Vertebral Metrics

Automation of a Non-invasive Instrument to Analyse the Spine

A. Gabriel

1,2

, C. Quaresma

2,3

and P. Vieira

1,2

Center of Atomic Physics, Faculty of Sciences and Technology New University of Lisbon, Caparica, Portugal

Department of Physics, Faculty of Sciences and Technology New University of Lisbon, Caparica, Portugal

CEFITEC, Faculty of Sciences and Technology New University of Lisbon, Caparica, Portugal

Keywords: Spinal Column, Biomechanics, Non-invasive, Medical Device, Stereo Vision.

Abstract: Back pain is a major health problem in modern society. It is known that the main responsible for this

symptom are the biomechanical changes in the spinal column. Thus, it is important to develop an instrument

that evaluates in a global way the spinal column in a standing position. For that purpose, a complete

innovative and non-invasive system – the Vertebral Metrics – has been built. The Vertebral Metrics is a

semi-automatic equipment designed to identify the spatial position of each spinal process from the first

cervical to the first sacral vertebra. Using a camera and a laser diode the recognition is achieved with

software capable of distinguishing prominent blue marks that identifies each spinous process. After the

validation process it was concluded that the Vertebral Metrics is a reliable and valid instrument. However,

the time required for a completely scan is still too long for practical use. Thus, we are presently working on

the automation of this medical device by developing a new prototype. The process will comprise several

modifications in the equipment as the introduction of fluorescent markers, UV lights and two video

cameras. The identification of the vertebras will be performed by the stereo vision method. The software

must be adapted to the new assembly.

1 INTRODUCTION

Back pain has affected humans throughout out

recorded history and it is well documented to be a

major health problem in modern society (Alexandre,

2001, Dankaerts, 2006). It is the leading cause of

activity limitation and work absence (Dionne, 2006)

so it has a significant impact on individuals,

families, communities, governments and business

(Hoy, 2010). In fact, the incidence of rachialgia is so

high that it must be studied as if it were an epidemic

and social disease (Knoplich, 2003; Galukande,

2005). Most researchers suggest that about 80% of

the individuals experience back pain at some point

of their life and, from those, 80 to 90% of the pain is

caused by mechanical changes in the spine (Najm,

2003; Quaresma, 2010).

Currently, many options are available to evaluate the

spine. There are several types of diagnostic imaging

technologies to assist in identifying the anatomical

changes, responsible by back pain. However, very

few of them are ionising radiation free and do not

allow the analysis of the spinal column in a vertical

standing position. Furthermore they tend to evaluate

only parts of this osseous structure (Secca, 2008).

The radiological studies, particularly X-rays, are the

most widely used methods for assessing the spinal

column curvatures in a global way, however, since it

uses ionizing radiation, it is not possible to use them

in general population (Harlick, 2007; Pinel-Giroux,

2006). X-rays are associated with 0.6 – 3.2% of

cumulative risk of cancer to age 75 years old

(Gonzales, 2004) so its use must be reduced to a

minimum.

Non-invasive equipment for evaluation of the spine

is commercially available, but often only perform

partial scans of the spine or are extremely expensive

(Hinman, 2004; Quaresma, 2013). Thus, the

development of non-invasive methods to evaluate

the spine in a standing position in a global way is

needed to attain a better insight into back disorders

(Quaresma, 2009).

For that purpose, the Vertebral Metrics was built. It

is a non-invasive system which is able to identify the

X, Y and Z position of each vertebra, from the first

cervical to the first sacral vertebra. The mechanical

equipment was originally planned and built to be

applied to pregnant women (Quaresma, 2009). After

150

Gabriel A., Quaresma C. and Vieira P..

Vertebral Metrics - Automation of a Non-invasive Instrument to Analyse the Spine.

DOI: 10.5220/0005276101500155

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2015), pages 150-155

ISBN: 978-989-758-071-0

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

an automation process, a semi-automatic prototype

was developed. This prototype can be applied in

general population. Its setup uses a camera and a

laser diode to measure the spatial coordinates of

each spinal process. Using an adequate blue marker

to identify the spinal processes, the recognition is

achieved with software capable of distinguish the

prominent blue marks in the skin. However, the time

required for data acquisition is about three minutes.

Artefacts in measurements are affecting the

performance of the system because individuals can

not stand still for such a long period of time. This

constraint limits the practical use of the equipment.

For that reason, the further development of the

Vertebral Metrics has become necessary. The aim of

this paper is to present a third improved prototype

that is being developed. This system will have

several differences comparing to the previous

prototypes. With the new equipment tests will be

faster with an improved resolution.

2 AUTOMATION OF THE

VERTEBRAL METRICS

The main purposes of the automation of the

Vertebral Metrics are the improvement of the

resolution and time required for data acquisition.

For a better understanding of the instrument that

is under study the axis system considered will be

defined as following: the transversal distance as X

coordinate, the antero-posterior distance as the Y

coordinate and the height as Z coordinate.

Several modifications will be performed

comparing to the existing prototype. The equipment

that is being projected will move along the Z

direction only. It has been designed to have two

video cameras and two UV lights that will be fixed

in specific positions. A fluorescent dye will be used

to identify the projection of the spinal processes.

This dye emits a specific wavelength when exposed

to the ultra-violet radiation. Also the triangulation

method, that uses a RGB camera and a laser diode,

will be replaced by the stereo vision method. The

software has to be adapted to perform scans with the

new model.

The characteristics of the automated system will

be described in the following topics.

2.1 Mechanical System

The design of the mechanical system has been

already projected using SolidWorks™ (Figure 1). It

is a vertical structure which has a bracket coupled.

This bracket consists of two identical video cameras

positioned between two ultraviolet lights.

Figure 1: Scheme of the mechanical system.

There are two main goals for the mechanical

structure: resolution and speed. The instrument is

being constructed based on the following pre-

requisites:

- The mechanical apparatus should move all the

hardware necessary for the image acquisition in the

Z direction (up and down).

- Must move 1000 mm in 30 seconds (33.3

mm/s).

- Must have a communication protocol

controlled by software.

For the vertical positioner a linear stage will be

used. It has a high helix pitch lead screw providing

high-speed positioner. The motor will be step-by-

step type for precision positioning. As in the

previous prototype, the motion and control of the

motor will be performed by a microcontroller

PIC16F877 from Microchip™. In addition, the

microcontroller will also monitor limit switches,

control serial RS232 communications and control

lights relays.

In terms of RS232 communication a protocol

will be developed based on instructions and replies.

VertebralMetrics-AutomationofaNon-invasiveInstrumenttoAnalysetheSpine

151

Therefore, the microprocessor will only answer and

execute commands from the computer.

Figure 2: Scheme of the communication system.

2.2 Determination of Distances Using a

Stereo Vision System

Stereo vision is a method that provides 3D

perception by means of two different images of the

same scene. To develop a stereo vision system some

specific elements are required. The geometry of the

system is also a relevant concern.

In order to engineer a stereo vision system, two

video cameras will be fixed side by side in a

horizontal support with parallel optical axes. Both

cameras will be equidistant from the middle of the

support (Figure 3). This distance has to be studied in

order to maximize the performance of the

equipment.

Figure 3: Scheme of the stereo vision system (x0z plane).

The system must have at least 1 mm of resolution

in all directions (X, Y and Z). To achieve the desired

resolution several calculations that involve

trigonometric equations have been studied in detail.

It was concluded that any camera available in the

market would be sufficient to identify the X and Z

coordinates. However, powerful cameras are

required to assess the Y coordinate. The selected

cameras are from IDS™, model UI-3480ML. They

are monochromatic cameras with 2560x1920

resolution. A lens with 12 mm of focal distance

(Optica Goyo from IDS™) will be coupled to each

camera.

Software must be developed in order to determine

the spatial position of each spinal process through

the stereo vision method. Different approaches of

the image processing algorithms are still being

studied at this stage of the project.

2.3 Detection of the Spinal Processes

Two programming softwares - Matlab™ and Visual

Studio™ - will be used to develop all software

including the algorithms to perform the detection of

the spinal processes. The image processing

algorithms will be implemented in Matlab™

(MATrix LABoratory, a numeric computer

environment for programming). Visual Studio™ will

be used to define functions for communication with

the mechanical equipment as well as the user

interface.

Selecting an appropriate marker to identify the

patients’ skin above the vertex of each the spinal

process is the first step required to the development

of the image processing algorithms. The marker will

be a fluorescent dye. When excited by ultraviolet

light this dye emits a well-known wavelength that

must be different from the wavelength of the

ultraviolet light. Presently, several spectroscopic

tests are being performed with multiple dyes in order

to choose the most adequate (Figure 4). It is

noteworthy that the dyes under study are suitable for

skin.

Only wavelengths emitted by the marker will be

detected by the cameras. The emission spectrum of

the ultraviolet lights (Figure 4c) has a clear peak at

400 nm so the wavelength of the emission peak of

the adequate markers must be quite different. By

observing all the emission spectrums it was

concluded that the dyes iv and v might be adequate

markers (Figure 4g and Figure 4h, respectively)

because they emit between 625 nm and 675 nm.

Spectral tests are still being performed with those

two dyes in order to achieve the most appropriate.

Optical filters must be positioned in front of each

camera in order to prevent other wavelengths. As the

emission wavelength of the iv and v dyes comprises

the red region of the visible spectrum, the red

component of the RGB image was used to simulate

the filter (Figure 4b). Both dyes are completely

distinguishable from the skin, therefore, the

positioning of the filters in front of the cameras may

simplify dramatically the definition of the

algorithms.

BIODEVICES2015-InternationalConferenceonBiomedicalElectronicsandDevices

152

Figure 4: Study of fluorescent dyes. a) RGB picture of the fluorescent dyes in skin; b) Red component of the previous

image; c) Emission spectrum of the UV lights; d) Emission spectrum of the first dye (i); e) Emission spectrum of the second

dye (ii); f) Emission spectrum of the third dye (iii); g) Emission spectrum of the fourth dye (iv); h) Emission spectrum of

the fifth dye (v).

While the vertebral metrics equipment is operating

the data acquisition process will be simple. In the

beginning of each scan the physician has to move

the mechanical system until the first sacral vertebra

is seen by the cameras. An intuitive interface is

being developed to assist the positioning. After the

physician orders the mechanical system begins its

upward movement. The software has a cyclical

VertebralMetrics-AutomationofaNon-invasiveInstrumenttoAnalysetheSpine

153

behaviour wherein the cameras acquire a pair of

images simultaneously every 20 ms. Each pair of

images is saved in memory and analysed by the

image processing algorithms when the scan is

completed. It is noteworthy that data acquisition will

be achieved in such a short period of time – about 30

seconds - because the image processing will be

performed after a full scan of the spine.

The start position of the mechanical system, its

vertical speed and the time between the acquisition

of each pair of images as well as the stereo vision

method are essential to determine the spatial position

of the vertebras. The physician has the possibility of

save data acquired in a single file. By running this

file in a specific software the 3D reconstruction of

the spinal column can be observed. An example that

uses data acquired with the second prototype is

presented in Figure 5.

Figure 5: 3D reconstruction of the spinal column.

3 CONCLUSIONS

The present study will contribute to the automation

of a non-invasive equipment that is a complete

innovation in health – the Vertebral Metrics.

Throughout this article, the characteristics of the

new prototype of the Vertebral Metrics were

described. At this moment a few steps were already

taken: the scheme of the new prototype was drawn,

the characteristics of the cameras were determined

through trigonometric equations and studies with

fluorescent dyes are being performed. The future of

the project will comprise the implementation of

software, image processing algorithms and

instrumentation. Furthermore, the equipment must

be validated. Tests in people with biomechanical

changes of the spine will be performed with

Vertebral Metrics and compared with X-ray (the

gold standard) in order to prove that it is reliable

system.

The differences between the previous prototypes

and the system that is being built will be particularly

relevant regarding the acquisition time required to

collect the individuals data. The stereo vision

method will improve the resolution and the accuracy

of the equipment. As data acquisition will be

performed faster it will also allow the analysis of the

postural adjustments. However, the patient’s skin

above the vertex of the spinal processes still needs to

be marked by the physician.

Due to its non-invasive characteristics the

Vertebral Metrics will be a powerful tool to evaluate

the spinal column. In a few seconds the new

equipment will provide a comprehensive overview

of the spine as it will perform a complete scan in

about 30 seconds. Furthermore, it will became

available a three dimensional analysis of the vertex

of the spinal processes.

This unique and already patented device will

avoid unnecessary invasive methodologies as well as

qualified staff for screening and prevention of spinal

disorders because it will allow repeated scans

without causing damage to the individuals. It may be

used in different contexts such as ambulatory public

and private (healthcare centre / doctor’s office),

hospital and as a potential screening device for back

pain in the general population. Using the Vertebral

Metrics will also became available the definition of

the most appropriate intervention methodologies to

each particular clinical case.

ACKNOWLEDGEMENTS

The authors would like to acknowledge NGNS-

Ingenious Solutions for the support provided.

REFERENCES

Alexandre, N., Moraes, M., 2001. Modelo de avaliação

fisio-funcional da coluna vertebral. In Rev. Latino-am

Enfermagem Março, 9(2): 67-75. RESEARCH GATE.

Dankaerts, W., O’Sullivan, P. B., Burnett, A. F., Straker,

L. M., 2007. The use of a mechanism-based

classification system to evaluate and management of a

patient with non-specific chronic low back pain and

motor control impairment - A case report. In Man Ther

12: 181-19. PUBMED.

Dionne, C. E., Dunn, K. M., Croft, P. R., 2006. Does back

pain prevalence really decrease with increasing age? A

systematic review. In Age Ageing35(3):229-34.

OXFORD JOURNALS.

Galukande, M., Muwazi, S., Mugisa, D. B., 2005.

Aetiology of low back pain in Mulago Hospital,

Uganda. In Afr Health Sci5:164-167. PUBMED.

BIODEVICES2015-InternationalConferenceonBiomedicalElectronicsandDevices

154

Gonzales, A., Darby, S., 2004. Risk of cancer from

diagnostic X-rays: estimates for the UK and 14 other

countries. In Lancet 345-341. PUBMED.

Harlick, J., Milosavlsejevic, S., 2007. Palpation

identification of spinous processes in the lumbar spine.

Man Ther. 12: 56-62. PUBMED.

Hinman, M., 2004. Comparison of Thoracic Kiphosis and

Postural Stiffness in Younger and Older Women. In

Spine Journal, 4: 413-417. PUBMED.

Hoy, D., Brooks, P., Blyth, F., Buchbinder, R., 2010. The

epidemiology of low back pain. In Best Practice &

Research Clinical Rheumatology 24:769-781.

SCIENCE DIRECT.

Jordão, A., Duque, P., Quaresma, C., Vieira, P., 2011.

Development of Vertebral Metrics – An instrument to

study the vertebral column. In Biodevices 2011 –

Proceedings of the International Conference on

Biomedical Electronics and Device: 224-229.

SCITEPRESS.

Knoplich, J., 2003. Enfermidades da coluna vertebral,

Robe Editorial. São Paulo. 3

edition.

Najm, W., Seffinger, M., Mishra, S., Dickerson, V.,

Adams, A., Reinsch, S., Murphy, L., Goodman, A.,

2003.Content validity of manual spinal palpatory

exams – A systematic review. In BMC

Complementary Altern Med 3:1. SPRINGER LINK.

Pinel-Giroux, F., Mac-Thiong, J., Guise, J., Labelle, H.,

2006.Computorized assessment of sagittal curvatures

of spine – Comparison between Cobb and tangent

circles tehniques.In Spinal Disord Tech vol. 19(7):

507-512. PUBMED.

Quaresma, C., João, F., Fonseca, M., Secca, M. F.,

Veloso, A., O’Neil, J. G., Branco, J., 2009. Validation

of Vertebral Metrics: A Mechanical Instrument to

Evaluate Posture of the Spinal Column. In World

Congress on Medical Physics and Biomedical

Engineering 25:711-713. SPRINGER LINK.

Quaresma, C., João, F., Fonseca, M.,Secca, M. F., Veloso,

A., O’Neill, J. G., Branco, J., 2010. Comparative

evaluation of the tridimensional spine position

measured with a new instrument (Vertebral Metrics)

and an optoelectronic system of stereophoto-

grammetry. In Med Biol Eng Comput 48: 1161-1164.

PUBMED.

Quaresma, C., Dias, I., Secca, M. F., O’Neill, J. G.,

Branco, J., 2013. Métrica Vertebral: a Aplicação de

uma Nova Tecnologia na Análise da Coluna Vertebral.

In Acta Med Port 2013 May-Jun 26(3): 226-230.

RESEARCH GATE.

Secca, M. F., Quaresma, C., Santos, F., 2008. A

Mechanical Instrument to Evaluate Posture of the

Spinal Column in Pregnant Women. In IFMBE

Proceedings 22: 1781-1784. SPRINGER LINK.

VertebralMetrics-AutomationofaNon-invasiveInstrumenttoAnalysetheSpine

155