The prosthetic socket plays a critical role in providing comfort, appropriate load transmission, and efficient movement control in people with transtibial amputation (TTA).1 The adverse consequences of ill-fitting socket or improper loading include soft tissue injury, bleeding, bruising, pressure sores, and pain, which in the short-term reduce the functional ability of people with amputation. If left unaddressed, these can diminish the individual’s health and quality of life.2,3
The interface pressure distribution between the prosthetic socket and the residual limb (residuum-socket interface [RSI]) is a critical consideration in the design and fit of a socket to a patient.4,5 Proper distribution of an individual’s weight across the surface of the residuum plays a critical role in the comfort and function of people with TTA.6,7 Furthermore, the design of the socket is dictated by the type of amputation that individuals undergo. Unlike traditional or standard approach to TTA, proponents of the transtibial osteomyoplastic amputation (TOA) procedure contend that the TOA allows the functional restoration of the residual limb extremity by providing end-bearing capacity.8 In the current case study, pressure distribution at the RSI of a patient with TOA and its relationship to the type of prosthetic foot was investigated.
Contrary to the traditional amputation technique, upon TOA, the residual limb is expected to function as normally as possible, including the load bearing along the axial length of the residuum. Such an end-bearing limb can theoretically stimulate the skin and deep tissue and minimize atrophy of the residuum due to disuse. End-bearing is also expected to reduce pain, improve sensation, and blood flow, and improve walking ability and prosthetic wear. Although many studies have examined pressure distribution inside the TTA prosthetic socket with a wide variety of prosthetic alignments,9,10 socket materials,11 types of socket,3,4,12,13 or types of gait activity,14,15 there is little reported research on the effect of different prosthetic feet on pressure distribution inside the socket of those with TOA.
End-bearing of the distal residuum of a patient with TOA procedure has been previously confirmed.16 The present case study investigated the effect of prosthetic feet on the contact pressure inside the total surface bearing (TSB) socket of a patient who had undergone TOA procedure. As the result of the TOA procedure, it is anticipated that during gait activities, loading pressure is observed at the distal end-bearing point of the residual limb (under the bony bridge, which stabilized the residual limb anatomy) as well as middle posterior point (where the length-tension relationship of the residual gastrocnemius muscle was retrieved). Moreover, such TOA procedure allows the patient to wear a TSB socket, which results in uniform pressure distribution at four proximal locations (anterior, posterior, lateral, and medial). The aims of this study were to 1) study the variation in RSI pressure produced by a person with TOA at the distal anterior end-bearing, middle posterior, and four proximal points during gait activities (normal and fast pace forward walking, stair ascending and descending, ramp ascending and descending) and 2) examine the effect of three prosthetic feet (Renegade Foot®, Venture Foot™, and Proprio Foot®) on pressure distribution and temporal characteristics during various gait activities. The results are expected to elucidate the potential benefits of the TOA procedure and provide insight into optimal prosthetic foot design to improve comfort and health of people with TTA.
The subject is a 43-year-old man with left TOA due to a traumatic event who is 1.91 m (75 in) in height and weighs approximately 109.1 kg (240 lb). He is otherwise healthy without comorbidity and gave informed consent to be a participant in this study. After amputation and provision of a prosthesis system, the subject returned to walking on various terrains without an assistive device. His prosthesis has an end-bearing, modified TSB socket with 6-mm Alpha™ Locking liner (Mt. Sterling, OH) and Bulldog™ Lock (Lewisburg, OH). The subject was able to confidently perform various walking activities on different terrains. The protocol was approved by the Institutional Review Board at The University of Oklahoma Health Sciences Center for the protection of human subjects.
In this study, the subject performed the following six gait activities: walking forward on level ground with self-selected pace (Forward), walking on level ground with fast pace (Forward brisk), ascending and descending a staircase (Stair-up and Stair-down; 53°, 24 steps), and ascending and descending a ramp (Ramp-up and Ramp-down) designed for wheelchair access (8°). Contact pressures at the RSI during each of these tests were recorded during three trials while the subject was using one of three types of prosthetic feet, all of the size 27 cm. The feet studied were the Renegade Foot from Freedom Innovations,17 Venture Foot from College Park,18 and Proprio Foot from Össur (Reykjavik, Iceland).19 The Renegade Foot has a deformable carbon fiber spring-type keel to absorb shock and provide dynamic response. The Venture Foot provides significant energy return and has a multi-axial configuration that allows maximal terrain compliance. The Proprio Foot is an advanced design with the ability to detect the walking terrain and automatically adjust the ankle angular position. During the study, the subject wore his own prosthetic socket for consistent RSI fit. All prosthetic feet were aligned by a licensed prosthetist for consistency in fit and function during each of the trials. The subject reported that the alignment was comfortable to wear during his usual daily activities.
Force resistive sensors were placed inside the socket to measure contact force (pressure) at the distal anterior end-bearing point, posterior point at the middle third level, and four points (anterior, posterior, lateral, and medial) at the proximal third level of the residual limb. The locations of these sensors are shown in Figure 1. Applied pressure on all sensors was sampled at the rate of 500 Hz and collected by 12-bit data acquisition and control board PC-CARD-DAS16/12 AO (Measurement Computing) that was installed in a tablet computer. The tablet computer and signal conditioning hardware were placed in a backpack and carried by the subject. The thinness of the force sensor and the compact size of acquisition hardware made it possible to measure the RSI pressures during data collection without having any adverse impact on the gait of the individual.
The overarching purpose of this case study was to investigate the effect of different prosthetic feet on RSI pressure in a person with TOA. We hypothesized that different prosthetic feet result in different pressure distributions. During the same gait activity with the same prosthetic foot, the greatest peak pressure (hypothesis H1) and mean pressure (hypothesis H2) would be detected at the distal anterior end-bearing location, compared with the highest peak and mean pressures recorded at other locations. The investigators also contend that during the same gait activity at the same measured location, peak (or mean) pressures would differ between any two prosthetic feet, that is, between Renegade Foot and Venture Foot (hypothesis H3), between Venture Foot and Proprio Foot (hypothesis H4), or between Proprio Foot and Renegade Foot (hypothesis H5).
Signals collected by the force sensors were converted into pressure values, filtered by a low pass filter with a cutoff frequency of 3 Hz, and normalized to 100% of the gait cycle. Each gait activity cycle was divided into stance phase when there was loading on the residual limb and swing phase when the residual limb was relaxed from loading pressure. From each step, the peak and mean pressures over the stance phase were calculated for each sensor. Figure 2 shows the average pressures recorded at each of the sensor locations with different prosthetic feet and gait activities. The mean sustained pressure (MP80+, average value of all pressure values above 80% of the peak pressure) was also calculated to represent sustained submaximal tissue loadings.9,14 For each prosthetic foot, the coefficient of variation (CV) of MP80+ over all gait activities was calculated at each measured location. It was assumed that, at each location, a prosthetic foot that results in a higher coefficient would create more variations in mean sustained pressure between the different gait activities. Similar calculations were performed for each gait activity and each prosthetic foot.
Hypothesis 1 will be verified by comparing the peak pressures measured by the distal anterior end-bearing sensor against the highest peak pressures from other sensors during the same activity and with the same prosthetic foot. If H1 is true, then maximal pressure recorded at the distal end-bearing location would be greater than the pressure measured at any other locations on the RSI. Similarly, for each prosthetic foot, support of H2 implies that the mean pressure is also greatest at the distal end-bearing location, compared with any other location studied. To demonstrate this, H1 and H2 will be tested using two-sample, right-tailed t-tests (p < 0.05). Because there are three prosthetic feet and six gait activities, H1 and H2 will each be tested for 18 different combinations.
Between any two prosthetic feet, testing of H3, H4, and H5 will be performed to evaluate an assumption that during the same gait activity, two different feet will result in different pressures at the same location on the RSI. Each of these hypotheses will be tested using two-sample, two-tailed t-tests (p < 0.05) and will be repeated for 72 combinations (six sensor locations × six gait activities × two measurements, i.e., peak and mean). For example, support of H3 when comparing the peak pressure at the distal anterior end-bearing during Forward gait implies that when the subject walked on level ground with self-selected pace, the Renegade and Venture feet result in different peak pressures on the RSI.
Figure 2 shows the average pressures (in kPA) recorded at each of the sensor locations during gait activities. The duration for each pressure curve is normalized as 100% of the corresponding stance phase. Swing phases are not shown because there is no contact during that gait duration.
The mean sustained pressures (MP80+) at the distal anterior end-bearing and middle posterior sensors during six gait activities with three prosthetic feet are shown in Figure 3. In each column, the MP80+ at the distal anterior end-bearing location is represented by the light horizontal bar, whereas the dark horizontal bar represents the MP80+ at the middle posterior. Across different gait activities, the mean sustained pressures (in kPA) at the distal sensor were 1049 ± 242, 1044 ± 194, and 723 ± 127 for Renegade, Venture, and Proprio feet, respectively. These measures at the middle posterior were 69 ± 8, 58 ± 10, and 385 ± 79, respectively, for the three feet.
In Figure 4, the mean sustained pressures (MP80+) (in kPA) at four proximal sensor locations are displayed for six gait activities. MP80+ at each sensor location is represented by a horizontal bar with corresponding dark level. At the proximal anterior location, the mean sustained pressures (MP80+) across all gait activities were 108 ± 16, 96 ± 28, 254 ± 44 for Renegade, Venture, and Proprio feet, respectively. Similarly for the three feet, the average MP80+ values across the other three sensors at the proximal level (proximal posterior, proximal lateral, and proximal medial) were 70 ± 12, 63 ± 10, and 122 ± 29, respectively.
Table 1 shows the test results of H1 and H2, which test for the significant differences in peak and mean pressure between the distal anterior end-bearing sensor and other sensors inside the socket. Hypothesis 1 was supported in all cases (p < 0.001). Hypothesis 2 was supported in 17 of 18 cases, except with Proprio Foot during walking up a ramp (p = 0.215).
Tables 2 to 7 show the test results of H3 to H5, which test for significantly different peak and mean pressures between different prosthetic feet during the same gait activity. Each table shows the test result for one individual sensor. At the distal anterior end-bearing (Table 2), H4 and H5 were accepted in all cases (p < 0.001). Hypothesis 3 was rejected in only three cases that test for different mean pressures at this location between the Renegade and Venture feet during forward brisk (p = 0.019), walking up a ramp (p = 0.775), and walking down a ramp (p = 0.026). In Table 3, H3 was rejected in two cases that test for the different peak pressures (p = 0.022) and mean pressures (p = 0.184) at middle posterior location between Renegade and Venture feet during walking down a staircase. Otherwise, H3 to H5 were supported (p < 0.001) in 34 of 36 cases at the middle posterior. In testing for different peak pressures among prosthetic feet at proximal anterior (Table 4), proximal posterior (Table 5), proximal lateral (Table 6), and proximal medial (Table 7), respectively, H3 to H5 were rejected in one, one, six, and three cases compared with four, two, seven, and five rejected cases in testing for different mean pressures at the same proximal locations.
The CVs of the mean sustained pressure (MP80+) over six gait activities are shown in Table 8. This calculation provides some insight into the effect of prosthetic feet in varying contact pressures at each sensor location when the subject changes his gait activities. At the distal anterior end-bearing sensor, when the subject changes between different activities, the mean sustained pressure varied from 17.6% (Proprio Foot) to 23.1% (Renegade Foot). The Renegade Foot varied the MP80+ by 12.1% at the middle posterior, 14.5% at the proximal lateral, and 7.8% at the proximal medial, whereas the Proprio Foot correspondingly affected up to 20.5%, 20.4%, and 33% at these locations. Venture Foot had little effect on MP80+ at the proximal posterior (16.8%) and much effect on MP80+ at proximal anterior (28.8%).
Finally, Table 9 shows the stance and gait cycle duration and their CVs across all feet and all gait activities. Across six different gait activities, the stance duration (in milliseconds) varied from 700 ± 29 to 840 ± 15 for Renegade Foot, from 650 ± 27 to 860 ± 23 for Venture Foot, and from 620 ± 35 to 780 ± 82 for Proprio Foot. Gait cycle duration (in milliseconds) also varied from 960 ± 36 to 1200 ± 27 for Renegade Foot, from 1000 ± 23 to 1210 ± 36 for Venture Foot, and from 990 ± 29 to 1130 ± 19 for Proprio Foot. Variations in stance phase across gait activities were 1.8% to 5.0% for Renegade Foot, 1.8% to 4.2% for Venture Foot, and 3.2% to 10.5% for Proprio Foot. Variations in gait cycle duration for Renegade, Venture, and Proprio feet were 2.2% to 3.8%, 1.8% to 6.2%, and 1.7% to 8.1%, respectively.
EFFECT OF PROSTHETIC FEET ON DISTAL ANTERIOR END-BEARING PRESSURE
First, significant peak and mean pressures that were recorded at the distal anterior end-bearing location with all three feet during all gait activities verify that the end-bearing occurred with the TOA procedure. As seen in Figure 2, the distal anterior end-bearing sensor measured the greatest pressure compared with the other sensors at any other locations. This was seen with all prosthetic feet during all of the gait activities. Figures 3 and 4 also show that the mean sustained pressure (MP80+) is also higher at the distal anterior end-bearing location compared with any other studied locations inside the socket. Moreover, such end-bearing is not only affected by the type of gait activities but also, to some extent, dependent on the type of prosthetic feet.
Table 1 shows the peak and mean pressures measured by the distal anterior end-bearing sensor and test results of H1 and H2. Hypothesis 1 was supported in all cases (probability of the null hypothesis p < 0.001), thus indicating that the TOA procedure, in conjunction with the end-bearing socket, resulted in the maximum peak pressure at the distal anterior location in the socket during gait. The only exception when H1 was rejected (p = 0.215) was in the case of the Proprio Foot, wherein the mean pressure at the distal anterior location while the subject was ascending the ramp (424 ± 43) was similar to the greatest mean pressure recorded at other sensor locations in the socket (411 ± 18). As shown in Table 2, during a gait activity, different mean pressures were observed when using different prosthetic feet. However, no appreciable differences were observed between Renegade Foot and Venture Foot during Forward brisk, Ramp-up, and Ramp-down gait activities (p values of 0.019, 0.775, and 0.026, respectively). All failed cases involved the comparison between the Renegade and Venture feet, whereas hypotheses involving Proprio Foot were supported. These results might be due to the adaptation nature of the Proprio Foot, which places it in a different configuration at the heel strike moment compared with the other feet. Table 8 shows that submaximal pressure at the distal anterior end-bearing location is affected to a greater extent by Renegade Foot (CV, 23.1%) than by the Proprio Foot (CV, 17.6%). In other words, the Proprio Foot generated a greater consistent sustained pressure at the anterior sensor at the distal residuum, when the subject changed his gait activities.
EFFECT OF PROSTHETIC FEET ON MIDDLE POSTERIOR PRESSURE
The RSI pressures at the middle posterior location were similar in the tests conducted when the subject was using the Renegade Foot and the Venture Foot but significantly greater when using the Proprio Foot (Figures 2 and 3). This effect of the Proprio Foot on the middle posterior location is in agreement with the results reported by Wolf et al.20 The peak and mean pressures observed at this location (Table 3) are both characteristic of the foot used and validate H4 and H5; that is, the observed values are distinct from those seen when either the Renegade Foot or the Venture Foot is used. Even when either Renegade Foot or the Venture Foot is used, the observed pressures at the middle posterior location on the RSI are distinct in a majority of the test cases, and H3 is validated in all but two instances when the subject was descending the stairs (Stair-down). Similar to hypothesis tests at the distal anterior end-bearing sensor, failed tests at the middle posterior did not involve Proprio Foot. Again, the adaptation mechanism of the Proprio Foot might have a distinct effect on the pressure observed at this location. In addition, there was less variability in mean sustained pressure (CV, 12.1%) using the Renegade Foot than the Proprio Foot (CV, 20.5%) when the gait activities were switched.
EFFECT OF PROSTHETIC FEET ON THE OBSERVED PRESSURE AT THE PROXIMAL LOCATIONS ON THE RESIDUUM-SOCKET INTERFACE
During all gait activities with each of the feet considered in this study, the observed pressure at the proximal locations was similar for all sensor locations, with slightly higher values in the proximal anterior location (Figures 2 and 4). Between any two prosthetic feet, significantly different peak pressures were detected in 61 of 72 cases, whereas significantly different mean pressures were detected in 54 of 72 cases (Table 4–7). The use of Renegade Foot resulted in less variability in mean sustained pressure at the proximal locations (7.8% to 15%) compared with the Proprio Foot (20.4% to 33%) for each gait considered in this study (Table 8). Moreover, when testing for different peak pressures at proximal locations, H3 to H5 are rejected in fewer cases compared with when testing for different mean pressures at the same proximal locations. These observations suggest that although different prosthetic feet produce similar mean pressures at the proximal level during the entire stance phase, they still have different effects on the peak pressures. Greater instances of failed tests were observed using data from the proximal sensors (anterior, 5 cases; posterior, 3 cases; lateral, 13 cases; medial, 8 cases) compared with data using the distal anterior end-bearing sensor (3 cases) and the middle posterior sensor (2 cases). Furthermore, at the proximal level, failed tests also involved Proprio Foot, which was in contrast to the outcomes of the hypothesis tests using data from other sensors (middle and distal). This suggests that the adaptation mechanism in the Proprio Foot has less effect on the pressure at proximal sensors than the pressure at distal anterior end-bearing and middle posterior locations. There are greater number of failed tests along the lateral-medial direction (proximal lateral, 13 cases; proximal medial, 8 cases) than failed tests along the anterior-posterior direction (proximal anterior, 5 cases; proximal posterior, 3 cases). This observation suggests that the difference in prosthesis mechanism might have more effect on the pressure distribution along the anterior-posterior direction than the lateral-medial direction.
EFFECT OF PROSTHETIC FEET ON TEMPORAL GAIT PARAMETERS
The duration of the stance phase and gait cycle when using each of the prosthetic feet is shown in Table 9. Although the duration of the gait cycle was similar for each type of gait considered in this study, significant variation in the duration of the stance phase was observed between the corresponding values for the three feet examined in this study. The duration of the stance when using the Proprio Foot was significantly lower than the duration observed for the Renegade or Venture feet during all types of gait except Ramp-up. Significant variability in the duration of stance and gait was also observed for all types of gait when the subject used the Proprio Foot. When the subject changed his gait activities, the stance phase and gait cycle duration, respectively, varied up to 10.5% and 8.1% with the Proprio Foot. These measures were, respectively, 5.0% and 3.8% with Renegade Foot and 4.2% and 6.2% with Venture Foot. Higher variations in stance phase and gait cycle duration when the subject used the Proprio Foot might be a result of the adaptation process of the joint.
The results from the tests conducted in this study demonstrate that the type of foot used by the subject has a significant impact on the contact pressure at distinct locations on the residual limb. This observation provides credence to the principles motivating the design of prosthetic feet, as well as the role of the dynamic response characteristics of these feet in generating desired distribution of pressure within the socket.
LIMITATIONS OF THE STUDY
This case study involved only one subject with TOA. In addition, a comparison between the TOA subject and subjects with traditional amputation was not performed. Future study will include transtibial amputees with various amputation techniques and prosthetic designs. Additional investigation along this line will further illuminate the effect of different prosthetic foot designs on the RSI pressure in subjects with TOA.
This article provides evidence that the choice of prosthetic foot impacts pressure distributions inside the socket during six gait activities in an amputee who underwent TOA procedure. Significant end-bearing, which is expected in a TOA amputee, was observed in all gait activities. Whereas nominal pressures were observed in the proximal region, significantly greater pressure was observed at the distal anterior end-bearing region inside the socket. Moreover, the series of statistical tests conducted in this study showed variability in contact pressures across each prosthetic foot. It is also evident that the prosthetic foot has a direct effect on temporal gait parameters during the gait activities. Taken as a whole, these results indicate that for the same socket design, loading characteristics inside the socket depend on the selected gait in combination with the type of prosthetic foot used by the subject. Hence, this study provides preliminary evidence that the design of the prosthetic foot modifies the loading response inside the socket and potentially impacts the user’s comfort during selected walking activities.
Although this study was limited to one amputee using three prosthetic feet during a variety of gait activities, it elucidates the important role of prosthetic feet in providing desired distribution of contact pressure on the residual limb. Such investigations provide insights that can inform in the selection of a proper prosthetic component to meet the comfort and performance requirements of each client with amputation.
We thank our colleagues at the School of Electrical and Computer Engineering, University of Oklahoma, and The University of Oklahoma Health Sciences Center for recruiting the patients and participating in collecting the data for this study.