400 N and 500 N were applied with a new developed device for intraoperative force introduction ( model Kraftsimulator Intraoperative Pedographie® (KIOP®), registered design no. 202004007755.8, German Patent Institute, Munich, Germany & St. Paul, MN, USA). The force measurements were displayed in real-time on the screen of the pedography-system. During the intraoperative pedography, KIOP® is entirely sterile and the force measurement mat (model pliance®, Novel Inc., Munich, Germany & St. Paul, MN, USA) is covered by a sterile plastic bag.

Validation process
To allow validation of the method, objectivity and reliability of the method were first analyzed in detail as a basis for the following validation process.

Objectivity
The objectivity of the technical system was analyzed and approved by an introduction of standardized forces to the mat in the KIOP® (data not shown) with a standard calibration device (model Trublu Calibration Device®, Novel Inc., Munich, Germany & St. Paul, MN, USA) as described before 19. A statistical analysis of the differences of introduced forces and the measured forces showed no significant differences (data not shown).

Reliability
The reliability of the technical system was analyzed and approved by a repeated introduction (ten times) of standardized forces to the mat in the KIOP® (data not shown) with a standard calibration device (model Trublu Calibration Device®, Novel Inc., Munich, Germany & St. Paul, MN, USA) as described before 19. A statistical analysis of the differences of the repeated measurements showed no significant differences (data not shown).

Validity
The validation was performed in two steps. A statistician determined the number of subjects for both steps after a review of the study design, and before starting the study. Step 1 included 30 healthy subjects without past history regarding both lower extremities (Table 1). A comparison between standard dynamic pedography (three trials, walking, third step, mid stance force pattern as described before 20;21, static in standing position (three trials) and pedography with KIOP® in supine position (three trials, total force determined by total force in standing position comparable to half body weight) (Table 1, Fig. 4). For dynamic pedography and pedography in standing position, a standard platform (model Emed AT®, Novel Inc., Munich, Germany & St. Paul, MN, USA) and software (model Emed ST®, version 12.3.18, Novel Inc., Munich, Germany & St. Paul, MN, USA) was used. Both sides were measured.

Fig. 4. Images from step 1 of the validation study; conscious subject; left, standard dynamic pedography; middle, static pedography in standing position; right, pedography with KIOP®. For dynamic pedography and pedography in standing position, a standard platform Emed®, Novel Inc., Munich, Germany & St. Paul, MN, USA) was used. All three images show increased forces beneath the 1st metatarsal head and the 1st toe.

Step 2 included a comparison of 30 patients in the department where the study was carried out. These were patients who had no past medical history regarding both lower extremities below the knee (Table 1). Pedography in standing position, pedography with KIOP® in non-anaesthetized and anaesthetized subjects (three trials, total force determined by total force in standing position comparable to half body weight) was performed (Fig. 5). For all measurements including the pedography in standing position the mat (model Pliance®, Novel Inc., Munich, Germany & St. Paul, MN, USA) and standard mat software (model Pliance Expert®, version 10.2.20, Novel Inc., Munich, Germany & St. Paul, MN, USA) were used. Both sides were measured. Subjects with operative procedures performed at the knee or distal to the knee were excluded. Only subjects with general or spinal anaesthesia were included. 17 subjects sustained general anaesthesia and 13 spinal anaesthesia. The anaesthetized subjects were measured in the anaesthetic room before entering the operating theater. The mat was covered with a plastic bag as described above.

Fig. 5. Images from step 2 of the validation study ; non-anaesthetized/ anaesthetized subject; left, pedography in standing position; middle, pedography with KIOP® in non-anaesthetized subject; right, IP in anaesthetized subject. All three images show increased forces beneath the 3 rd and 4 th metatarsal.

For both steps a standard computerized mapping to create a distribution into the following foot regions was performed with the standard software (model Automask®, version 12.3.18, Novel Inc., Munich, Germany & St. Paul, MN, USA): hindfoot, midfoot, 1 st metatarsal head, 2 nd metatarsal head, 3 rd metatarsal head, 4 th metatarsal head, 5 th metatarsal head, 1 st toe, 2 nd toe, 3 rd -5 th toe (Fig. 6, Table 2). The percentages of the overall forces at the different regions were compared according to the different measurements (Table 2).

Fig. 6. Image from Intraoperative Pedography (IP) after computerized mapping. The following regions are defined by the mapping process: M1, hindfoot; M2, midfoot; M3, 1st metatarsal head; M4, 2nd metatarsal head; M5, 3rd metatarsal head; M6, 4th metatarsal head; M7, 5th metatarsal head; M8, 1st toe; M9, 2nd toe; M10, 3rd-5th toe.

Ethical approval
The study was approved by the Ethical Commission of the Hannover Medical School , Hannover , Germany . Informed consent was obtained from all subjects included in the study.

RESULTS
Table 2 indicates force distributions of step 1 and 2 and the results of the statistical analysis. No significant differences between the all measurements of step 1 and 2 were found for the hindfoot versus midfoot/forefoot force distribution. For the medial versus lateral force distribution and the 10-region-mapping, significant differences were found when comparing all measurements (step 1 & 2), and when comparing the measurements of step 1 only. No differences were found for these distributions when comparing the measurements of isolated step 2, and when comparing the measurements of step 1 and 2 without the platform measurements of step 1 (dynamic walking pedography and static standing pedography) (Table 2). No significant differences in the force distributions were found in step 2 when comparing subjects with general anaesthesia with subject with spinal anaesthesia (data not shown; t-test, p>0.05).

The null hypothesis was rejected for the hindfoot versus midfoot/forefoot force distribution for all tests. The null hypothesis was rejected for the medial versus lateral and the 10-region-mapping within step 2, and within step 1 and 2 without the platform measurements of step 1 (dynamic walking pedography and static standing pedography) (Table 2). The null hypothesis was not rejected for the medial versus lateral and the 10-region-mapping within step 1 & 2, and within 2 step 1 only (Table 2).

Table 1: Percentages from total force for different foot regions for step 1 and 2 >> download

DISCUSSION

Standard pedography
The objective documentation of foot function before and after therapeutic intervention is greatly enhanced by the utilization of devices capable of measuring dynamic foot force distribution 21. Efforts to develop this technology date back to the late 19th century, but only with recent advances in computers has it been possible to produce quantitatively accurate high resolutions of foot force distribution with high sampling rates and easily interpreted graphic displays 21. Over the years, a variety of methods have been employed to study foot pressure 21-23. Many of these techniques have already improved our understanding of the foot and its function, and have had an impact on the way we practice 15;21;24.

The idea of an Introperative Pedography (IP)
This investigation was driven by the idea to profit from the advantages of the pedography-based functional analysis as described above not only pre- and postoperatively but also intraoperatively 17. The main idea was to use the data from an intraoperative pedography to detect non-optimal biomechanical conditions and to have the opportunity for immediate changes of the correction or reduction in the same procedure. This process may later lead to an improved biomechanical function in the clinical course.

Development of a system for IP
The main problem during the invention of the system for intraoperative pedography was an adequate introduction of force to the sole of the foot in an anaesthetized subject in supine position. This introduced force should be as similar as possible to a static pedography in standing position of even better to a dynamic pedography during the stance phase of walking. For that purpose, a device named Kraftsimulator Intraoperative Pedographie® (KIOP®) was developed (Fig. 1 & 2). The measurements were performed in near neutral ankle position. In this neutral ankle position, the influence of the missing muscle activity in the anaesthetized subject was considered to be minimal since the EMG in conscious standing subjects with comparable ankle position has been demonstrated to be silent 25-27. Our main concern was of course, that there are still significant differences in the force distribution patterns between the standard pedography and the intraoperative pedography. The null-hypotheses that there are differences between the different pedography-methods was therefore formulated.

The validation process
Two steps for the validation of the method were planned. Before these steps were taken, the technical objectivity and reliability of the introduced system were approved. In the first step a standard dynamic pedography during the stance phase of walking and a static pedography, both performed with a standard platform for dynamic pedography (model Emed AT®, Novel Inc., Munich, Germany & St. Paul, MN, USA), was compared to a pedography in supine position with the KIOP® and the mat (model Pliance®, Novel Inc., Munich, Germany & St. Paul, MN, USA). This step was performed in a gait laboratory on healthy subjects. The intention of this trial was to find out whether the developed system is at all able to simulate dynamic or static pedography in conscious subjects. We were aware that a dynamic pedography force distribution pattern is principally not comparable to a static pedography, neither in standing nor in supine position. Still, we believed that a comparison of the maximum forces of the entire stance phase in dynamic pedography to the static measurements may lead to a better understanding of the mechanisms during the validation process. The second step was performed on patients from our department without past medical history regarding the lower extremities below the knees. These subjects had operations on at the hips or on a higher level. For this step only the mat was used for the measurements. Static pedographic measurements in the conscious subjects in standing and supine position and measurement in the anaesthetized subject were performed and compared. For all measurements with the KIOP®, the total force was determined by the total force in standing position in the same subject comparable to half body weight to simulate the pedography in standing position as good as possible.
The analysis was focused on the force distribution and not on the force values. To allow a validation of the method, objectivity and reliability of the method were analyzed in detail as a basis for the validation process as described above. An analysis of the intra-observer reliability and the inter-observer objectivity during the validation process was not performed since the entire measurement and evaluation process is observer- or investigator-independent. The differences of the force distribution between the different methods were analyzed by the software without interaction of observers or investigators. During this analysis, the mapping of the pedographic force patterns into different regions was performed by software, as well as the calculation of the percentages from the total force in the single regions. The mapping process has been demonstrated to allow a better standardized analysis of force patterns than an observer based subjective analysis 28. Following, all resulting values were statistically analyzed. No subjective-based analysis was performed during the entire validation process at all.

Findings
Our results demonstrate that the force distribution measured with the introduced method in anaesthetized subjects was not significant different to the force distribution in standing position for all compared regions, i.e. hindfoot versus midfoot/forefoot, medial versus lateral, and the 10-region-mapping (step 2). Furthermore, the hindfoot versus midfoot/forefoot force distribution did not differ between all measurements including the dynamic pedography in two different subject cohorts and bilaterally. This is very important result because a adequate 60 : 40 hindfoot : midfoot/forefoot force distribution is a condition sine qua non in healthy subjects 16. Surprisingly, we found no significant differences in the force distribution between the KIOP® measurements of step 1 and step 2, despite examining different subjects and two different sides. Based on our results we were able to reject most of the null hypothesis. Consequently, we consider the introduced method for intraoperative pedography as valid for introperative biomechanical assessment despite the methodological weaknesses as following.

Methodological weaknesses
Our main concern was that the force induction with KIOP® to the sole of the foot was insufficient. Secondly, we considered the supine position and, of course, the anaesthesia as an important cause for an unphysiological pedographic force distribution pattern. To minimize the influence of the missing muscle activity we measured in near neutral position in the ankle since the EMG in conscious, standing subjects with comparable ankle position has been demonstrated to be silent 25-27. Still, we could not eliminate the other potential disturbing factors.
Therefore, the null-hypotheses were formulated to detect the potential differences between the standard pedography and the introduced methods that were affected by these factors. Subjects without known foot pathologies were included in the validation process. One could argue that this validation is therefore only sufficient for “healthy” feet. However, we detected also non-physiological force distribution patterns in the included subjects as for example shown in figure 5. Consequently, we consider the introduced method to be comparable to the standard pedography for detection of pathologies.
Finally, we did not really measure intra-operatively but in the in the anaesthetic room before the subjects entered the operating theater. The setting for the validation process was absolutely comparable to a real intra-operative pedography as the same devices, including the plastic bag covering the mat was used as for the intra-operative setting as shown in figure 3.

In conclusion, the introduced method allows a valid intraoperative pedography since no statistical significant force distribution differences were found between standing subjects and anaesthetized subjects in supine position. This system may also be able to detect non-optimal biomechanical conditions of the foot intraoperatively and may lead to modifications of a reduction or correction in the same operative procedure 17. A clinical study is planned to evaluate the benefit of the introduced system during the operative correction of foot pathologies.

Acknowledgements
The authors thank Birgitt Wiese, PhD (Institute for Biometry, Hannover Medical School, Hannover, Germany) for her help and support in carrying out the extensive statistical analysis and for her unbiased prior evaluation, Novel Inc., Munich, Germany & St. Paul, MN, USA for making the extensive equipment available, and Julia Mues, Leighton Hospital, Crewe, UK for her review of the manuscript and the far-reaching language editing.

REFERENCE LIST