Construction of a Reference Corridor to Assess the Value of Combining Innovative Morphological and Functional Analysis Techniques in Pediatric Patient Care Based on Different Pathological Models

Overview

Reference motion analysis systems require a dedicated, specific laboratory environment, with self-reflective markers placed on anatomical points by trained personnel, limiting gait analysis to specific pathologies. Some markerless systems have emerged but still require numerous cameras and are expensive. The implementation of a markerless motion capture system based on only four cameras in routine clinical practice would broaden the indications for gait analysis to include any pathology of the lower limbs and/or spine in the pediatric population.

Deformities of the lower limbs (LL) and spine have a severe functional impact on quality of life, whether they are present at birth (malformation) or develop with the acquisition of walking (deformity). These abnormalities result from de novo or hereditary genetic mutations, and their cause often remains unknown. These deformities do not only affect the skeletal system. They are abnormalities of the entire musculoskeletal system, associated with muscle damage and changes in the ligament and capsular systems. All of these conditions result in functional walking disorders that impair patients' quality of life. The development of innovative techniques in recent years has enabled accurate and safe morphological (EOS stereoradiography and muscle MRI in different sequences) and functional (quantified gait analysis, QGA) analyses, allowing for better diagnostic and therapeutic management of pediatric patients. Radiological examinations (EOS, MRI) are readily available in routine clinical practice. Gait analysis still requires a specialized laboratory and a dedicated team to place markers and analyze data, which limits its use despite its proven usefulness in various pathologies. However, it is an indispensable tool as it provides an objective, quantified measurement of gait function. It has long been demonstrated that this tool is invaluable in aiding diagnosis, therapeutic decisions, and monitoring the effectiveness of treatments undertaken. In recent years, new markerless motion capture systems have been developed with the aim of simplifying access to this tool in clinical practice. One such motion capture system was developed by teams of engineers at the ENSAM biomechanics laboratory and was tested and validated on a population of 31 volunteers. This system obtained positive results in terms of gait joint position error when compared to standard three-dimensional AQM. With regard to gait parameters, it has been validated as sufficiently accurate for measuring the spatiotemporal parameters of gait in clinical applications, but still requires adjustments to the kinematic parameters of gait for clinical application. Quantified gait analysis (QGA) enables three-dimensional analysis of movement using motion capture systems (kinematic analysis) and ground reaction force analysis (kinetic analysis). These systems consist of infrared cameras that capture retro-reflective markers positioned on anatomical landmarks of the human body for kinematic analysis. For kinetic analysis, the forces exerted (vectors, moments, and powers) on the joints of the trunk, hip, knee, and ankle are measured using force platforms. This 3D joint movement data aids in the functional diagnosis of patients with many different pathologies and enables a selective therapeutic approach. This "classic" three-dimensional AQM is performed in dedicated, secure laboratories, requiring the presence of an engineer and a medical team to acquire, process, and interpret the data from these analyses, which limits its use due to the time required to interpret the data. Markerless motion capture systems with fewer cameras have been developed in recent years. Such a system has been developed at ENSAM's LBM and is based on "human pose estimation," a computer vision method. Using four RGB (red, green, blue) cameras to calculate walking parameters and artificial intelligence techniques, this system determines the position of the joint centers. It has been tested and validated on a sample of 31 subjects (adults and children) at the Georges Charpak Institute of Human Biomechanics. This system has been validated for the use of spatiotemporal parameters in clinical applications but still requires adjustments to the kinematic parameters. The validation of this markerless AQM for various pathologies, such as hereditary and/or congenital abnormalities of the lower limbs and/or spine, followed by its implementation in routine clinical practice, would represent a major advance in the understanding of different musculoskeletal models and in the diagnostic, therapeutic, and follow-up management of these pathologies. The existence of pediatric reference corridors is uncommon, mainly due to the changes that occur throughout growth. A comparison between a sick child and a healthy adult is therefore not possible, due to the specific properties of growth. Furthermore, within pediatrics itself, a child who is still learning to walk cannot be compared to an athletic teenager. Obtaining reference standards for this markerless motion capture system based on healthy children would allow for subsequent comparison with any pathology of the lower limbs and/or spine of children undergoing this examination in routine clinical practice.