Secondary Logo

Journal Logo

Original Research Article

Determination of Dynamic Prosthetic Alignment Using Forceline Visualization

Talaty, Mukul PhD; Esquenazi, Alberto MD

Author Information
JPO Journal of Prosthetics and Orthotics: January 2013 - Volume 25 - Issue 1 - p 15-21
doi: 10.1097/JPO.0b013e31827afc29
  • Free


Objective, science-based prosthetic alignment is an essential ingredient missing in the care of patients with lower limb amputation for uniform achievement of optimal functional mobility outcomes.1,2 The term alignment in this article is used to refer to the relationship, during walking, of the center of mass, subject anatomy, and the orientation of the prosthetic components that comprise the prosthesis. How components are oriented can have a profound effect on the quality of the performance of the user.1,3,4 Specifically, walking difficulty, pain, skin irritation and breakdown,1 safety, energy expenditure,5 maximum walking speed and distance, and others, can all be affected by the alignment. There is no universally correct alignment because individual factors such as specific componentry involved, anatomical joint range of motion, strength, residual limb length, and ability to tolerate pressure can all affect alignment choice.1,3 Furthermore, it is appropriate to note the distinction between static and dynamic alignments. Static prosthetic alignment refers to that performed during quiet standing posture. Some objective recommendations have been made for this.6–8 Although there is certainly some relationship to how individuals stand and how they walk, it is not sufficient to use solely a static alignment as the basis for optimizing walking performance.6 The unique strength, coordination, sensation, and other demands that contribute to walking are intrinsically considered during a dynamic alignment. Dynamic alignment is the focus of this article.

Currently, there is great variability in prosthetic alignment outcome level9: some clinicians and gait laboratories are able to achieve excellent outcomes, whereas others routinely come up short in achieving this standard. The principles of optimizing dynamic alignment are not codified or part of standard instruction for clinicians who are often responsible for providing this care.10 Consequently, there are no universally accepted and used quantitative approaches to achieving this ideal; however, some techniques recently introduced have attempted this.

ComPAS® (Computer Prosthetic Alignment System) is a fairly recently developed unique system that uses a computer algorithm to provide real-time feedback to the clinician about specific changes to achieve optimal alignment.11 To do so, forces across the socket are measured using a special instrumented alignment pyramid. These forces are then wirelessly relayed to a computer and compared against a database of alignments to suggest what alignment changes are needed to achieve the “targeted” alignment. The objectivity and repeatability of this system seem to be quite high in the hands of an experienced user. Furthermore, fairly well-established statistical techniques and a wide variety and amount of input data seem to have been used in developing the algorithm that functions as the “brain” of the system. However, in our opinion, it seems that the clinician still needs to know how to follow the recommendations made to address the deviations demonstrated by the algorithm. Feedback we have heard from individuals who have either used ComPAS or otherwise familiar with ComPAS seems mixed, with some arguing that the system takes away from the clinical intuition and provides a rigid guidance. More broad, peer-reviewed scientific evaluation of the system was unavailable at the time of this writing. Nonetheless, this system generally seems to be a step in the direction of adding objectivity and repeatability to the dynamic lower limb prosthetic alignment process. Other systems, like the Otto Bock static forceline, are limited to the static alignment and cannot integrate the dynamic contributions of walking to the process.

Gait analysis can be considered a set of methods and tools to help the clinician in more objectively evaluating walking and optimizing walking function. To that end, it should and does have a role in achieving good prosthetic alignment and, subsequently, walking outcomes.12 Some of the relevant tools include video, motion, and force analyses. Video at 30 fps permits slowing down of rapid motions during walking and a means to qualitatively compare multiple views in multiple conditions (e.g., alignment settings). Three-dimensional motion (3D) capture allows more accurate and objective assessment of the walking movement, but at the expense of having to instrument the subject and process data. In addition, there are nontrivial system and skilled operator costs. Force platforms allow objective measurement of forces under the foot during walking and can be used to assess limb loading and center of pressure position, compare lateral asymmetry, and compare with normal gait data. In addition, when force data are combined with 3D motion capture data and biomechanical models, an estimate of net joint loading can also be obtained. Two major limitations of gait analysis are that it is resource intensive and still cannot guarantee results. In the end, the knowledge and experience of the clinician are still essential to properly interpret and use the data generated to reach a successful clinical outcome.

The dynamic forceline visualization (DFV) system is a simple approach that combines and enhances aspects of the force plate data and video analyses without the need for subject instrumentation. In this approach, an image of the net line of force under each foot is superimposed on a video image of the subject as he/she walks and displayed on a screen for the clinician to assess in real time. An image of this from our laboratory is shown in Figure 1. Such a technology has been around for several decades and used in special contexts—prosthetic and orthotic “alignment” being one.13 A similar line has been used for static or bench alignment purposes.5,6 The primary benefit of these forceline approaches is real-time feedback of effect of a change in the device alignment coupled with the ability to allow the patient to walk without any encumbrances from instrumentation. This approach can potentially reduce the iterative nature1 of sending patients home with a trial alignment, having them try it out for a while and return for further adjustment. Although there is a monetary cost with a forceline visualization system, it is an order of magnitude less than that of a fully outfitted gait laboratory. In addition, the marked reduction in the space needed to use such a tool will likely make this approach a viable option for many facilities and clinicians who provide prosthetic alignment services. Although forceline visualization technology has existed and been used by some for many years, a concise methodology for using it has yet to be developed. The objective of our work was to suggest a methodology and to demonstrate that such an approach can be easily learned and applied by novice clinicians to produce a safe, functional, and reproducible alignment.

Figure 1:
Sample forceline image. The dark straps and jigs shown on the legs and pelvis are for the motion capture system and are unrelated to the forceline imaging.


Three clinician trainees were taught the basic biomechanical principles of gait kinetics in the context of the ground reaction force location relative to the lower limb joints. Then, using experienced prosthesis users as test subjects, the novice clinicians were given deliberately misaligned prostheses to identify and correct. For each case, the clinicians worked initially as a group but made final decisions on their own with help of the forceline system only. In the final decision process, the clinicians could make any alignment changes they felt appropriate and could iteratively evaluate the amputee’s performance in that condition. The only tools they could use were observation, subject feedback, and video footage that included the forceline. The goal was to achieve three alignment targets as taught to them in the training session.


The essence of the alignment targets was derived from accepted normal gait data14–16 that suggest the following polarities of joint kinetics (using the internal joint moment convention): 1) In early stance (just before contralateral toe off): positioning the line to be anterior to the hip, posterior to the knee, and posterior to the ankle would indicate hip extension, knee extension, and ankle dorsiflexion moments. 2) In midstance: a line passing close to or through the hip and knee and anterior to the ankle would indicate no moment at the hip or knee and a plantarflexor moment at the ankle. 3) In terminal stance (one to two frames just before double support): a line passing posterior to the hip, posterior to the knee, and relatively far anterior to the ankle would indicate hip flexion, knee extension, and a large ankle plantarflexor moment. These joint kinetics are roughly in line with published and well-accepted phases of normal gait kinetics for comfortable, self-selected walking speeds.15 The “internal joint moment convention” refers to the moments being calculated with the polarity of the net amount generated by the muscles surrounding each joint (Figure 2).

Figure 2:
Illustration of a time in the gait cycle when the forceline indicates internal moments with hip extensor, neutral at the knee, and ankle plantarflexor polarities. The yellow lines from each joint center to the forceline clarify the polarity of the existing moment at each joint.

For target 1, the timing was critical. Going into single support with the forceline behind the knee joint indicates that knee extensor muscle action is present to maintain stance phase stability, and this may also increase forces at the residual limb socket interface.

For target 2, an abnormal location of the line behind the knee reflects limb instability, whereas a line anterior to the knee joint often indicates stability (delayed unloading and transition to the swing phase).

For Target 3, the timing is less critical. The force is not likely to produce increase joint instability but facilitates limb unloading and initiation of the swing phase.

To recap, these “targets” essentially correspond to the following polarity joint moments at the hip, knee, and ankle (respectively) during the gait cycle: extensor, extensor, and dorsiflexor during early stance (double support); neutral, neutral, and plantarflexor during midstance; and flexor, extensor, and plantarflexor during terminal stance.


Two prosthetists in their internship and one young physical therapist with 2 years of clinical experience participated in this study. Four experienced prosthesis users with mature residual limbs (at least 1 year and able to walk without assistive devices) were the amputee test subjects (Table 1). Three had unilateral transtibial amputations; one had a transfemoral amputation. All had used their current prosthesis for several months without complaints. In addition, one skilled clinician experienced in the use of forceline visualization for prosthetic alignment and one engineer experienced in the use of forceline visualization performed the training and determination of laboratory “nominal” alignments, made all requested alignment changes suggested by training clinicians, and recorded all interventions tried.

Table 1:
Amputee subject characteristics


Once instructed on the basic “targets” for alignment, we allowed three dimensions of alignment along which modifications could be made to achieve the desired force line positioning. The three dimensions of alignment used in this study were foot dorsiflexion, knee flexion, and foot rotation (toe in/out). Our experience suggests the following relationships:

  • Foot dorsiflexion moves the forceline posterior to the knee.
  • Knee flexion moves the force line posteriorly.
  • Toe out moves the force line posteriorly, but only if the foot was already toed out, not toed in, or in neutral because it decreases the toe lever. This intervention would have an effect only in midstance and late stance.

We chose this limited set of alignment degrees of freedom because they are commonly used by clinicians; changing them has clear and relatively straightforward ramifications to prosthetic gait kinetics, and thus, they are relatively simple to convey and learn, even for those who may not be intimately familiar with gait kinetics. Clinician subjects were limited to making changes to these three dimensions. Furthermore, we limited them to make changes in increments of at least 2.5° (one screw turn) or integer multiples of this.


Each amputee subject was initially evaluated in the gait laboratory by the investigator team to determine the nominal alignment that satisfied the alignment targets that were taught for the study. This configuration was recorded. The prosthesis was then deliberately misaligned along some combination of the three selected alignment dimensions. Next, the study clinicians evaluated the amputee subject in the deliberately misaligned prosthesis and determined what changes to make. First, they did so as a group but without help of the line; then they made final adjustments individually but using forceline visualization. The final alignment of each novice clinician was recorded and subsequently compared with the laboratory nominal alignment for each of the four amputee subjects. Not all three dimensions were changed for an amputee subjects’ prosthetic alignment in generating a misalignment condition. Overall, eight distinct changes were made for the four amputee subjects (i.e., on average, two dimensions were changed per amputee). Thus, the overall data set consisted of 3 clinicians × 4 subjects × 2 dimensions of alignment changes = 24 distinct alignment comparisons. As an ancillary measurement, temporospatial footfall parameters via the electronic gaitmat17 was recorded for each subject in the final alignment condition for each clinician as well and compared with the nominal alignment outcome.


The primary outcome measure was whether the clinician-selected alignments were closer to the nominal alignment when the forceline visualization system was used or not used. To this end, we compared the actual orientation of the prosthesis along the three alignment dimensions with that of the laboratory nominal alignment. The number of degrees the foot was toed in/out, the socket was flexed/extended, and the foot was dorsiflexed/plantarflexed was tracked using a combination of turns of set screws and a specially designed goniometer. The difference between the clinician subject alignments obtained both with and without the use of the line was then compared with the nominal alignment to assess which was closer.

We also assessed whether amputee walking was improved in the forceline-based alignment. A set of temporospatial measures were obtained from the electronic GaitMat™ II (E.Q., Inc, Chalfont, PA) system to assess how, if at all, the amputees walked differently in the alignments obtained both with and without the forceline. Gait velocity and symmetry ratios were calculated of the following measures, as a ratio of the left and right sides for each amputee: step length, step time, swing time, stance time, and double support time. If a ratio got closer to 1 (perfect symmetry) when the forceline was used, it was considered “improved.” Two thresholds (5% and 2%) for what was a significant difference in a change in the symmetry ratio were assessed to evaluate the impact of the arbitrary threshold on the outcome. In addition, an increase of greater than 10% in velocity was also considered an improved outcome. The scores were tallied (five variables for each of three clinicians) and reported.


A comparison of the clinician subjects’ performance both with and without the use of the forceline visualization technique is presented in Table 2. Half of the changes (11/22) that needed to be made to restore the nominal alignment from the misaligned condition were influenced by the use of the forceline. Eight of those 11 were improved by the use of the forceline. Three seemed to have become slightly worse when the forceline was used to adjust the prosthetic alignment. The remaining 11 were not changed by the use of the forceline.

Table 2:
Approximation of clinician subjects’ alignments to the “nominal” alignment with and without use of the forceline visualization system

The effect of forceline use during alignment on temporospatial symmetry and walking velocity is shown in Table 3. The ratings “improved” and “worsened” refer to whether the use of the forceline made the temporospatial parameters more or less symmetric and whether velocity increased or decreased, respectively.

Table 3:
Impact of forceline visualization use during alignment on temporospatial symmetry and walking velocity


The DFV technique was easily learned and incorporated into alignment performance by all three novice clinicians. After a single training session of approximately 90 minutes that consisted of viewing images and listening to an experienced clinician describing the alignment protocol, the novice clinicians were able to internalize and independently apply the protocol to determine alignments for the four new cases they were presented. The novice clinicians were relatively new to the field and in part selected for this reason. To be considered for participation in the study, they needed to have a basic understanding of gait and gait kinetics as well as some experience with lower limb prostheses. Limited experience would allow them the fundamental knowledge to grasp the forceline-based alignment concepts. At the same time, their relative inexperience in administering alignments would allow them to apply the learned techniques without too much reliance on other methods they may have already developed. We assume that a clinician with any level of experience would be able to use these forceline visualization–based techniques, given an earnest desire to use them as a guidance while avoiding the use of other “favorite” techniques. Certainly, it is possible that a better method may indeed result from combining individual clinical knowledge and techniques acquired by the experienced user with the DFV method than strictly applying just the proposed methods. There is yet to be consensus on what exactly constitutes optimal prosthetic alignment or how to achieve it. We suggest that this method may facilitate achievement of at least the definition of optimal and subsequently perhaps an algorithm by which to achieve it by making use of the clinician’s skill and the guidance provided by real-time visualization of the interaction of alignment modifications and kinetics.

This work has demonstrated that an empirical approach using the forceline visualization technique can be easily learned and applied by novice clinicians to produce a safe, functional, and reproducible prosthetic alignment. The qualifiers of “safe” and “functional” of the resulting alignment are inferred from the decades of use in our laboratory with a high level of patient and clinician satisfaction with functional outcome such as increased walking velocity and improvement in the symmetry of the temporal spatial parameters and reported increase in comfort, without the need for repeated alignment or/and with no reported falls. We are not proposing that the alignment technique discussed is optimal or superior but rather one that is reproducible and can lead to good outcomes regardless of the experience level of the clinician or the need for other tools or more complex equipment. For just the sagittal forceline, only a single Web camera, standard laptop, forceplate, and visualization software are needed.

The alignment technique is one that has evolved over the years at our laboratory and is based on relatively simple and accepted biomechanical principles of gait. Clinicians who may use the proposed methodology still need to be familiar with prosthetic alignment and basic gait biomechanics and learn to use the DFV technique as a guide for alignment. We want to acknowledge that the alignment technique we used in the study is not a comprehensive approach, but rather a simplified version of a more complex routine used at our center. We circumscribed a set of guiding principles that would be sufficient to address the relatively simple cases of (mis)alignment that were used in the study. This set of principles forms the basis of extended alignment techniques for more complex scenarios that clinicians will encounter in clinical practice. Some of these “extended techniques” can be extrapolated from the simplified set described in this article. The simplified set had a very practical motivation as well—to allow for a relatively quick learning curve and reasonable data collection times.

Using the DFV system resulted in alignments closer to the desired (the nominal based on our laboratory standard) alignment 36% of the time (8/22 cases) and further away from the desired only about 14% of the time (3/22 cases). The net effect was to standardize the alignments, that is, bring them closer to a desired endpoint. This endpoint is one that was determined by an experienced clinician and has been used to provide alignment for hundreds of amputees over a span of more than 20 years. Most of these patients exhibit a consistently high level of function and subsequently have reported satisfaction with their prosthetic performance with regard to functional outcome (increased walking velocity, improved symmetry of the temporal spatial parameters, and reported increase in comfort and no falls). This report is anecdotal, however, and has not been quantified.

For test subject AT5, the toe in/out in the misaligned condition was 10° toed in, whereas the nominal was 10° toed out. Thus, they were symmetric about zero and likely had equivalent functional consequences. This may explain why the clinicians likely did not make any significant adjustment in this degree of freedom yet still obtained reasonable outcomes with respect to the forceline. In general, the toe in/out dimension of alignment was least affected by the use of the forceline. Only two of the possible nine evaluations relating to this dimension were affected, with one improving slightly and one worsening slightly.

The forceline did not seem to have a consistent positive (or negative) effect on walking performance as gauged by temporospatial symmetry and velocity measures (Table 3). Overall, the changes made by the clinicians with the help of the forceline were relatively minor (in most cases, less than 5°). This level of change can have a measurable and significant impact on amputee walking performance.4,8,18–20 The alignment changes in our study may have been implemented for too short a time to allow stabilization in response to be observed or that our experienced study subjects were able to overcome the minor alignment differences with their own internal compensations. The latter is an unavoidable confound when using experienced prostheses users; the former is a difficult-to-avoid constraint given our study design. We chose to have all clinicians evaluate the same test subject consecutively to minimize any performance inconsistencies across that particular subject that may have arisen from testing a single individual across several days. This, however, made for a relatively long testing session (about 4 hrs). To minimize fatigue biases, we aimed to expedite the testing procedure wherever possible. There may well have been differences in other measures, such as joint kinematics or kinetics, but these data have not yet been analyzed. In addition, energy expenditure, which is known to reflect alignment changes, was not collected.

Although there is no consensus on what optimal alignment is or how to reach it, we believe that kinetics will be an integral part of that definition—whenever it is developed. Kinetics reflect the net effect of how our muscles, joints, and prostheses function to counter gravity and provide propulsion—two key aspects of walking.14,21,22 To that end, whatever the definition of optimal is, it should be generally amenable to being codified into a method that the forceline visualization technique can assist. The forceline allows the visualization of the ground reaction force in real time and in a spatially accurate manner relative to the image of the person as he/she walks. We stress that we are not suggesting that the alignment technique we presented can supplant clinical thinking or decision making. Rather, quite the contrary, we envision this approach to assist by removing some of the guesswork and uncertainty involved. Providing accurate and quick feedback can only help improve the understanding of how changes to prosthetic alignment influences kinetics and, ultimately, function and comfort. But this can work only in the hands of clinicians who understand basic biomechanics and who can incorporate the feedback given both by the forceline visualization and the patient into a cogent and rationale theory. Finally, these clinicians will need to have sufficient experience and/or creativity to modify the prosthesis to test that theory, incorporate that feedback, and still be willing to accept the possibility of a revision when the patient returns in the future.


The demonstrated forceline visualization–based methodology for assisting lower limb dynamic prosthetic alignment was easily learned and applied by novice clinicians. It helped to standardize the alignment procedure across the clinicians and brought the final alignments closer to what is desired about 36% of the time. By providing real-time visual feedback on how gait kinetics are altered by subtle changes to prosthetic alignment, such a visual tool can reduce guesswork in the effect of alignment. Coupled with relatively simple usage guidelines (as presented), the suggested tools and methodology can help to develop a widely accepted definition of an optimal prosthetic alignment and can also help in determining that alignment for each patient.


We thank the clinician and amputee subjects who participated in the study; Barbara Hirai and Oliver Woods Jr, for help with data collection; and Sayed Naseel Mohamed Thangal, for help with the literature review. Last, but not least, we thank Bertec Corporation for working with us and providing the opportunity and resources for us to carry out this study.


1. Zahedi MS, Spence WD, Solominidis SE, et al.. Alignment of lower-limb prostheses. J Rehabil Res Dev 1986; 23: 2–19.
2. Mizrahi J, Susak Z, Seliktar R, et al.. Alignment procedure for the optimal fitting of lower limb prostheses. J Biomed Eng 1986; 8: 229–234.
3. Fridman A, Ona I, Isakov E. The influence of prosthetic foot alignment on trans-tibial amputee gait. Prosthet Orthot Int 2003; 27: 17–22.
4. Yang L, Solomonidis S, Spence W, et al.. The influence of limb alignment on the gait of above-knee amputees. J Biomech 1991; 24: 981–997.
5. Schmalz T, Siegmar B, Jarasch R. Energy expenditure and biomechanical characteristics of lower limb amputee gait: the influence of prosthetic alignment and different prosthetic components. Gait Posture 2002; 16: 255–263.
6. Blumentritt S. A new biomechanical method for determination of static prosthetic alignment. Prosthet Orthot Int 1997; 21: 107–113.
7. Blumentritt S, Schmalz T, Jarasch R. Significance of static prosthesis alignment for standing and walking of patients with lower limb amputation. Der Orthopade 2001; 30: 161.
8. Blumentritt S, Sghmalz T, Jarasch R, et al.. Effects of sagittal plane prosthetic alignment on standing trans-tibial amputee knee loads. Prosthet Orthot Int 1999; 23: 231–238.
9. Geil MD. Variability among practitioners in dynamic observational alignment of a transfemoral prosthesis. J Prosthet Orthot 2002; 14: 159–164.
10. Van Der Linde H, Geertzen JHB, Hofstad CJ, et al.. Prosthetic prescription in the Netherlands: an interview with clinical experts. Prosthet Orthot Int 2004; 28: 98–104.
11. Boone DA. Investigation of Socket Reactions From Transtibial Prosthesis Malalignment [dissertation]. Hung Hom, Kowloon, Hong Kong: Department of Health Technology and Informatics, The Hong Kong Polytechnic University; 2006.
12. Van Velzen J, Houdijk H, Polomski W, et al.. Usability of gait analysis in the alignment of trans-tibial prostheses: a clinical study. Prosthet Orthot Int 2005; 29: 255.
13. Cook TM, Cozzens BA, Kenosian H. A Technique for Force-Line Visualization. Philadelphia, PA: Rehabilitation Engineering Center 2, Moss Rehabilitation Hospital; 1979.
14. Perry J. Gait Analysis: Normal and Pathological Function. Thorofare, NJ: Slack Incorporated; 1992.
15. Winter DA. Biomechanics and Motor Control of Human Movement. 2nd ed. New York, NY: Wiley Interscience Publication; 1990.
16. Inman VT, Ralston HJ, Frank T. Human Walking. 2nd ed. Baltimore, MD: Williams & Wilkins; 1994.
17. Taylor D. An Instrumented Gait Mat. Paper presented at The International Conference on Rehabilitation Engineering; 1980; Toronto, Canada, June 16–20, 1980.
18. Beyaert C, Grumillier C, Martinet N, et al.. Compensatory mechanism involving the knee joint of the intact limb during gait in unilateral below-knee amputees. Gait Posture 2008; 28: 278–284.
19. Chow DHK, Holmes AD, Lee CKL, et al.. The effect of prosthesis alignment on the symmetry of gait in subjects with unilateral transtibial amputation. Prosthet Orthot Int 2006; 30: 114.
20. Pinzur MS, Cox W, James Kaiser C, et al.. The effect of prosthetic alignment on relative limb loading in persons with trans-tibial amputation: a preliminary report. Development 1995; 32: 373–378.
21. Inman V, Ralston H, Todd F. Human Walking. Baltimore, MD: Williams & Wilkins; 1981.
22. Kepple TM, Siegel KL, Stanhope SJ, et al.. Relative contributions of the lower extremity joint moments to forward progression and support during gait he influence of limb alignment on the gait of above-knee amputees. Gait Posture 1997; 6: 1–8.

lower limb prosthetics; alignment; forceline visualization; gait

© 2013 American Academy of Orthotists & Prosthetists