*4.2.2.3.3 Laser triangulators*

Čelan et al. [78] and Poredoš et al. [79] used the laser triangulation method to evaluate the three-dimensional human spine curvature. The main purpose of these studies was to estimate the spatial bend of the thoracic and lumbar spine curvatures in all three planes. The laser triangulation imaging system used in Poredoš et al.'s study consisted of two basic elements: a greyscale camera (A) and a laser line projector (B) (see **Figure 12**). The spinal path or region of interest (ROI) of the human model is manually marked by the palpation of the subject's bony landmarks. The laser projector illuminates the light onto the subject's back, and the intersection of the laser line with the spinal path or ROI provides the intersection curve, which is then measured using a greyscale camera. The distance between the laser projector and the camera is known. The intersection angle in 3D space is calculated using the triangular method [80].

The laser scanning triangulation method was assessed for both validity and repeatability. Using a point-to-point analysis, the average error (±1 mm S.D) (distance between markers) for a regular shape (cylinder) was as low as 4.99 ± 1.56 mm, versus 6.91 ± 2.29 mm for an irregular shape (mannequin) [81]. Research by Majid et al. [82] demonstrated the performance of the 3D laser scanning system. In this laboratory-based study, craniofacial measurements of mannequins demonstrated that the photogrammetric/3D laser scanning system had an accuracy of ±0.7 mm (1 standard deviation [SD]).

The same measurement in human models demonstrated an accuracy of ±1.2 mm. This decrease in accuracy was due to facial movement during data acquisition.

However, this method also has limitations. The manual spinal path determination is also likely to cause palpation errors. This limits the usage of the system to only experienced healthcare practitioners who have good palpation skills.

**Figure 13.** *Microsoft Kinect sensor.*

Additionally, this tool is capable of only measuring the shape of the human spine and not the complete back or human body.

### *4.2.2.3.4 Kinect sensors*

Microsoft kinetic sensors are currently being used in a range of disciplines from biomechanics to clinical applications [83, 84]. Castro et al. [85] described the use of the Microsoft's Kinect™ to measure back surface and posture. The Kinect sensor consists of two cameras, a colour camera (RGB camera) (A) and a depth (infrared IR) camera (B), and a projector (C) (please see **Figure 13**). These cameras do not require passive markers to determine anatomical landmarks. By measuring the deformations of the projected speckle pattern, a 3D map of the dorsal skin surface is created by using the appropriate software.

The results from previous studies have demonstrated that the depth sensor is valid in measuring 3D back surface in patients with scoliosis and in healthy volunteers [85, 86]. The Microsoft Kinect™ system had comparable intertrial reliability (ICC difference = 0.06 ± 0.05; range, 0.00–0.16) and excellent concurrent validity against a benchmark reference, a multiple-camera 3D motional analysis system, with Pearson's r-values >0.90 for the majority of measurements (r = 0.96 ± 0.04; range, 0.84–0.99).

Whilst the Microsoft Kinect™ is inexpensive, portable and offers good repeatable of the 3D map of the back surface, it also has a few limitations. The measurements are limited only to the back surface and not the whole body. Additionally, the Kinect system software is mainly restricted to the Microsoft operating system and is not applicable to any other mobile applications.
