**1. Introduction**

The interest in spinal‐ and postural‐related pathologies and the evaluation of their related functional impairment is widely represented in both biomechanical and clinical research

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2017 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

literature. The need for quantitative posture and spine shape analysis is recognized as crucial for clinical assessments in physical medicine and rehabilitation [1] and very important in designing and developing treatment programmes, planning of orthopaedic surgical proce‐ dures [2], and monitoring the progression of pathology and/or treatment outcomes [3, 4].

Posture, that is, the attitude in the space of the body whilst sitting, walking or standing, is a dynamic event, even in relation to the simple neutral‐standing‐erect position. In fact, even for the neutral erect standing, which is usually considered as a static posture, we know that, in reality, the body is continuously oscillating. So, the standing posture could be characterized by an 'equilibrium status' (i.e. the mean standing position) together with the intrinsic vari‐ ability in terms of oscillations around this status (i.e. the standard deviation associated to the mean). In an analogous way, also cyclical‐repetitive movements such as gait, in which the lower limbs move in an alternating cyclic way, can be described by an 'equilibrium status' (i.e. the mean gait cycle) together with the associated variability. It is very well‐known that posture (i.e. equilibrium status and associated variability) is strictly related to any given men‐ tal and/or physiological status (healthy, pathological, voluntarily maintained, fatigued, under physical and/or psychological stress etc.). Further, different factors can affect one's postural demeanour including familial physical aspects, anatomical structural impairments, postural habits and work activities.

In this way, from a neurophysiological point of view, by analysing this mean status and con‐ nected variability it is possible to derive important information about the functional status of a human body system as well as the related control mechanisms provided by the central nervous system (CNS) [5, 6].

The quantification of such functional evaluation in an unobtrusive and innocuous but com‐ plete way is a big challenge from an instrumental point of view.

This is particularly true when the analysis of the full‐skeletal posture, including the three‐ dimensional (3D) shape of the spine, is considered. In fact, although in the last decades, a real enhancement in the diagnostic technologies based on image processing (e.g. digital X‐ray, digital 3D stereo X‐ray reconstruction, computed axial tomography (CAT) scans and mag‐ netic resonance imaging (MRI)) has led to a significant improvement in accurate and detailed information, in the evaluation of skeletal anatomical structures and spine‐related pathologies. However, except for dynamic X‐ray and the very recent dynamic MRI, no single one of these techniques is able to provide information about the functional state of the vertebral column and related patient posture [7, 8].

Indeed, two‐dimensional (2D) X‐ray‐based images with their potentially harmful ioniz‐ ing effects and their 'single‐shot' nature are still commonly used in clinical examination. Moreover, they are not free from technical limitations such as the presence of image noise, distinctive characteristics of imaging techniques and the variable positioning of the patient during image acquisition, which represent a major source of variability and create the risk of evaluation errors that may conceal the actual geometrical relationship between anatomical structures [9].

Many efforts have been made in recent years to develop non‐invasive techniques to overcome such X‐ray‐based shortcomings. Unfortunately, many of these new approaches still embody certain limitations. Namely, they are either able to provide only partial measurements or alter‐ natively are unable to perform simultaneous 3D measurements throughout the whole spinal column. In some cases, they require that the subject maintain a specific and restricted postural demeanour, which significantly affects the outcomes, as occurs for both electro‐goniometric and/or flexicurve devices [1, 10, 11]. More recently, some interesting low‐cost photographic methods have appeared in the literature. However, even if these new methods present prom‐ ising results they still exhibit significant intrinsic limitations: the single‐shot approach, lack of genuinely instantaneous 3D posture measurement (the coronal and sagittal planes are not recorded simultaneously) together with weak calibration procedures, all of which limit their use to follow‐up monitoring [12–15].

Furthermore, even the more commonly available rastereography back‐surface measurement technique raises questions and doubts that require further clarification. Curiously, even if it has been introduced for the evaluation and follow‐up of scoliosis, particular concern is related to discrepancies found in spine shape with respect to X‐ray techniques in the coronal plane [10, 16].

Given the restrictions above, it has been demonstrated that, in this context, a technique, named opto‐electronic stereo‐photogrammetry, offers a significant solution for the cap‐ ture of functional information necessary for addressing clinical problems in rehabilitation medicine and is increasingly being reported in the literature for use in exploring different original approaches [1, 6–8, 17–30]. Basically, this approach is reliant on the possibility of obtaining 3D measurement of points in space using a number of calibrated TV cam‐ eras (at least two) using stereo‐vision principles. With this technique, the measurement is restricted to few specific body landmarks, labelled by retro‐reflective or active markers, neglecting other information such as the back surface of the trunk (as in rastereogra‐ phy) to lower the computational effort allowing very fast measurements (hundreds per second).

So, for a static‐erect posture both the 'equilibrium status' (averaged shape measurement) and the intrinsic variability (standard deviation around the mean) in term of oscillations around this status can be easily obtained. Additionally, the analysis can be expanded to quantitatively document the kinematics of the full skeleton during movement. Given that this method has no harmful effects, it provides a 'natural' approach for both the capturing and monitoring of the progression of pathology and/or treatment outcomes.

Among the various 3D opto‐electronic stereo‐photogrammetric original approaches pre‐ sented in the literature [6–8, 18–30], we focus here on a new recently proposed integrated stereo‐photogrammetric opto‐electronic system named GOALS1 (Global Opto‐electronic

literature. The need for quantitative posture and spine shape analysis is recognized as crucial for clinical assessments in physical medicine and rehabilitation [1] and very important in designing and developing treatment programmes, planning of orthopaedic surgical proce‐ dures [2], and monitoring the progression of pathology and/or treatment outcomes [3, 4].

Posture, that is, the attitude in the space of the body whilst sitting, walking or standing, is a dynamic event, even in relation to the simple neutral‐standing‐erect position. In fact, even for the neutral erect standing, which is usually considered as a static posture, we know that, in reality, the body is continuously oscillating. So, the standing posture could be characterized by an 'equilibrium status' (i.e. the mean standing position) together with the intrinsic vari‐ ability in terms of oscillations around this status (i.e. the standard deviation associated to the mean). In an analogous way, also cyclical‐repetitive movements such as gait, in which the lower limbs move in an alternating cyclic way, can be described by an 'equilibrium status' (i.e. the mean gait cycle) together with the associated variability. It is very well‐known that posture (i.e. equilibrium status and associated variability) is strictly related to any given men‐ tal and/or physiological status (healthy, pathological, voluntarily maintained, fatigued, under physical and/or psychological stress etc.). Further, different factors can affect one's postural demeanour including familial physical aspects, anatomical structural impairments, postural

In this way, from a neurophysiological point of view, by analysing this mean status and con‐ nected variability it is possible to derive important information about the functional status of a human body system as well as the related control mechanisms provided by the central

The quantification of such functional evaluation in an unobtrusive and innocuous but com‐

This is particularly true when the analysis of the full‐skeletal posture, including the three‐ dimensional (3D) shape of the spine, is considered. In fact, although in the last decades, a real enhancement in the diagnostic technologies based on image processing (e.g. digital X‐ray, digital 3D stereo X‐ray reconstruction, computed axial tomography (CAT) scans and mag‐ netic resonance imaging (MRI)) has led to a significant improvement in accurate and detailed information, in the evaluation of skeletal anatomical structures and spine‐related pathologies. However, except for dynamic X‐ray and the very recent dynamic MRI, no single one of these techniques is able to provide information about the functional state of the vertebral column

Indeed, two‐dimensional (2D) X‐ray‐based images with their potentially harmful ioniz‐ ing effects and their 'single‐shot' nature are still commonly used in clinical examination. Moreover, they are not free from technical limitations such as the presence of image noise, distinctive characteristics of imaging techniques and the variable positioning of the patient during image acquisition, which represent a major source of variability and create the risk of evaluation errors that may conceal the actual geometrical relationship between anatomical

plete way is a big challenge from an instrumental point of view.

habits and work activities.

20 Innovations in Spinal Deformities and Postural Disorders

nervous system (CNS) [5, 6].

and related patient posture [7, 8].

structures [9].

<sup>1</sup> Bioengineering & Biomedicine Company Srl Pescara, Italy.

Approach for Locomotion and Spine based on Optitrack2 hardware) fully founded on the protocol and procedural techniques formerly presented by D'Amico et al. [7] together with the subsequent development and upgrading of the system [6–8, 17, 27–30].

The actual implementation of the GOALS system and related biomechanical skeleton model allows for the analysis of the full human skeleton 3D posture and movement taking into account the 3D spine shape considering each vertebral level as well as the postural attitude of the head, the trunk, the pelvis, the legs and when necessary the upper limbs. It is able to per‐ form a multi‐sensor approach, fully integrating data deriving from force platforms, surface electro‐myography (SEMG) and foot pressure maps. In addition to kinematic measurements, depending on the specific analysis requirements to be fulfilled, this can also include the mea‐ surement of the forces, torques and electro‐muscular activity of participants or patients. By means of data fusion and optimization procedures, all these inputs can be used in the skeleton model to assess internal joint forces, torques and muscular effort. This allows for the correla‐ tion of the full‐functional evaluation of subjects with their morphological characteristics.

The possibility of assessing and extracting mean behaviours for cyclic or repetitive tasks (such as multiple strides in gait) has been included as well [8, 27, 28, 31].

The applications and the valued contribution in clinical‐functional diagnoses of differ‐ ent classes of posture, locomotion and spine‐related pathologies using the GOALS system together with its original biomechanical approach have been presented in literature [6–8, 27–31]. Thousands of patients are currently analysed and followed up with this methodology.

The highly sophisticated and demanding computing tasks to acquire data and to solve the whole skeleton model's equations and algorithms can be approached even on relatively low‐ cost powerful PC workstations. Examples of multi‐sensor quantitative functional descriptions of pathological cases are presented to describe the actual level of development of the GOALS system/skeleton model and the actual capability to use them as a clinical tool.
