Stereophotogrammetry in human movement analysis: novel methods for the quality assurance, biomechanical analysis and clinical interpretation of gait analysis

The study of movement has always fascinated artists, photographers and researchers. Across the years, several attempts to capture, freeze, study and reproduce motion were made. Nowadays, motion capture plays an important role within many fields, from graphical animation, filmmaking, virtual reality, till medicine. In fact, movement analysis allows to measure kinematic and kinetic performance of the human body. The quantitative data obtained from measurements may support the diagnosis and treatment of many pathologies, allowing to take clinical decisions and supporting the follow-up of treatments or rehabilitation. This approach is nowadays named evidence based medicine. In this work, motion capture techniques and advanced signal processing techniques were exploited in order to: (i) develop a protocol for the validation and quality assurance of the clinical strength measurements, (ii) develop an algorithm for clinical gait analysis data interpretation and identification of pathological patterns, and (iii) design user-friendly software tools to help clinicians using the novel data processing algorithms and reporting the results of measurements. This work was divided into three sections: Part 1 contains a survey about the history of motion analysis and a review of the earliest experiments in biomechanics. The review covered the first historical attempts, that were mainly based on photography, till the state-of-the-art technology used today, i.e. the optoelectronic system. The working principle of optoelectronic system was reviewed as well as its applications and modern setups in the clinical practice. Some modern functional evaluation protocols, aimed to the quantitative evaluation of physical performance and clinical diagnosis of motor disorders, were also reviewed. Special attention was paid to the most common motion analysis exam that is nowadays worldwide standardized, i.e. the Gait Analysis. Examples of Gait Analysis studies on subjects with pathology and follow-up were reviewed. Part 2 concerns the design of an experimental setup, involving motion analysis, for the quality assurance of clinical strength measurements. Measurements of force are popular in the clinical practice as they allow to evaluate the muscle weakness, health status of patients and the effects of therapies. A variety of protocols was proposed to conduct such measurements, implying the acquisition of forces, angles and angular velocities when the maximum voluntary force is exerted. Hand held dynamometry (HHD), based on single component load cell, was extensively used in clinical practice; however, several shortcomings were identified. The most relevant were related to the operator’s ability. This work was aimed to investigate the inherent inaccuracy sources in knee strength measurements when are conducted by a single component load cell. The analysis was conducted by gathering the outputs of a compact six-component load cell, comparable in dimension and mass to clinical HHDs, and an optoelectronic system. Quality of measurements was investigated in terms of quantifying, by an ad-hoc metrics, the effects induced in the overall inaccuracy by: (i) the operator’s ability to place and to hold still the HHD and (ii) ignoring the transversal components of the force exchanged between the patient and the experimenter. The main finding was that the use of a single component HHD induced an overall inaccuracy of 5% in the strength measurements, when operated by a trained clinician and angular misplacements are kept within the values found in this work (≤15°) and with a knee ROM ≤ 22°. Even if the measurement outputs were reliable and accurate enough for both knee flexion and extension, extension trials were the most critical due to the higher force exerted, i.e. 249.4±27.3 N vs. 146.4±23.9 N of knee flexion. The most relevant source of inaccuracy was identified in the angular displacement of HHD on the horizontal plane. A dedicated software, with graphical user interface, was designed and implemented. The purposes of this software were to: (i) speed up data processing, (ii) allow user to select the proper processing workflow, and (iii) provide clinicians with a tool for quick data processing and reporting. Part 3 concerns the research study about gait analysis on subjects with pathology. Gait analysis is often used for the assessment of the gait abilities in children with cerebral palsy and to quantify improvements/variations after a treatment. To simplify GA interpretation and to quantify deviation from normality, some synthetic descriptors were developed in literature, such as the Movement Analysis Profile (MAP) and the Linear Fit Method (LFM). The aims of this work were: (i) to use synthetic descriptors in order to quantify gait variations in subjects with Cerebral Palsy that underwent surgery involving bone repositioning and muscle/tendon lengthening at the level of the femur and hamstring group (SEMLS); (ii) test the effectiveness of a recently proposed index, i.e. the LFM, on such patients; (iii) design and implement a novel index that may overcome the limitations of the previous methods. Gait Analysis exams of 10 children with Cerebral Palsy, pre and post treatment, were collected. Data were analysed by means of MAP and LFM indices. To overcome the limitations observed for the methods, another index was designed as a modified version of the MAP, namely the OC-MAP. It took into account the effect on deviation due to offset and allowed to compute the deviation from normality on tracks purified by the offset. An overall improvement of the gait pattern was observed for most of the subjects after surgery. The highest effect was observed for the knee flexion/extension angle. Patients who had initial high deviations also had the largest improvements. Worsening in the kinematics of the pelvis could be explained as a consequence of SEML involving a lengthening of hamstring group. Pre-post differences were higher than the Minimally Clinical Important Difference for all parameters, except hip flexion. An improvement towards normality was observed for all the parameters, with exception of pelvic tilt for which a worsening was observed. LFM provided results similar to OC-MAP offset analysis but could not be considered reliable due to intrinsic limitations. As offset in gait features played an important role in gait deviation, OC-MAP synthetic analysis is recommended to study gait pattern of subjects with Cerebral Palsy. A dedicated software, with graphical user interface, was designed and implemented. The purpose of this software was to compute the synthetic descriptors on a large amount of data, to speedup data processing and to provide clinicians with a quick access to the results