T. M. Bach, L. J. Barnes, O. M. Evans
La Trobe University, Melbourne, Australia

INTRODUCTION

Inertial characteristics of lower extremity prostheses have received increasing attention in recent years as a result of considerable speculation that the asymmetric gait of transfemoral (TF) amputees is related at least in part to altered limb mass and mass distribution (Mena et al., 1981; Tsai and Mansour, 1986). Nevertheless, systematic investigations of the inertial characteristics of limb segments of amputees are very rare. In studies of amputee locomotion, investigators usually make estimates of prosthetic limb segment parameters on an individual basis, an approach which is well justified given the variety of amputation levels and prosthetic componentry available. However, for computer modelling applications, it would be useful to have normative data based on averages for an amputee population similar to the often cited proportional data for normal subjects (eg. Winter, 1979). The purpose of this investigation was to provide descriptive data on segment inertial characteristics of the prosthetic limbs and amputation stumps of a sample of TF amputees.

METHOD

Subjects were nine transfemoral amputees who attended teaching clinics at the Division of Prosthetics and Orthotics, La Trobe University. Four subjects had definitive limbs of exoskeletal construction which they agreed to provide for a short time while measurements were made. Modular endoskeletal limbs manufactured by students to the stage of dynamic alignment were available for eight subjects. These limbs consisted of ISNY sockets, Otto Bock 3R15 stance control knees, tubular aluminum pylons with stainless steel adaptors and SACH feet.

The prosthetic limbs were disassembled into socket and shank/foot segments. Segment mass was determined by weighing. Segment center of mass (CM) location was determined by a suspension method. Segment mass moment of inertia was determined by pendular suspension. Stump characteristics were estimated by modelling the stump as a stack of cylindrical sections representing the portion of the stump contained within the socket and a circular paraboloid representing the portion of the stump between the ischial seat and the hip axis (Hatze, 1979). Dimensions of these shapes were estimated from internal measurements of the socket obtained using a perimeter guage. Body mass and stature of the amputees were recoded while wearing prostheses. Corrected body mass (CBM) was estimated by subtracting prosthesis mass and estimated stump mass and by assuming that the remainder was .839 of intact body mass (Winter, 1979).

RESULTS AND DISCUSSION

The results of this survey demonstrated clear differences between shank/foot segments of endoskeletal and exoskeletal prostheses but little difference between thigh (stump/socket) segments (Table 1). The inertial characteristics and CM locations of the segments of transfemoral prostheses differed substantially from those of a normal lower limb (Table 2). The prosthetic thigh segments measured less than 70% of the mass of a normal segment. The CM was substantially more distal and the radius of gyration substantially smaller. For the prosthetic shank/foot segment, similar differences were observed: the prosthetic segments had substantially less mass and substantially more distal CM locations but had similar radii of gyration.

In order to estimate the possible effect of these differences in inertial parameters on swing phase duration of TF amputees, the half periods of oscillation of the shank/foot only and of the total lower limb with the knee extended were made for an anthropometrically average male (1.78 m stature) using the data presented in Table 2 and the proportionality data reported by Winter (1979). Although Grieve and Gear (1966) have demonstrated that such estimates exceed swing phase durations in a normal population, the error observed was a systematic one and differences would therefore be at least indicative of the likely effect on swing phase duration. These results are shown in Table 3. The estimated half periods of oscillation for the shank/foot were increased by 7% to 8% over normal but the half periods for the total limb were decreased by 2% to 5% depending on prosthesis type. Furthermore, despite large apparent differences in mass and mass distribution, the endoskeletal and exoskeletal limbs had half periods which were very similar. In samples of unilateral male TF amputees, Murray et al. (1980, 1983) observed increases in swing phase duration averaging 32% for amputees using constant friction knee components in exoskeletal limbs and 14% for amputees using hydraulic knee components in exoskeletal limbs. Although the analysis presented here is simplistic, it suggests that the changes in prosthetic limb inertial characteristics could account only in part for the increase in swing phase duration observed in transfemoral amputee gait.

REFERENCES

Grieve, D.W. and R.J. Gear (1966). Ergonomics 5:379-399
Hatze, H. (1979). A model for the computational determination of parameter values of anthropomorphic segments. Pretoria: National Research Institute for Mathematical Sciences.
Mena, D. et al. (1981). J. Biomech. 14:823-832.
Murray, M.P. et al. (1980). Bull. Pros. Res. (BPR 10-34) 17:35-45.
Murray, M.P. et al. (1983). Arch. Phys. Med. Rehabil. 64:339-345.
Tsai, C.S. and J.M. Mansour. J. Biomech. Eng. 108:65-72.
Winter, D.A. (1979). Biomechanics of Human Movement. New York: John Wiley.

TABLE 1. Measured characteristics for amputation stumps, prosthetic sockets and prosthetic shank/foot sections. Center of mass (CM) location for the stump and socket is measured from the greater trochanter. CM location for the shank/foot is measured distal to the knee axis. Moment of inertia (I) is measured about a transverse axis through the CM. Standard deviations are indicated in parentheses. * indicates statistically significant difference between exoskeletal and endoskeletal prostheses (p<.001).

TABLE 2. Proportionality data for prosthetic shank/foot and thigh (stump/socket) with comparative data for normals based on Winter (1979). Masses are expressed as proportions of CBM. CM locations are expressed as proportions of segment length from the proximal axis. Radii of gyration (K) about a transverse axis through the CM are expressed as proportions of thigh length and shank length respectively. Standard deviations are indicated in parentheses.

TABLE 3. Estimated half periods of oscillation for normal and prosthetic limbs. Periods were estimated for a subject of 1.78m stature based on data from Table 2.