In-Vitro Detection of Small Isolated Cartilage Defects: Intravascular Ultrasound Vs. Optical Coherence Tomography
In-Vitro Detection of Small Isolated Cartilage Defects: Intravascular Ultrasound Vs. Optical Coherence Tomography
0 Department of Biomechanical Engineering, Delft University of Technology , Delft , The Netherlands
1 Department of Orthopedic Surgery, Academic Centre for Evidence-based Sports Medicine (ACES), Academic Medical Centre , Amsterdam , The Netherlands
2 Department of Orthopaedics, UMC Utrecht Regenerative Medicine Centre , Utrecht , The Netherlands
3 Research Centre Smart Devices, Zuyd University of Applied Sciences , Heerlen , The Netherlands
4 chanical Engineering, Delft University of Technology , Delft , The Netherlands. Electronic mail:
-This experimental work focused on the sensor selection for the development of a needle-like instrument to treat small isolated cartilage defects with hydrogels. The aim was to identify the most accurate and sensitive imaging method to determine the location and size of defects compared to a gold standard (lCT). Only intravascular ultrasound imaging (IVUS) vs. optical coherent tomography (OCT) were looked at, as they fulfilled the criteria for integration in the needle design. An in-vitro study was conducted on six human cadaveric tali that were dissected and submerged in saline. To simulate the natural appearance of cartilage defects, three types of defects were created via a standardised protocol: osteochondral defects (OCD), chondral defects (CD) and cartilage surface fibrillation (CSF), all sized between 0.1 and 3 mm in diameter. The detection rate by two observers for all diameters of OCD were 80, 92 and 100% with IVUS, OCT and lCT, for CD these were 60, 83 and 97%, and for CSF 0, 29 and 24%. Both IVUS and OCT can detect the presence of OCD and CD accurately if they are larger than 2 mm in diameter, and OCT can detect fibrillated cartilage defects larger than 3 mm in diameter. A significant difference between OCT-lCT and IVUS-lCT was found for the diameter error (p = 0.004) and insertion depth error (p = 0.002), indicating that OCT gives values closer to reference lCT. The OCT imaging technique is more sensitive to various types and sizes of defects and has a smaller diameter, and is therefore preferred for the intended application.
Orthopedics; OCT; IVUS; Needle intervention
Ankle sprains and cartilage fractures are common.
In up to 50% of all ankle sprains and fractures,
isolated cartilage defects occur.20 These defects involve
the articular cartilage and subchondral bone, which
can result in deep ankle pain, stiffness of the joint and
impaired movement, and if left untreated may evolve
into posttraumatic osteoarthritis.36 Therapeutic
treatments have been developed to alleviate pain, to restore
functionality and to extend the time until total joint
replacement.4,5,25 The treatments depend on the size
and severity of the cartilage defect, starting with
conservative treatment for International Cartilage Repair
Society (ICRS) Grade I and II defects, followed by
surgical treatment for ICRS Grade II?IV defects and
implants for partial joint replacement.5,21,25,34
However, evidence indicates less favourable results after
surgical treatment for larger and older cartilage defects
with disturbed joint homeostasis16,26 and indicates
some healing potential for small cartilage defects
(ankle < ? 8 mm).9,12,15
This knowledge combined with the recent
development of injectable hydrogels2,13,17,29 sets the stage for
even less invasive interventions at an early stage, for
example by using a needle intervention to seal small
isolated cartilage defects with hydrogel. Figure 1 is an
artist?s impression of a multifunctional steerable needle
device currently in development for such a medical
intervention. For these slender needle-like instruments,
it is crucial that the needle tip finds the exact location
of the isolated cartilage defect within the joint. This
requires local cartilage tissue characterisation.
An overview of tissue characterisation techniques is
given by Nieminen.19 We foresee the use of a tissue
2018 The Author(s)
characterisation imaging system within a needle-like
instrument outside the operation room in various
scenarios, such as day-care clinics and, ultimately,
sport fields. Therefore, the targeted imaging system
should be slender, flexible, small (diameter < 2 mm)
and mobile. This rules out traditional imaging
techniques such as CT and MRI (not mobile32), as well as
arthroscopes or arthroscopic characterisation
techniques, such as the streaming potential method
(diameters > 2 mm and rigid).8 Two possible methods
were left that meet the design criteria: intravascular
ultrasound (IVUS)11,14,33 and optical coherence
tomography (OCT).6,7,27 IVUS and OCT use the
reflections of sound and light, respectively, from tissue
to create an image of the tissue. Both methods are
clinically available. Their predominant application is in
vessel imaging during coronary interventions, and in
the last decade they have also been studied for
application in cartilage characterisation.11,14,33 IVUS and
OCT provided quantitative information on repaired
chondral and osteochondral defects 2?8 mm in
diameter in fresh bovine, equine and caprine
models.6,7,11,14,27 In these studies, IVUS was capable of
discriminating between fibrillated and intact cartilage
and providing quantitative information on defect
depth and surface roughness.11,14,33 OCT has been
shown to be a reproducible technique in quantitative
analysis of cartilage thickness and to a lesser extent
when measuring repair tissue area and surface
roughness.1,6 However, to detect and treat isolated defects at
an early stage, we estimated that defect sizes as small as
1 mm in diameter and fissures 0.1 mm in width should
be measured.9,28,30 This should be done in different
human intra-articular joints, with ankle joints being
challenging as the cartilage thickness is relatively small
(mean of 1 mm).1,28 Furthermore, the appearance of
isolated defects differs depending on whether the
trauma mechanism was wear or tear, namely a
complete crack with or without bone damage, bruising or
Therefore, the aim of the present research was to
assess the performance of IVUS and OCT in detecting
small and different types of cartilage defects in human
cadaveric tali compared to a reference micro-computed
tomography (lCT), in order to select the most accurate
and sensitive technique for integration in the
multifunctional steerable needle device.
An in-vitro study was conducted on human
cadaveric ankle specimens. Tali of six different human ankles
with no known age that had previously been stored at
2 26 C were prepared to expose the bone and the
cartilage surface by removing the soft tissue. The bones
were frozen again directly after preparation and
thawed an hour before the experiment, during which
each specimen was hydrated with saline. Directly
afterwards the specimens were frozen and imaged with
the lCT scanner on the same day. Three types of
cartilage defects were created: osteochondral defects
(OCD), chondral defects (CD) and chondral surface
fibrillation (CSF) (Fig. 2).
Creation of Defect Types and Sizes
A setup was designed to create each defect type in a
standardised manner (Fig. 3). All defects were created
perpendicular to the cartilage surface by manual
adjustment of the specimen vice under visual
inspection (Fig. 4e). OCD and CD defects were created using
a hand-operated drill with 0.5, 1, 2 and 3 mm drill bits.
The drilling depth was set at either 4 mm to create an
OCD or 1 mm to create a CD. A surgical ??Beaver??
knife was used to create CD 0.1 mm in diameter. CSF
was created with 140 grit sandpaper33 that was glued to
disks 1, 2 and 3 mm in diameter, which were then
attached to the bottom of a bolt. The bolt was pressed
against the cartilage surface and a manual half turn
back and forth of the bolt was repeated five times to
create the surface damage. Care was taken to create
each defect condition at another location on each
specimen to compensate for the effect of site
dependence.9,23,28 Other than that, the locations were
assigned randomly and each defect condition was created
at least six times.
To avoid changes in tissue status due to degradation
or unintentional repositioning, the cartilage surface
location was measured consecutively by both IVUS
and OCT with a custom-made plastic catheter holder
that holds the catheter tips in the same position and
orientation above the target (Fig. 4, left side). The
setup allows the catheter holders to be changed
without moving the specimen or setup. The complete
measurement setup (Fig. 4, right side) consisted of a
robust stainless steel frame (a) to which the plastic
plate with bolts (b) was attached. A catheter was
inserted through holes in the bolts (c). The frame allows
the catheter to be manually adjusted both horizontally
and vertically to accurately position it above the target
location on the surface. The specimen (d) was fastened
in a customised specimen vice (e) that was placed in a
glass tank filled with saline solution (f). The specimen
vice was equipped with custom-made beaks that
contained a grid of sharpened adjustable stainless steel
bolts that facilitate firm fixation. The vice?s ball joint
allows the alignment of the specimen perpendicular to
the catheter and allows the positioning of the cartilage
surface as close as possible to the catheter transducer
by visual inspection.
Two mobile catheter imaging systems were used:
intravascular ultrasound (IVUS) and optical coherence
tomography (OCT). The IVUS measurement system
consisted of a clinical IVUS main unit (Volcano
Corporation, CA, USA) in combination with a Volcano
Revolution 45 MHz rotational Imaging Catheter
(Volcano Corporation, CA, USA) with a distal outer
diameter of 3.2 Fr (~ 1.2 mm) and an intravascular
resolution of 50 lm. The automatic pullback system
can scan a trajectory of up to 150 mm. The
high-resolution IVUS images were captured and stored as 2D
images (TIFF format). The OCT imaging system was
the C7-XRTM Intravascular Imaging System (St. Jude
Medical, Inc., MN, USA), which emits near-infrared
light with a data sampling rate of 100 frames per
second to produce real-time, ultra-high definition images
combined with the DragonflyTM Imaging Catheter (St.
Jude Medical, Inc., MN, USA). This catheter, which
has an outer diameter of 2.7 Fr (~ 0.9 mm), produces
images with an axial resolution of 15 lm and a lateral
resolution of 30?35 lm at 1300 nm.30 The automatic
pullback system can scan a trajectory of up to 520 mm.
For both systems, we set a pullback trajectory of
30 mm, which covered the surface area of the tali. The
high-resolution OCT images were captured and stored
as 2D images (TIFF format). As reference imaging,
lCT scans were made of all prepared specimens before
and after creating the defects in air. The lCT scanner
(Quantum FX, Perkin Elmer, Waltham, MA) was set
with the following parameters: 90 kV tube voltage,
180 lA tube current, 3 min. scan time and isotropic
voxel size of 42 9 42 9 42 lm. These settings in
combination with scanning in air gave sufficient
contrast to image cartilage. The 3D reconstructed images
were converted to 2D image of the slices (TIFF
format) by using the built-in software of the lCT scanner
The six frozen specimens were scanned with the lCT
scanner and checked for any existing defects. Each of
the defrosted specimens was then fixed in the vice and
placed inside the tank. The ball joint of the vice was
adjusted to align the selected defect location parallel to
the catheter. A single measurement trial was conducted
as follows: first, the pre-defect measurements with
IVUS and OCT were conducted. Both IVUS and OCT
catheters were pulled along a trajectory of 30 mm
along the cartilage surface to mimic the intended
application in a needle-like instrument. This implied
that the distance and orientation between the
transducer and cartilage surface could show variations.
Second, the defect was created at a predefined location.
Third, post-defect measurements were conducted with
IVUS and OCT. After this trial, the vice?s ball joint
was adjusted so that the next defect location was
aligned with the catheter, and the process was repeated
until all grid areas on the specimen surface had been
used. After all defects had been created and measured
with both imaging systems, the specimen was removed
from the setup and refrozen. Finally, a second lCT
scan was made of all six frozen specimens.
The reconstructed images of IVUS and OCT were
stored without further processing. The open-source
image processing software Image J (version 1.50a) was
used for image post-processing. The lCT images were
reoriented to show the defect using a stack alignment
(Align3 TP, version 2010/11/12)
.23 All TIFF
stacks were examined and the image that showed the
largest defect diameter at the defect bottom was
defined as the centre of the defect. The cartilage was
inspected adjacent to the defect in the same image to
prevent misinterpretation due to unintended cartilage
damage of any type. The cartilage thickness of each
specimen was measured directly adjacent to each
defect at the first location the lCT image slice showed a
smooth, undisturbed cartilage layer. The defect
diameter and depth values provided by the
measurement software were rounded off to the nearest
Drill Hole Visualization Errors
Five different holes (0.1, 0.5, 1, 2 and 3 mm in
diameter) were drilled to a depth of 1 mm (CD) and
4 mm (OCD). In this study, lCT was regarded as the
gold standard and OCT and IVUS were investigated
for their accuracy and sensitivity compared to lCT. By
subtracting the measured lCT diameter from the
measured OCT and IVUS diameters, a value remains
that represents the measured diameter difference (thus
OCT?lCT and IVUS?lCT). Similarly, the drill depth
measured with lCT was subtracted from the drill
depth measured with OCT and IVUS and the
difference was used for comparison purposes.
In the first step of this study, two observers were
used to assess each surface defect on the cartilage. The
success rate per imaging system was expressed as the
percentage of correctly determined defects per type of
surface defect (OCD, CD, and CSF). A defect was
considered real only when both observers were able to
detect it. Its dimensions were then measured during the
second step of this study. The sensitivity of the imaging
systems were considered equal when an equal number
of defects were found within a 95% accuracy.
Statistical analyses were performed with IBM SPSS
Statistics (v22, IBM Corp., NY, USA) to indicate the
presence of any difference in the value of the diameter
and insertion depth between the imaging systems
compared to the gold standard lCT (i.e., IVUS?lCT
vs. OCT?lCT). An f test was used to test the null
hypothesis that the variances of the populations are
equal (p > 0.05). In the case of equal variance, a
double-sided paired Student?s t test was used to test the
null hypothesis that the means of two populations are
equal. The statistical significance was set at p < 0.05.
In total, 25 OCD, 30 CD and 21 CSF defects were
created on the six tali. Figure 5 shows a similar sized
OCD defect as visualized with the different imaging
techniques. Both observers correctly identified the
presence of all diameters of OCD defects in 80% of
the cases with IVUS, 92% with OCT and 100% with
lCT. For all diameters of CD defects, the
figures were 60% with IVUS, 83% with OCT and 97%
with lCT, and for all diameters of CSF defects, 0%
with IVUS, 29% with OCT and 24% with lCT
(Fig. 6). The percentage of observed CSF defects was
too low and the diameter and depth of this type of
damage was too difficult to measure. Therefore, only
the OCD and CD defects were further analysed and
The cartilage thickness for all specimens was on
average 0.9 mm (standard deviation of 0.1 mm). The
measured diameter and insertion depth for OCD and
CD for IVUS and OCT are shown in Fig. 7. The
differences in measured diameter and depth for OCD and
CD for OCT?lCT and IVUS?lCT are shown in
Fig. 8. The statistical outcomes are presented in
Table 1. When testing for equality of variance, no
significant difference was found between the imaging
systems. Therefore, a Student?s t test for equal variance
was used for further statistical analysis.
Analysis of the diameter and insertion depth
difference for OCD showed no significant difference
between OCT?lCT and IVUS?lCT. Analysis of the
diameter difference for CD showed a significant
difference between OCT?lCT and IVUS?lCT, with an
underestimation for OCT and an overestimation for
IVUS. Analysis of the insertion depth difference for
CD showed a significant difference between OCT?lCT
and IVUS?lCT, both giving an underestimation, with
OCT giving a lower value.
With the introduction of injectable hydrogels, the
need has arisen for new needle-like instruments to
allow early-stage interventions for small cartilage defects
in joints. Such a new type of steerable needle is under
development (Fig. 1). To allow the correct localisation
of the needle tip on top of the isolated damaged
cartilage prior to injection and sealing with a hydrogel, we
studied two catheter imaging systems for their
suitability for integration in the slender steerable needle.
The detection rate of OCT and IVUS was measured
compared to the gold standard lCT, and it was at least
80% for OCD and 60% for CD (Fig. 6) for diameters
ranging from 0.5 to 3 mm, which were smaller than
previously tested.6,33 OCT and IVUS are depth-limited
techniques where the depth of penetration varies from
1?2 mm for OCT to 4?8 mm for IVUS. The achievable
resolution with OCT is 10?20 lm whereas that for
IVUS is 100?150 lm.24
The results indicates that the thin talar cartilage
regions can be successfully visualised by both
techniques. The detection rate was higher for OCT
compared to IVUS, especially when the defects were
smaller and less deep, because the IVUS resolution
limits its detection rate. This was also confirmed by the
data given in Fig. 7, which indicate that it was not
possible to measure the detected 0.1 and 0.5 mm
diameter holes with IVUS. The higher detection rate in
combination with the smaller diameter of the OCT
catheter makes OCT the preferred candidate for
integration in needle-like hydrogel delivery devices to be
used for OCD and CD defect types (Fig. 1).
The detection rate of CSF was rather low for all
imaging systems (Fig. 6). This is in contrast to earlier
studies by Viren33 and by Saarakkala et al.27 In-depth
analysis indicates that this is predominantly caused by
the differences in study design: rather old, previously
frozen human bones with a thin cartilage layer vs. fresh
Osteochondral defects (OCD)
Chondral defects (CD)
animal bones with a thick cartilage layer, fine grinding
(140 grit) vs. course grinding (60 grit), small defect size
(< ? 3 mm) vs. large defect size (> ? 5 mm),
variation in distance vs. optimal constant distance to the
surface, B-mode analysis vs. ToF signal analysis.
Furthermore, during this study, the bone specimens
underwent multiple freeze?thaw cycles for logistic
reasons. We expect that this marginally influenced the
morphology of the bone and cartilage defects. As this
was a predominant outcome parameter that influenced
all conditions equally, the influence on the relative
comparison of the two imaging systems can be
considered negligible. However, further studies are needed
to quantify the true influence of freeze?thaw cycles, as
this limitation makes comparison with other studies
such as Viren33 and Saarakkala27 difficult.
We also processed a single measurement per
condition vs. taking the average of up to 10 measurements
in the other studies.33 The latter approach obviously
increases the signal-to-noise ratio and enhances the
image quality of CSF for both OCT and IVUS. IVUS
images show quite some noise, which is typical in
sound reflection imaging modalities (Fig. 4). It could
be that the CSF was performed too gently without
causing enough macroscopic structural cartilage
disruption and that the noise is in the same range as that
caused by CSF. In retrospect, we could have made
arthroscopic or microscopic photographic images to
assist the lCT recordings. On the other hand, these
differences indicate the contribution of our study. Our
study design of taking a single measurement and using
a pullback trajectory that allows the distance between
the transducer and cartilage surface to vary, mimics
more closely the actual clinical practice and the
intended manual steering of the foreseen needle-like
instrument. Apparently, this does imply that certain
defect types and sizes can no longer be detected.
In the present study, outcome measures were based
on geometrical differences rather than full tissue-quality
characterisation. The dimensions of the defects can be
accurately assessed using the photographic images of
IVUS and OCT. Furthermore, lCT is known to be
superior in detecting bony tissue and has the advantage
of providing 3D geometrical information as opposed to
histology.22 In this sense, the use of lCT as the gold
standard is defendable, especially to image cartilage, as
we verified that cartilage could be seen with our
scanning protocol in air. The latter is also confirmed by the
cartilage thickness measurements that we performed,
which are in line with the results of other studies.1,3,18
Contrast-enhanced CT using Hexabrix or CA4 + or
histology could be used in future studies to assess the
cartilage quality.35 Finally, histological analysis of
multiple sections taken from and around the defect sites
could have highlighted more details on site dependence
and cartilage composition.
Although OCT has a 46% higher success rate than
IVUS when detecting CSF defects, it is unlikely that the
steerable needle device will be used for the treatment of
surface fibrillation with hydrogel as there is not much
space to fill. Thus, OCT imaging applied in the
proposed application provides a promise for the future of
early-stage defect detection and intervention in a
onestop solution for patients who have sustained a trauma
known to have a high likelihood of cartilage damage.
Although both catheter imaging techniques allow
the measurement of defect diameters and depths for
OCD and CD, the results presented in Fig. 7 indicate
that the measured values can deviate by twice the
actual diameter and depth as defined by the drill bits that
were used to create the defects. One reason for this is
that the composition of cartilage and bone material did
not allow the creation of holes with higher accuracy.
Another reason is that the small OCD and CD defects
were in the spatial resolution range of all three imaging
modalities and the thickness of one trabecula (ca.
0.3 mm). Therefore, just penetrating a trabecula wall
(wall thickness of around 0.05 mm) with the drill tip
can result in a depth measurement error of up to
0.3 mm, as the open space in the trabecula is possibly
added to the drill depth during the measurement.10
Finally, the cylindrical 2D images from the catheters
do show some errors due to catheter surface debris
(Fig. 5, 0.5 mm IVUS), non-uniform rotation
distortion (NURD) effects (Fig. 5, 0.5 and 2 mm IVUS),
and reflection and refraction caused by the harder
bone that result in artefacts and deformation, which
makes it more difficult to estimate the dimensions.31
Therefore, further experiments should also include a
focus on improvements in catheter purging, catheter
tip actuation and optimisation of the distance between
surface and catheter.
Figure 7 shows that a mean difference of 0.6 mm
was found for the estimated defect depth of 4 mm deep
holes between IVUS and OCT, showing that IVUS has
more difficulty in assessing the correct depth.
However, comparing the mean insertion depth difference
for the 4 mm deep hole of both catheter imaging
techniques with the gold standard (1.1 mm IVUS and
1.6 mm OCT vs. 2.5 mm lCT) indicates that neither
imaging technique should be used to determine the
depths of deeper holes. For clinical application, the
found accuracy and variance are acceptable, because
for the intended application an estimate of the volume
of the defect is relevant to deciding the volume of
hydrogel that should be injected, but this could also be
verified real-time when the catheter is present.
In the present study, a method was employed that
uses an automated retraction mechanism to scan a
particular area of an open joint. In order to investigate
its applicability in a more realistic setting, a closed
joint study should be performed to evaluate the ease of
use of the total needle imaging concept. In the
presented setup, the distance between scan area and
catheter is fixed. In order to determine the accuracy of
the images in a real-life setup, off-centre tests and the
creation of accuracy plots vs. distance from the centre
should be investigated in a series of additional
experiments. As catheter holders were interchanged during
the experiments, a lot of effort was made to
standardize the experiment with a strong focus on
repeatability. However, due to some tolerances in the
components, it is still possible that not exactly the same
slice was taken when making the images with each
different imaging method. To reduce this risk in further
experiments, it is advised to use small metal markers as
Previous studies where OCT catheters have been
manually inserted into joints7 showed that
manoeuvring the catheter inside the joint can be challenging.
To investigate whether OCT catheters have a
minimally invasive clinical application in orthopaedics, the
next step would be to implement the OCT catheter in
the design shown in Fig. 1, followed by a series of
cadaver experiments conducted with orthopaedic
surgeons to investigate the practical aspects of this
combined device. In the last phase of this research, the
reliability of this kind of system inside and outside the
hospital should be determined.
Mechanically created defects as small as 0.1 mm can
be detected with an OCT catheter in in-vitro human
talar cartilage and bone, making it the preferred
candidate for future integration in needle-like instruments
for the precise localisation of the instrument relative to
The authors thank Dr. Gijs van Soest of the
Erasmus MC for sharing his insights into IVUS and OCT
imaging techniques and for helping to interpret the
disturbances found in some of the 2D images. We also
thank NWO for financing the ASPASIA project that
made this research possible.
This article is distributed under the terms of the
Creative Commons Attribution 4.0 International
which permits unrestricted use, distribution, and
reproduction in any medium, provided you give
appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons license, and
indicate if changes were made.
1Adam, C. , F. Eckstein , S. Milz , and R. Putz . The distribution of cartilage thickness within the joints of the lower limb of elderly individuals . J. Anat . 193 ( 2 ): 203 - 214 , 1998 .
2Amini , A. A. , and L. S. Nair . Injectable hydrogels for bone and cartilage repair . Biomed. Mater . 7 ( 2 ): 024105 , 2012 .
3Athanasiou, K. A. , G. G. Niederauer , and R. C. Schenck . Biomechanical topography of human ankle cartilage . Ann. Biomed. Eng. 23 ( 5 ): 697 - 704 , 1995 .
4Bhosale , A. M. , and J. B. Richardson . Articular cartilage: structure, injuries and review of management . Br. Med. Bull . 87 ( 1 ): 77 - 95 , 2008 .
5Cain , E. L. , and W. G. Clancy . Treatment algorithm for osteochondral injuries of the knee . Clin. Sports Med . 20 ( 2 ): 321 - 342 , 2001 .
6Cernohorsky, P. , A. C. Kok , D. M. D. Bruin , M. J. Brandt , D. J. Faber , G. J. Tuijthof , et al. Comparison of optical coherence tomography and histopathology in quantitative assessment of goat talus articular cartilage . Acta Orthop . 86 ( 2 ): 257 - 263 , 2015 .
7Cernohorsky, P. , S. D. Strackee , D. M. de Bruin , M. B. van Herk , D. J. Faber , T. G. van Leeuwen , and J. Bras . Insitu imaging of articular cartilage of the first carpometacarpal joint using co-registered optical coherence tomography and computed tomography . J. Biomed. Opt . 17 ( 6 ): 060501 , 2012 .
8Changoor , A. , J. P. Coutu , M. Garon , E. Quenneville, M. B. Hurtig , and M. D. Buschmann . Streaming potentialbased arthroscopic device is sensitive to cartilage changes immediately post-impact in an equine cartilage injury model . J. Biomech. Eng . 133 ( 6 ): 061005 , 2011 .
9De, M. M. , M. Shahabpour , P. R. Van , and C. Pouders . MRI of cartilage and subchondral bone injury. A pictorial review . JBR-BTR 91 ( 1 ): 6 - 13 , 2008 .
10den Dunnen, S. , J. Dankelman , G. M. Kerkhoffs , and G. J. Tuijthof . How do jet time, pressure and bone volume fraction influence the drilling depth when waterjet drilling in porcine bone ? J. Mech. Behav. Biomed. Mater . 62 : 495 - 503 , 2016 .
11Huang, Y. P. , and Y. P. Zheng . Intravascular ultrasound (IVUS): a potential arthroscopic tool for quantitative assessment of articular cartilage . Open Biomed. Eng. J . 3 : 13 , 2009 .
12Jackson , D. W. , P. A. Lalor , H. M. Aberman , and T. M. Simon . Spontaneous repair of full-thickness defects of articular cartilage in a goat model: a preliminary study . JBJS 83 ( 1 ): 53 - 64 , 2001 .
13Jin, R. , L. M. Teixeira , P. J. Dijkstra , C. A. Van Blitterswijk , M. Karperien , and J. Feijen . Enzymatically-crosslinked injectable hydrogels based on biomimetic dextranhyaluronic acid conjugates for cartilage tissue engineering . Biomaterials 31 ( 11 ): 3103 - 3113 , 2010 .
14Kaleva , E., T. Vire? n, S. Saarakkala, J. Sahlman , J. Sirola , J. Puhakka , et al. Arthroscopic ultrasound assessment of articular cartilage in the human knee joint: a potential diagnostic method . Cartilage 2 ( 3 ): 246 - 253 , 2011 .
15Khan , I. M. , S. J. Gilbert , S. K. Singhrao , V. C. Duance , and C. W. Archer . Cartilage integration: evaluation of the reasons for failure of integration during cartilage repair. A review . Eur. Cell Mater . 16 ( 3 ): 26 - 39 , 2008 .
16Loomer, R., C. Fisher, R. Lloyd-Smith , J. Sisler , and T. Cooney . Osteochondral lesions of the talus . Am. J. Sports Med . 21 ( 1 ): 13 - 19 , 1993 .
17Mart ??nez- Sanz , E., D. A. Ossipov , J. Hilborn , S. Larsson , K. B. Jonsson , and O. P. Varghese . Bone reservoir: injectable hyaluronic acid hydrogel for minimal invasive bone augmentation . J. Control. Release 152 ( 2 ): 232 - 240 , 2011 .
18Millington, S. A. , M. Grabner , R. Wozelka , D. D. Anderson , S. R. Hurwitz , and J. R. Crandall . Quantification of ankle articular cartilage topography and thickness using a high resolution stereophotography system . Osteoarthr. Cartil . 15 ( 2 ): 205 - 211 , 2007 .
19Nieminen, H. J. , Y. P. Zheng , S. Saarakkala , Q. Wang , J. Toyras , Y. P. Huang , and J. S. Jurvelin . Quantitative assessment of articular cartilage using high-frequency ultrasound: research findings and diagnostic prospects . Crit. Rev. Biomed. Eng . 37 ( 6 ): 461 - 494 , 2009 .
20O'loughlin , P. F., B. E. Heyworth , and J. G. Kennedy . Current concepts in the diagnosis and treatment of osteochondral lesions of the ankle . Am. J. Sports Med . 38 ( 2 ): 392 - 404 , 2010 .
21O'driscoll , S. W. Current concepts review-the healing and regeneration of articular cartilage . JBJS 80 ( 12 ): 1795 - 1812 , 1998 .
22Palmer , A. W. , R. E. Guldberg , and M. E. Levenston. Analysis of cartilage matrix fixed charge density and threedimensional morphology via contrast-enhanced microcomputed tomography . Proc. Natl. Acad. Sci . 103 ( 51 ): 19255 - 19260 , 2006 .
23Parker, J. A. Stack Alignment Align3 TP Plugin , 2012 /12/ 12 ed., 2010 .
24Rathod, K. S. , S. M. Hamshere , D. A. Jones , and A. Mathur . Intravascular ultrasound versus optical coherence tomography for coronary artery imaging-apples and oranges . Interv. Cardiol. Rev . 10 ( 1 ): 8 - 15 , 2015 .
25Redman, S. N. , S. F. Oldfield , and C. W. Archer . Current strategies for articular cartilage repair . Eur. Cell Mater . 9 ( 23 -32): 23 - 32 , 2005 .
26Robinson , D. E. , I. G. Winson , W. J. Harries , and A. J. Kelly . Arthroscopic treatment of osteochondral lesions of the talus . J. Bone Joint Surg . 85 ( 7 ): 989 - 993 , 2003 .
27Saarakkala, S. , S. Z. Wang , Y. P. Huang , and Y. P. Zheng . Quantification of the optical surface reflection and surface roughness of articular cartilage using optical coherence tomography . Phys. Med. Biol . 54 ( 22 ): 6837 , 2009 .
28Shepherd , D. E. T. , and B. B. Seedhom . Thickness of human articular cartilage in joints of the lower limb . Ann. Rheum. Dis . 58 ( 1 ): 27 - 34 , 1999 .
29Spiller, K. L. , S. A. Maher , and A. M. Lowman . Hydrogels for the repair of articular cartilage defects . Tissue Eng. B 17 ( 4 ): 281 - 299 , 2011 .
30Tuijthof, G. J. M. , A. C. Kok , M. P. Terra , J. F. A. Aaftink , G. J. Streekstra , C. N. van Dijk , and G. M. M. J. Kerkhoffs . Sensitivity and specificity of ultrasound in detecting (osteo) chondral defects: a cadaveric study . Ultrasound Med. Biol . 39 ( 8 ): 1368 - 1375 , 2013 .
31Unal, G., S. Lankton , S. Carlier , G. Slabuagh, and Y. Chen . Fusion of IVUS and OCT through semi-automatic registration . Proc. CVII-MICCAI , 163 - 170 .
32van Bergen, C. J. , R. M. Gerards , K. T. Opdam , M. P. Terra , and G. M. Kerkhoffs . Diagnosing, planning and evaluating osteochondral ankle defects with imaging modalities . World J. Orthop . 6 ( 11 ): 944 , 2015 .
33Vire? n, T., S. Saarakkala , E. Kaleva, H. J. Nieminen , J. S. Jurvelin , and J. To? yr a?s. Minimally invasive ultrasound method for intra-articular diagnostics of cartilage degeneration . Ultrasound Med. Biol . 35 ( 9 ): 1546 - 1554 , 2009 .
34Williams, R. I. , and R. H. Brophy . Cartilage repair procedures: clinical approach and decision making . Instr. Course Lect . 57 : 553 - 561 , 2008 .
35Xie , L. , A. S. Lin , M. E. Levenston , and R. E. Guldberg . Quantitative assessment of articular cartilage morphology via EPIC-lCT . Osteoarthr. Cartil . 17 ( 3 ): 313 - 320 , 2009 .
36Zengerink, M. , P. A. Struijs , J. L. Tol , and C. N. Van Dijk . Treatment of osteochondral lesions of the talus: a systematic review . Knee Surg. Sports Traumatol. Arthrosc . 18 ( 2 ): 238 - 246 , 2010 .