Secondary Logo

Journal Logo


Roentgen Stereophotogrammetric Analysis of Motion between the Bone and the Socket in a Transtibial Amputation Prosthesis

A Case Study

Söderberg, Bengt; Ryd, Leif; Persson, Björn M.

Author Information
JPO Journal of Prosthetics and Orthotics: July 2003 - Volume 15 - Issue 3 - p 95-99
  • Free


Sport activities for lower limb amputees have increased during the last decade. Paralympic games have created publicity and increased expectations for ordinary users to participate in sports and to function better in daily life activities. Energy-returning feet, carbon fiber laminated constructions, and improved socket constructions have gradually been developed to benefit a majority of users. To optimize the function of a prosthesis for a lower limb amputee, the attachment of the prosthesis to the residual limb is of critical importance. Some authors believe that the suspension needs to be secure and the stability, or with mechanical terms, the stiffness between the bone and the socket of the prosthesis should be as high as possible. This means that the position of bone versus socket should not change much between swing-phase and stance. This has led to the development of osseous integration attachment using a titanium plug in the distal end of the transected femur has shown this. Such patients have reported a feeling of improved balance. This has been called osseous perception by the inventor group in Gothenburg, Sweden.1 This must be accomplished, however, without causing pain or skin breakdown in conventional stumps. Therefore, soft interfaces or liners (e.g., TEC or Iceross) have been developed to protect the skin despite a tight fit and high loads during intense activities, such as running and lifting, but also to improve the suspension.

Motion between the prosthesis and the residual limb occurs in several layers (i.e., between the hard socket and the soft-liner, within the soft-liner, between the soft-liner and the skin, and between the skin and the bone). Therefore, the total motion depends on the choice of material, the shape of the socket, the type of suspension and its fit to the skin. The stability is also influenced by the quality of the stump itself concerning the amount of subcutaneous fat and the muscular coverage of the bone. Irrespective of such soft tissue stump qualities, a longer stump is easier to enclose and stabilize in a socket to create a more forceful lever arm to move the prosthesis.

Plain radiographs have repeatedly been used to analyze the motion between bone and socket during different parts of a gait cycle.2-5 The vertical motions between mid-stance and swingphase have been measured to be 2 to 4 cm and the movement in the anteroposterior direction 1 to 3 cm. With conventional radiography techniques, it has only been possible to describe the stump motion in one plane at a time. We wanted to apply the modern Roentgen stereophotogrammetric analysis (RSA) to completely characterize the motion between socket and stump in all 6 degrees of freedom of motion. The RSA technique6 has had an enormous impact on orthopaedic research, especially for analysis of migration and loosening of joint replacement prostheses. We applied RSA in a study of four different types of socket suspensions with the hypothesis that a modern thin liner should provide better stability.


A 69-year-old male patient with a transtibial amputation due to trauma 46 years before provided informed consent to participate after oral and written information. He had a 10-cm tibial length in the stump (Figures 1 and 2) and functioned well with his prosthesis without walking aids.

Figure 1.:
Sagittal plane radiograph in socket in the RSA box.
Figure 2.:
Frontal plane radiograph in socket in the RSA box.

Under local anesthesia, the tibia was provided with 6 tantalum markers Ø 0.8 mm, inserted well spread out into the metaphysis by use of a spring-loaded custom-made pistol. A wider distance between the markers increases the sensitivity of measurements proportionally and they need to be fixed to bone so as to not move with soft tissues. The method has a program to exclude such movers. Twice the minimal number of markers is always used during RSA examination to guarantee accuracy.7 The patient suffered only minor discomfort postoperatively and could ambulate fully again after 10 days. After 4 weeks, he was subjected to the first RSA stress examination. RSA exposures (A-P and lateral) were obtained with the knee inside the calibration cage with markers of defined positions on each of the four walls. Four different prosthetic suspension modes were used on the same hard socket and the same interface (TEC) (Figure 3). The position of the patient was controlled by a standing frame around the body and the knee was inside the RSA box, eliminating the possibility of lateral movement. A plumb line was marked on the prosthesis to ensure the same vertical position during the tests.

Figure 3.:
Stump in the liner used for all setups.


  1. Supracondylar suspension—creates a suspension over the femoral condyles.
  2. Patellar tendon bearing strap—a leather strap is attached to the hard socket and tightened over the femoral condyles.
  3. Distal pin suspension—a peg is attached to the bottom of the liner and locked into the hard socket.
  4. Vacuum suspension with expulsion valve—an airtight sleeve is placed over the hard socket and up onto the skin on the thigh, creating an airtight seal on the skin and the hard socket. The expulsion valve admits trapped air in the socket to expulse but does not permit air to come into the socket.

The hard socket was provided with eight tantalum markers around the periphery at two levels. For each type of suspension, four different positions were tested according to Lilja et al.4 To simulate heel contact and push-off, a 15° wedge was positioned underneath the food, inclined forwards or backwards, respectively (Figure 4).

Figure 4.:
The prosthesis on a 15° tilted plane to simulate heel strike.

The patient could stand vertically on either the toes or the heel on the tilted floor and load with the whole body weight. The floor reaction force thus affected the knee joint with an extending or flexing knee movement, as in push-off and heel contact, respectively, during normal gait. In mid-stance, the patient was standing on the plain floor.4 Normally the floor reaction force is about 80% of the body weight in mid-stance and about 115% of the body weight in push-off and heel contact at freely selected walking speed.8 With this set-up, the floor reaction force was equal to the body weight in all three positions. To simulate the swing-phase, the prosthesis was positioned at an angle of 45° relative to the floor, and a 1-kg weight was attached to the foot to simulate inertial effects.4

One RSA-examination comprised x-ray pairs of pictures at heel-strike, mid-stance, toe-off, and swing phase, hence four RSA pairs for each of four suspensions for a total of 16 paired radiographs. These radiographs were digitized using a manual digitizing table (Hasselblad Engineering, Gothenburg, Sweden). The accuracy is about 20 μm according to both the manufacturer and our earlier experiences in other similar orthopedic experiments. A virtual marker was drawn on the radiographs of the first RSA pair at the tip of the tibia; this marker was transposed to all other radiographs. The RSA software provided results concerning rotation and translation of the center of gravity of the markers in the tibia (segment motion). The accuracy of RSA has previously been determined to 0.3 degrees and 0.2 mm at the 95% confidence limit.6 The RSA software is provided with an internal control to verify that markers in each rigid body construct, i.e., that of the bone and socket, respectively, do not move relative to one another. No such loose markers were identified.


The examination was easy to perform and the patient experienced no hindrance from the RSA equipment. There was about 35 mm vertical motion of the tibia into the hard socket with the Kondylen bettlung Munster (KBM) and PTB-strap, whereas the results with the pin and the sleeve were less than half that value. In fact, there was only about 7 mm of vertical movement in the sleeve type. For all four suspension types the movement was largest between swing phase and stance phase with all four suspension types (Figure 5). The data points represent one radiographic examination each in two projections at right angles according to the method description above and the calculated true vertical segment motion of the tibia in relation to the outer prosthetic socket.

Figure 5.:
RSA measures of vertical motion of distal tibia in relation to the outer prosthetic socket.

The motion among the three phases of standing was small. In the anteroposterior direction, the tip of the tibia translated approximately 14 mm anteriorly in the socket from the toeoff to swing-phase position with the supra condylar and PTB socket (Figure 6) and again with the distal peg and sleeve suspension. With the pin and sleeve type, the maximum anterior position occurred at heel strike but was otherwise of about the same amplitude. The total mediolateral movement of the distal tip in the socket was 6 mm, but variations between the different trials in stance-phase implicated that the patient's body center of mass in relation to the prosthesis varied (Figure 7).

Figure 6.:
RSA measures of anteroposterior motion of distal tibia in relation to prosthetic socket during simulated swing phase.
Figure 7.:
Mediolateral translation of distal tibia in relation to prosthetic socket.

Segment motion analysis showed that in the transversal plane with the supracondylar socket tibia rotated 4° outwards at heel strike. This rotation continued through mid-stance to toe-off reaching 7.5° and then during swing phase, the tibia rotated inward again to the 0 position. The rotation with the pin type of socket was smaller and more erratic (Figure 8).

Figure 8.:
Rotation in horizontal plane of tibia in relation to prosthetic socket.


The results showed several centimeters of telescoping (vertical movement) of the bone in the socket during the gait cycle and the magnitude of this motion using the supracondylar and PTB-strap suspensions is compatible with previous results4 using two-dimensional radiographic techniques. The use of a peg in the bottom of the socket for suspension and the airtight sleeve mode showed less motion both in the vertical direction and in the anteroposterior direction. Rotation about the vertical axis of the tibia in the socket has not been described before. In this case, it was about 7° with the KBM and sleeve suspension and less with the pin and PTB strap. This is a logical difference because the vertical motion depends more on the suspension, whereas the anteroposterior movement depends more on the stump length, the muscle force during stance, and the socket fit. Similarly, the rotation around a vertical axis is of significant magnitude, measured here to about 7° to 10°. During natural gait, this rotation is probably even larger. This shows the importance of permitting a controlled rotation either within the interface material or through a torsion adapter. The interface used for all trials is made of 6.2-mm polyurethane. There is a flow characteristic of this polyurethane, which results in a slight sinking of the stump into the material during stance phase. This normally represents approximately 3 to 5 mm of vertical movement but the liner was the same in all four tests.

It is important to realize that this is a pilot investigation on one patient using a somewhat laboratory set-up. The study is not truly dynamic, but the gait-cycle is mimicked by four positions using full body weight. Normally the floor reaction force is approximately 80% of the body weight in mid-stance and approximately 115% of the body weight in push-off and heel contact at freely selected walking speed.4

For ethical reasons with regard to radiation burden, we could not perform multiple gait-cycles assessments. The exposure characteristics during RSA are set so that a minimum of radiation is absorbed. It has been determined previously that an RSA stress series of 12 different positions corresponded to one ordinary knee examination.7

With this new technique for measuring motion of the bone segment in a stump in relation to the socket of a prosthesis, a series of new studies can be conducted. The radiographic apparatus and the calibration cage that create the reference points on the films to be used during digitizing, however, restrict the system. Possible reference places that permit the exclusion of the calibration cage can be used to somewhat remedy this restriction. The results also highlighted the importance of controlling the upper body position. If the patient can shift his body center of mass in the horizontal plane, this will influence the sensitive measurements of motions of the distal bone segment in the socket. Different types of socket shapes and materials can be studied, and different loads and weight-bearing situations can be created. The motion can be separated by addition of markers on the skin to describe the motion between bone and skin and between bone and hard socket. Even markers on a polyurethane soft liner can be introduced. The method is used routinely at several university orthopaedic departments. Other types of studies and comparisons will be possible (e.g., prosthetic fixture directly to bone by use of titanium osseous integration technique9).

The RSA method is now finding an increased use for a multitude of research projects at different university departments. The stiffness between stump and prosthesis is maximal with an osseous integrated anchorage of a prosthesis, but a biphasic nature of stiffness might be desirable to create both security and comfort.


1. Branemark R, Branemark PI, Rydevik B, Myers RR. Osseointegration in skeletal reconstruction and rehabilitation: a review. J Rehabil Res Dev. 2001;38:175-181.
2. Eriksson U, Lundberg R. Roentgenological study of movements of the amputation stump with the prosthesis socket in belowknee amputees fitted with a PTB prosthesis. Acta Orthop Scand. 1969;40:520-529.
3. Grevsten S, Eriksson U. Roentgenological study of the stumpsocket contact and skeletal displacement in the PTB-suction prosthesis. Ups J Med Sci. 1975;80:49-57.
4. Lilja M, Johansson T, Öberg T. Movement of the tibial end in a PTB prosthesis socket: a sagittal x-ray study of the PTB prosthesis. Prosthet Orthot Int. 1993;17:21-26.
5. Newton RL, Morgan D, Schreiber MH. Radiological evaluation of prosthetic fit in below-knee amputees. Skeletal Radiol. 1988;17: 276-280.
6. Selvik G. Roentgen stereophotogrammetry. A method for the study of the kinematics of the skeletal system. Acta Orthop Scand Suppl. 1989;232:1-51
7. Ryd L. Micromotion in knee arthroplasty. A roentgen stereophotogrammetric analysis of tibial component fixation. Acta Orthop Scand Suppl. 1986;220:1-80.
8. Hermodsson Y, Ekdahl C, Persson BM, Roxendal G. Gait in male transtibial amputees: a comparative study with healthy subjects in relation to walking speed. Prosthet Orthot Int. 1994;18:68-77.
9. Gunterberg B, Brånemark PJ, Brånemark R, Bergh P, Rydevik B. Osseous integrated prosthesis in lower limb amputation. The development of a new concept. International Society for Prosthetics and Orthotics. Abstracts Amsterdam June 28-July 3, 1998.
© 2003 American Academy of Orthotists & Prosthetists