Intelligent artificial spinal disk Paper


Design of intelligent spinal artificial disk prosthesis as an in-body
implantable load cell to measure in-vivo loading on the spine and in-vitro
experiments in an animal spine using the same disk
M. P. Pancholi, P. A. Kyriacou, and J. Yeh
Abstract—The in-vivo loading on the spinal disk is probably the missing link in the jigsaw puzzle of total chronic low back pain
solution. In this study the artificial spinal disk is used as a base for making in-body intelligent telemetrized implantable load-cell
which can measure the in-vivo loading on the spinal disk non-invasively in human. This study serves two unique purpose one is to
measure non-invasively in-vivo multi-direction loading on the spinal disk in human and other is to develop new generation
intelligent telemetrized artificial spinal disk prosthesis. The artificial spinal disk is loaded with eight straingauges installed on
different locations and two piezoresistive sensors to give complete load mapping on the disk. The in-vitro loading results with and
without animal spine vertebrae gives promising results of sensors output which is consistent, reliable and can be used for intended
object of making intelligent artificial spinal disk prosthesis which can measure multi-direction in-vivo loading on the disk.
Keywords: in-vivo spinal loading, spinal disk, Artificial spinal disk prosthesis, Disk Degenerative Diseases, Lumbar spinal disk,
Low back pain, Intelligent artificial disk prosthesis, Biomechanics.
I INTRODUCTION
THe low back pain is a economic and social burden to the society and its total solution requires very systematic, long term,
multi-angle and multi-disciplinary approach. The causes of the low back pain are mainly the back tissue-muscles, the
degenerative spinal disk and the damaged bones/vertebrae. The low back pain due to tissues or muscles is not considered as
chronic and can be treated easily but the low back pain due to degenerative disk and damaged vertebrae are considered
chronic. The root causes for degenerative disk is extremely hard to find out. This is strongly related to the mechanical loading
on the spine [1] [2] [3] [15]. Any problems like degenerative disk and low mineral density vertebrae imbalance the dynamics
of biomechanics of the spine and further damaged the other healthy parts of the spinal biomechanics. The essential but still
missing part is unavailability of in-vivo data of loading of the spinal disk. Many efforts had been made and still being made
by researchers to find the correct in-vivo loading data on the spinal disk in human using different techniques which are not
actually in-vivo techniques and hence, their findings are questionable [1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [15] [13]. Not
only correct in-vivo measurement of spine loading, but also the distribution of the loading on the spinal disk are of prime
importance in many of the ways like for proper understanding of biomechanics of the spine and its parts, total solution of the
low back pain, better solution for disk degenerative disease, better revolutionary-efficient designs of the prosthesis, better
post-surgery management of the patient, continuous real time monitoring of spinal loading etc. It is also very helpful for
treatment of the vertebrae compression fractures due to trauma, low bone mineral density or multiple myeloma. It seems this

is the only unknown thing which is missing in jigsaw puzzle of total solution of the low back pains and of proper
understanding of spinal biomechanics.
II Material and method for developing load-cell to measure in-vivo loading on the spinal disk
In this study the artificial spinal disk is used as a base for monitoring in-vivo loading on the spinal disk. The required
material and method mainly comprises of artificial disk with sensors as a load-cell, experimental set-up including loading
UTN machine and specifically designed mechanical tools-fixtures, data-acquisition system including hardware and software,
freshly harvested animal spinal vertebrae, statistical software for analysis-presentation of the results.
A Design and development of the load cell with experimental set-up
Fig. 1 Aesculap Activ-L™ Artificial Disc (size M)
The correct and proper design of the loading cell is very crucial to measure the right in-vivo loading on the spinal disk. The
artificial spinal disk prosthesis was selected as a
Fig. 2. Experimental set-up with each part’s actual pictures.
base for the development of the load cell. The biomechanically very important reason for this selection is that all the load
passes through the original spinal disk must also be passed through the artificial spinal disk prosthesis. The commercial
Activ-L™ artificial spinal disk prosthesis (Aesculap, B-Braun, Germany) is used for this experiment. The artificial disc used
was suitable for placing in L4/L5 (between lumbar 4 and 5 vertebrae), which is one of the most common location of
degenerated disk in humans. The artificial disk comprises of mainly three parts (see figure 1), the upper end-plate, the lower
plate (both made-up of Cobalt-Chromium alloy) and the inlay material (UHMW Polyethylene).
Out of eight straingauges four strain gauges were installed on upper end-plate and other four on lower end-plate as shown in
fig.2. Out of two piezoresistive (FlexiForce®, Tekscan Inc., MA, USA) sensors one placed on top of the inlay material and
other below the inlay material as shown in fig.2. The more details on this set-up are described in a research paper [1]. Hence,
all design will give comprehensive monitoring of loading on the disk which is one of the main objects of this study.

B Harvested animal spine vertebrae
The sensor loaded disk is placed between two vertebrae of animal spine as it normally placed in human
spine. The selection of animal is very important because the normally available animals have spine in horizontal direction
where as humans have spine placed in vertical direction. Due to that direction of loading on the disk is different in animal and
human, width to breadth ratio of vertebrae-disk contact surface is different of human
and animals. As mentioned before the sensor loaded artificial spinal disk have size is M and as per dimension given in the
fig.1. After searching and checking many animal spines like Sheep, Goat, Cow and calf. The calf’s spine vertebrae are used
for proper size matching of the spinal disk. Freshly harvested calf spine is used at laboratory of Royal Veterinary College,
UCL, Potter’s bar, Hertfordshire, UK. As shown in fig,2, the vertebrae are hold in specifically designed and developed
cylindrical hollow aluminum tool. The polyester filler has been filled around the spinal vertebrae in the hollow cylindrical
tool to hold the vertebrae strongly. The filler is left 24 hours for drying to make it strong/hard enough for loading
experimentation.
C Signal conditioning and data acquisition system
A signal processing and data acquisition system has been developed to process all the signals acquired from all sensors,
digitise, display and store them on a computer (Fig. 3). All sensor output signals were digitized (sampling rate at 100 Hz)
using an NI CompactDAQ USB Data Acquisition System (National Instruments Corporation, Austin, Texas). The digitized
signals were analyzed by a Virtual Instrument (VI) implemented in LabVIEW®
(National Instruments Corporation, Austin,
Texas). This VI read the voltage outputs from all sensors, converted them into a spreadsheet format and saved them into a file
specified by the user and displayed the signals in real time on the screen of the computer.
D E
xperim
ental
metho
d
The
compr
essive
loadin
g was
applie
d in
the
norma
l
directi
on to
the
artifici
al disc
(with
all
sensor
s embedded)
Fig. 3. Block Diagram of Signal Conditioning and Data Acquisition System.
using a DARTEK®
, Universal Testing Machine (computer controlled by Instron®
, Bucks, UK). In this study the main
objective was to evaluate the experimental set-up and confirm that all sensors produce meaningful outputs when the artificial
disc (loading cell) subjected to loading from 0-1KN with and without the two animal spinal vertebrae. The secondary object

is to study the effect on sensor’s output with and without animal spinal vertebrae. The load that was applied to the disc was
from 0 to 1 kN, which is well below the natural operative range of the loads that the human spinal disc can be exposed [14].
With animal spinal vertebrae and as per experimental set-up the loading more than 1KN the disk start shifting from its
original location between the spinal vertebrae. That’s why the loading upto 1 kN is taken into consideration. In this
experiment loading speed was 10 NPS (Newton per Second).
III Result and analysis
Fig.4 and Fig.5 shows the results of all sensors output when loading from 0-1 kN with 10 NPS loading speed without
placing sensor loaded artificial spinal disk (load-cell) between animal spinal vertebrae. Fig.6 and Fig.7 shows the results of
all sensors output when loading from 0-1 kN with 10 NPS loading speed and placing sensor loaded artificial spinal disk
(load-cell) between animal spinal vertebrae. The Strain gauge-0 to strain gauge-3 are installed
on inferior disk and strain gauge-4 to strain gauge-7 are installed on superior disk. One Piezoresistive sensor-Flexiforce®
named F.force_upper installed top on the inlay material and other installed below the inlay material named F.force_Lower. In
Fig.4, the all graphs show time in second
on x-axis and the first eight graphs in first two rows show strain gauge’s output in microstrain on y-axis. In Fig.4, the first
and second graphs from left show piezoresistive sensor-Flexiforce’s output in volt on y-axis and in third graph from left
shows analog output in Newton.In Fig.5, the all graphs show applied compressive load in Newton on x-axis and the first
eight graphs in first two rows show strain gauge’s output in microstrain on y-axis. In Fig.4, the first and second graph from
left show piezoresistive sensor-Flexiforce’s output in volt on y-axis. Fig.4 & Fig.5 show that without animal spinal vertebrae
with the same loading range of 0-1000N the graphs of the sensor’s outputs are almost very much linear. In Fig.4 & Fig.6
show the red colour graph shows the original row data where as black colour line-graph shows result after data filtering and
best curve fitting. In Fig.5 & Fig.7, the Load Vs Sensor’s output graphs drawn after filtering and best curve fit of raw data. In
Fig.6, the all graphs show time in second on x-axis and the first eight graphs in first two rows show strain gauge’s output in
microstrain on y-axis. In Fig.4, the first and second graphs from left show piezoresistive sensor-Flexiforce’s output in volt on
y-axis and in third graph from left shows analog output in Newton. In Fig.7, the all graphs show load in Newton on x-axis
and the first eight graphs in first two rows show strain gauge’s output in microstrain on y-axis. In Fig.4, the first and second
graph from left show piezoresistive sensor-Flexiforce’s output in volt on y-axis. In case of loading without animal spinal
vertebrae outputs of strain gauges are higher with compare to loading with animal spinal vertebrae specifically for inferior
plate. The values of outputs of Flexiforce® sensors vary dramatically and unpredictably because the little dislocation of
Flexiforce® sensor at the time of apply loading on the disk.

Fig.4: All Straingauge’s output Vs Time (straingauge 0 to straingauge 7 – eight strain gauges-First two row graphs), Piezoresistive sensor’s output Vs Time
(Flexiforce®_Upper - Top of Inlay material and Flexiforce_Lower-Below the inlay material-First and second graphs from left in third row) and Analog
output of applied Load Vs Time (Third graph from left in third row) when loading from 0-1 kN without animal spinal vertebrae.

Fig. 5: All Straingauge’s OUTPUT Vs Applied load (straingauge 0 to straingauge 7 – eight strain gauges-First two row graphs), Piezoresistive sensor’s
OUTPUT Vs Applied load (Flexiforce_Upper - Top of Inlay material and Flexiforce_Lower-Below the inlay material-First and second graphs from left in
third row) when loading from 0-1 kN without animal spinal vertebrae.

Fig 6: All Straingauge’s output Vs Time (straingauge 0 to straingauge 7 – eight strain gauges-First two row graphs), Piezoresistive sensor’s output Vs Time
(Flexiforce_Upper - Top of Inlay material and Flexiforce_Lower-Below the inlay material-First and second graphs from left in third row) and Analog output
of applied Load Vs Time (Third graph from left in third row) when loading from 0-1 kN with animal spinal vertebrae.

Fig.7: All Straingauge’s output Vs Time (straingauge 0 to straingauge 7 – eight strain gauges-First two row graphs), Piezoresistive sensor’s output Vs Time
(Flexiforce_Upper – Top of Inlay material and Flexiforce_Lower - Below the inlay material-First and second graphs from left in third row) when loading
from 0-1 kN with animal spinal vertebrae.
III
IV Conclusion
The graphs of the outputs of strain gauges in Fig.4, Fig.5, Fig.6 and Fig.7 show that all strain gauge’s output characteristics
are similar in nature and that the upper plate exhibits more strain than the lower plate. It was also interesting to notice in these
experiments the outputs of the sensors describe a little visco-elastic behavior which is obvious. After these experiments it is
very clear that the sensor’s outputs to measure loading on the artificial spinal disk prosthesis are in strong and reliable
relationship with the applied load with some tolerance but at the same time, it is very difficult to interpret the results from the
graph about loading on the artificial disk. Moreover the calibration of sensors is also very crucial and require to research
further. Further research experiments are required by changing speed of loading, application of load at different angle.
Further live animal trials are also required on successful positive results second level of experiments with cadaver spine. At
the same time, design of low power wireless, batteryless circuit needs to be developed for final product.
References
1 P. A. Kyriacou, M. P. Pancholi and J. Yeh, ―Investigation of the in-vitro loading on an artificial spinal disk‖, J. Phys.: Conf.Ser. 178012023, doi:
10.1088/1742-6596/178/1/012023, Volume 1, 2009.
2 M Liuke, S Solovieva, A Lamminen, K Luoma, P Leino-Arjas, R Luukkonen and H Riihimäki, ―Disc degeneration of the lumbar spine in relation to
overweight‖ International Journal of Obesity (2005) 29, 903–908. doi:10.1038/sj.ijo.0802974; published online 17 May 2005
3 IAF Stokes, JC Iatridis – “Mechanical Conditions That Accelerate Intervertebral Disc Degeneration: Overload Versus Immobilization.‖ Spine, 2004 -
spinejournal.com
4 Nachemson A, Elfstrom G. Intravital dynamic pressure measurements in lumbar discs. Scand J Rehabil Med Suppl;S1:1–40, 1970.
5 Ledet EH, Sachs BL, Brunski JB, Gatto CE, Donzelli P. Real time in vivo loading in the lumbar spine. Part 1: interbody implant load cell design and
preliminary data. Spine;25(20):2595–600, 2000.
6 Ledet EH, Tymeson MP, DiRisio DJ, Cohen B, Uhl RL. Direct Real-Time measurement of in vivo forces in the lumbar spine. The Spine Journal 5, 85-
94, 2005.
7 McGill SM. A myoelectrically based dynamic three-dimensional model to predict loads on lumbar spine tissues during lateral bending. J
Biomech;25(4):395–414, 1992

8 Morlock MM, Schneider E. Determination of the magnitude of lumbar spinal loading during different nursing activities. Proceedings of the 44th
Annual Meeting of the Orthopaedic Research Society, March 16–19, New Orleans, Louisiana, Chicago: Orthopaedic Research Society, 1998.
9 Han JS, Goel VK, Ahn JY, et al. Loads in the spinal structures during lifting: development of a three-dimensional comprehensive biomechanical
model. Euro Spine J;4:153–68, 1995.
10 Schultz AB, Ashton-Miller JA. Biomechanics of the human spine. In: Mow VC, Hayes WC, editors. Basic orthopaedic biomechanics. New York:
Raven Press, Ltd,:337–74, 1991.
11 Dolan P, Adams MA, Kingma I, de Looze MP, van Dieen J, Toussaint HM. The validity of measurements of spinal loading during manual handling.
Proceedings of the 44th Annual Meeting of the Orthopaedic Research Society, March 16–19, New Orleans, Louisiana, Chicago: Orthopaedic Research
Society, 1998.
12 Rohlmann A, Bergmann G, Graichen F. Loads on an internal spinal fixation device during walking. J Biomech;30(1):41–7, 1997.
13 Patwardhan, AG, Meade KP, Lee B. A ―follower load‖ increases the load carrying capacity of the lumbar spine in axial compression. Spine; 24(10):
1003-9, 1999.
14 Book: White, A.A. III & Panjabi, M.M.: Clinical Biomechanics of The Spine, 2nd
ed., 1990, Lippincott Williams & Wilkins.
15 Cholewicki J, McGill SM, Norman RW. Lumbar spine loads during the lifting of extremely heavy weights. Med Sci Sports Exer; 23(10):1179–86,
1991.

Intelligent artificial spinal disk Paper

More Related Content

What's hot (20)

Similar to Intelligent artificial spinal disk Paper (12)

Recently uploaded (20)

Intelligent artificial spinal disk Paper