Co-Investigators: Rebecca Stine, MS
Project Director: Stefania Fatone, PhD
Funded by: National Institute on Disability and Rehabilitation Research (NIDRR)
The NU-RERC State of the Science Conference conducted in 2006 highlighted the need for the evaluation of functional outcomes that can be collected in the P&O clinic and combined from multiple facilities to create sufficient sample size for the resulting data to be meaningfully utilized to support clinical practice. The evaluation of functional outcomes provides information that empowers clinicians, patients, payers and policy makers, and facilitates patient care based on identifying the most effective interventions. Outcomes are an important component of evidence based practice and quality patient care.
Many clinical questions regarding function with P&O devices are spatially planar and could be addressed with 2D motion analysis. Although relatively simple, many such questions have only expert opinion to support them. For example, the 2003 ISPO Consensus Conference on the Orthotic Management of Stroke recommended that both nonarticulated AFOs and articulated AFOs with plantar flexion stops could be used to control knee recurvatum following stroke. However, these recommendations were only supported by low level evidence or expert opinion. Insufficient evidence exists for the functional outcome of many P&O interventions.
We propose that the functional outcome of P&O interventions could be determined in a sufficiently large sample to be statistically meaningful if a simple tool to measure motion could be deployed and used in multiple P&O facilities. Two-dimensional motion analysis has been used successfully in the past and, with improvements in digital camera technology and marker tracking software, can be used effectively today. The advantages of such systems are that the equipment required is relatively inexpensive, data collection protocols can be simple, and data can be centrally processed. As with any measurement tool, there are also limitations, the main one being that more complex questions regarding 3D movement cannot be easily addressed.
Our objectives for this project include (1) developing a simple two-dimensional (2D) motion analysis tool; (2) developing training materials to assist in deployment of the tool at clinical sites; (3) deploying the tool for use at several P&O clinical sites and calibrating the set-up; (4) conducting a pilot investigation of the effectiveness of ankle foot orthoses (AFO) at controlling knee recurvatum at mid stance in persons following stroke; and (5) evaluating the utility of the tool and application to the pilot study.
Results and Discussion
This project explored the feasibility of performing multicenter functional outcome studies using relatively low-cost video cameras and centralized data processing. This feasibility study is important for developing future research projects that involve making measurements of persons using prostheses and orthoses in a clinical setting, but this information could also be used in other fields of rehabilitation to address clinical questions regarding movement pathologies.
Our experience conducting a small multicenter clinical trial collecting videos in orthotic practices suggested that guidance was needed regarding the recording of quality videos for 2D motion analysis (Fatone & Stine, 2015). Conditions that affect video quality include:
- Illumination - a bright image is needed to track the joints or segments,
- Image size - camera should be positioned as close as possible to the image field while maintaining sufficient space to view the activity,
- Camera position and steadiness - the camera needs to be held steady, level, and placed perpendicular to the image field,
- Marker placement on the patient - correct marker placement on the joints of interest improves measurement of joint angles (Baker et al., 2013).
FIGURE:This figure illustrates two potential layouts for camera position to record gait: (A, C) through a doorway and (B) in a large room. Camera must be positioned parallel to the floor and perpendicular to the image plane (C). Markers placed on key anatomical landmarks such as greater trochanter, femoral epicondyle and lateral malleolus (D) can help with post-processing of joint angles. Be sure that arm swing does not obscure greater trochanter marker and consider de-identifying the image by capturing from the shoulders or waist down depending on data of interest.
A good quality clinical video can produce data that are comparable to that of instrumented motion analysis systems, especially for sagittal plane motion of large joints like the knee (Grunt et al., 2010). We processed pilot data from 3 able-bodied subjects in 2D using a single Casio video camera and MaxTraq software (Innovision Systems Inc., Columbiaville, MI, USA), again in 2D using an 8-camera digital motion analysis system and Cortex software (Motion Analysis Corporation, Santa Rosa, CA, USA), and in 3D using the same 8-camera system and OrthoTrak software (Motion Analysis Corporation) (Fatone & Stine 2015). For each of these techniques, reflective markers were placed on the subjects using a modified Helen Hayes model (Kadaba et al. 1990) with an extra marker on the greater trochanter for use in the 2D calculations of sagittal plane knee kinematics. Variations in knee angle between the 2D approaches were smaller than between the 2D and 3D approaches. This was likely due to 3D angle calculations using an embedded coordinate system to calculate Euler rotations rather than 2D included angles. Sagittal plane knee angles calculated from hand-digitized video using MaxTraq were comparable to 2D calculations using the computerized motion analysis system with automatic digitization of reflective markers (Fatone & Stine. 2015). Our results confirm the work of other authors who reported that 2D video-based systems can provide valid gait data assuming good quality videos are used (Churchill et al., 2002 and Nielsen and Daugaard, 2008).
Care must be taken during the setup process and during data capture to ensure that the resulting video is recorded in a manner that is appropriate with respect to patient privacy concerns.
Baker R, Harvey A, Rodda J. Clinical video. In: Baker R, (Ed.) Measuring Walking: A Handbook of Clinical Gait Analysis. London: Mac Keith Press; 2013.
Churchill AJ, Halligan PW, Wade DT. RIVCAM: a simple video-based kinematic analysis for clinical disorders of gait. Comput Methods Programs Biomed 2002;69:197-209.
Grunt S, van Kampen PJ, van der Krogt MM, et al. Reproducibility and validity of video screen measurements of gait in children with spastic cerebral palsy. Gait Posture 2010;31:489-494.
Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower extremity kinematics during level walking. J Orthop Res 1990;8:383-392.
Nielsen D, Daugaard M. Comparison of Angular Measurements by 2D and 3D Gait Analysis [dissertation]. Jonkoping: Orthopaedic Technology, Jonkoping University; 2008.
Fatone S, Stine R. (2015) Capturing Quality Clinical Videos for Two-Dimensional Motion Analysis. Journal of Prosthetics and Orthotics, (1)27-32.
Stine R and Fatone S (2017) Capturing Quality Clinical Videos for Two-Dimensional Motion Analysis. In the Symposium “Improving Patient Care with Smart Data.” The 43rd Academy Annual Meeting & Scientific Symposium of the American Academy of Orthotists & Prosthetists (AAOP), March 1-4, Chicago, IL.
Fatone S and Stine R (2016) Capturing Quality Clinical Videos for Two-Dimensional Motion Analysis, as part of the Gait Society and Lower Limb Society Forum Technological Advancements That Promote Practical Application of Outcome Measures, 42nd American Academy of Orthotists and Prosthetists Annual Meeting & Scientific Symposium, March 9-11, Orlando, FL.
Fatone S. (2012) 2D video based assessment of gait. Invited Speaker, 3rd Annual Convention in Prosthetics and Orthotics, July 6-8, Manila, Philippines.