-
1
Clinical and postmortem diffusion MRI for deep brain stimulator
electrode localization in Essential Tremor patients
by
Jingxian Zhang
Trinity College Department of Neuroscience
Duke University
Thesis committee
Nandan Lad, MD, PhD; Supervisor
G. Allan Johnson, PhD
Evan Calabrese, PhD
Thesis submitted in fulfillment of the requirements for the
degree of
Graduation with Distinction in Bachelor of Science in the
Department of Neuroscience in Trinity College
of Duke University
2015
-
2
Dedication
This thesis is dedicated to my dearest siblings Alicia and
Michael Zhang.
It is also dedicated to Karishma Popli, carry on Karishma!
-
3
Contents Abstract
....................................................................................................................................
5
List of Tables
............................................................................................................................
6
List of Figures
..........................................................................................................................
7
Acknowledgements.................................................................................................................
8
1. Introduction
.........................................................................................................................
9
2.
Methods...............................................................................................................................12
2.1 DBS patient data
.........................................................................................................12
2.1.1 Pre-operative anatomic
MRI.............................................................................13
2.1.2 Pre-operative diffusion tensor MRI
.................................................................13
2.1.3 Post-operative
CT..............................................................................................14
2.1.4 Post-operative electrode testing and
outcomes...............................................14
2.2 Fiber
tractography......................................................................................................14
2.2.1 Deterministic fiber tractography in clinical diffusion MRI
............................15
2.2.2 Analysis of DBS contact proximity with clinical DRT
tractography .............15
2.2.3 Probabilistic fiber tractography in postmortem diffusion
MRI .....................16
2.2.4 Analysis of DBS contact proximity with postmortem DRT
tractography ....17
2.2.5 Analysis of DBS contact proximity with postmortem
rendering of Vim......18
3. Results
.................................................................................................................................18
3.1 DBS Lead and Contact segmentation and modeling
...............................................19
3.2 Diffusion Tensor Imaging and Fiber tractography
..................................................20
3.3 Deterministic
tractography........................................................................................21
3.4 Probabilistic
tractography..........................................................................................21
3.5 Registration of postmortem MRI to clinical
datasets...............................................22
-
4
3.6 Statistical analysis of DBS contact position and clinical
efficacy ............................23
4. Discussion
...........................................................................................................................24
4.1 In vivo patient dataset analysis
..................................................................................24
4.2 Ex vivo patient dataset analysis
.................................................................................26
References
...............................................................................................................................28
-
5
Abstract:
Deep brain stimulation (DBS) is now the preferred surgical
treatment for a variety of movement disorders. We propose
standardized protocols for modeling implanted
DBS electrodes to better visualize and understand the
correlation between contact
positioning, underlying tractography and efficacy of treatment
outcomes. Imaging
datasets (stereotactic CT, MRI-FLAIR and DTI) of patients
treated for essential tremor with bilateral ventral intermediate
(Vim) nucleus DBS were analyzed, and a standardized protocol was
developed to accurately model the placement of DBS leads
and individual contacts. This was paired with consistent fiber
tractography of the
relevant dentatorubrothalamic tract in each patient dataset:
deterministic fiber tracking
was performed on clinical MRI in Brainlab neuronavigation
software, while probabilistic
fiber tracking was performed on a postmortem diffusion MRI
template of the brainstem
and thalamus in Avizo 3D imaging software. A reliable and
reproducible method to
segment DBS lead and contact positions in relation to the DRT
validates the feasibility
of including DTI fiber tractography-based analyses when studying
targeting, lead
location and programming for DBS. This work could provide
further insight into circuit
modulation of underlying white matter pathways that appear to be
the true targets of
neuromodulation by DBS.
Keywords: Deep brain stimulation; diffusion tensor imaging;
fiber tractography;
Essential Tremor
-
6
List of Tables1
Table 1: Stereotaxic coordinates of DBS electrode contacts for
each patient......................40
Table 2: Patient
Demographics..............................................................................................43
Table 3: Summary of post-operative electrode testing outcomes
.......................................44
Table 4: Results of Spearman rank
correlation.....................................................................47
1 Table formatting credit to Dr. Evan Calabrese; Duke University
Medical Center Dept. of Radiology CIVM
-
7
List of Figures
Figure 1: DRT mapped through three ROIs in the clinical MRI
image set..33
Figure 2: Contact creation and fiber
tractography...............................................................34
Figure 3: Fiber tractography through contact ROI in 10 patients
.......................................35
Figure 4: Probabilistic tractography of postmortem diffusion MRI
DRT36
Figure 5: Registration between in vivo and postmortem
MRI.............................................37
Figure 6: Probabilistic DRT shown with red nuclei, Vim, and
implanted leads ...............38
Figure 7: DBS electrode position relative to postmortem template
DRT in 12 patients....39
-
8
Acknowledgements:
This work was supported by the National Institutes of Health and
the National
Institute of Biomedical Imaging and Bioengineering (grant number
P41 EB015897). I would like to sincerely thank and credit my team
members and major contributors from their various departments in
Duke University Medical Center: Dr. Evan Calabrese and
Dr. G. Allan Johnson in Radiology, Mr. Peter Masso from
Brainlab, Dr. Patrick Hickey in
Neurology, Dr. Christine Hulette in Pathology, Ms. Beth Parente
in Surgery, Dr. Dennis
Turner in Surgery, Dr. Allen Song in Brain Imaging, and Dr.
Guillermo Sapiro in Duke
Biomedical Engineering. I would also like to thank Percy
Rochelle and Robert
Satterwhite for their help in procuring brain specimens, Mark
Martin for help with clinical
image data, Porsche Atwater and Anne Jarvis for their guidance
and support, and Dr.
Shouyin Zhang and Ms. Xiaoling Lu.
Most importantly, I would like to thank my principal
investigator Dr. Nandan Lad
in Surgery, who has been an unwavering source of inspiration and
support, for his
mentorship, time, and encouragement these past two years.
-
9
1. Introduction
Essential tremor (ET) is one of the most common neurological
disorders occurring in approximately 4.0% of individuals aged 40
years and older (Dogu et al., 2003). ET is characterized by action
and postural tremor of the arms, and may involve the head and voice
as well. Most patients experience progression in tremor
severity
over time and other neurological signs such as rest tremor and
ataxia may also arise.
Many patients with ET report only mild symptoms, however, about
three quarters of
patients have significant disability and decreased quality of
life. More than 90% of
patients who seek medical care report disability and 10% of
patients who present to a
movement disorder clinic report severe motor disabilities,
including tremor that
interferes with eating, drinking, writing, or communication
(Louis et al., 2001) . First-line medical treatments include
propanolol and primidone, though medications are typically
effective in only 50% of patients and rarely reduce the tremor
to asymptomatic levels
(Deuschl et al., 2011; Elble and Deuschl, 2009). For the subset
of medically refractory cases surgery is an effective option.
The ventralis intermedius (Vim) nucleus of the thalamus is the
typical intervention target structure in the ventral thalamus for
patients with ET (Deuschl et al., 1998; Hassler et al., 1979;
Schaltenbraaand et al., 1977). Two primary surgical procedures have
been performed in patients with ET: thalamotomy and high frequency
deep brain
stimulation (DBS) (Flora et al., 2010; Lehericy et al., 2001;
Schuurman et al., 2008). Both procedures target the Vim, however
equal success has been reported with
posterior subthalamic area (PSA) modulation, prompting further
investigation regarding the optimal neuromodulation target for
tremor suppression.
-
10
The Vim receives its major afferent projections from the deep
cerebellar nuclei, which then project to the motor cortex.
Microelectrode recording of the Vim in patients with ET identifies
cells discharging in bursts that are time locked to the patients
tremor,
indicating that tremor is associated with an abnormal discharge
in the cerebellothalamic
pathway (Benabid et al., 1996). An interruption in this pathway
from lesion or stimulation provides some theoretical basis for the
empirical observation of tremor
improvement, but a more precise understanding is still
lacking.
High frequency deep brain stimulation (DBS) has largely replaced
the ablative procedures used in the past to treat such movement
disorders (Flora et al., 2010). DBS has been shown to be a
reproducible, adjustable, and reversible neuromodulation technique
(Barkhoudarian et al., 2010; Kumar et al., 2003). Accurate
targeting and selective stimulation are essential in optimizing
symptom alleviation and minimizing
potential side effects. The size and position of the neural
targets is variable, and
indirect targeting is based on atlas-defined coordinates rather
than patient-specific
anatomy, although new acquisition and processing approaches are
addressing this
target localization challenge (Duchin et al., 2012; Kim et al.,
2014; Lenglet et al., 2012). Historically, various imaging
modalities and targeting methods have been used to
achieve successful clinical outcomes, including MR imaging, CT
scanning,
ventriculography, and microelectrode recording (De Salles et
al., 2004; Lee et al., 2005; Mori et al., 1999; Sedrak et al.,
2008). Recent advances in imaging modalities have refined the
visualization of surgical targets and landmarks. Multi-modal and
advanced
neuroimaging techniques such as diffusion tensor imaging (DTI)
show great promise for providing increased sensitivity and
specificity of the underlying structurefunction
-
11
relationships (Basser et al., 2000; Coenen et al., 2011; Hyam et
al., 2012; Mori et al., 1999; Poupon et al., 2000). These
techniques are also being explored for the more immediate clinical
needs of DBS lead targeting, intraoperative testing,
postoperative
programming, patient-specific maps, algorithms and treatment
strategies (Lenglet et al., 2012; Kim et al., 2014).
Patients undergoing DBS offer a unique opportunity to study the
functional
anatomy of stimulation targets in humans. Here, protocols are
presented for modeling
implanted DBS electrode leads and evaluating electrode contact
positions relative to
key fiber tracts mapped by diffusion tensor imaging and fiber
tractography (FT). This has utility and implications not only for
understanding circuit modulation of current grey
matter DBS targets (Vim, subthalamic nucleus, internal globus
pallidus) in movement disorder surgery, but is also critical for
future white matter targets (e.g. fornix in memory disorders,
cingulate in mood disorders) (Gutman et al., 2009; Laxton et al.,
2010; Lozano et al., 2012; Mayberg et al., 2005). The analysis has
been implemented on individual clinical patient datasets as well as
on a postmortem diffusion MRI template of
the brainstem and thalamus transformed into patient image space.
The purpose of
developing these protocols was to standardize the imaging of
these patients and better
visualize and understand the correlation between DBS electrode
contact positioning and
efficacy of treatment outcomes in individual patients. The high
resolution diffusion MRI
postmortem template with 3D nuclei and DRT tract reconstruction
can be registered to
individual in vivo clinical images of DBS patients to correlate
electrode proximity to the
DRT and test for surgical outcomes. Likewise, the direct
analysis of electrode
positioning with DRT tractography in clinical images using
Brainlab software can shed
-
12
light on how DBS lead localization in individual patient anatomy
affects treatment
efficacy and potential side effects, while expanding the
analysis pre-operatively will
pave the way for even better surgical results and more precise
understanding of
structure-function relationships.
2. Methods
2.1. DBS patient data
All experiments on patient image datasets were approved by the
Duke University
institutional review board (IRB). Twelve patients with medically
refractory ET received DBS targeted to the ventralis intermedius
nucleus of the thalamus (Vim). Atlas based targeting of the Vim
nuclei was performed according to standard neuroanatomical
targets relative to the AC-PC plane (Schaltenbrand et al.,
1977). The Vim target was identified on FLAIR imaging using target
coordinates 13 to 15 mm lateral to the anterior
commissure-posterior commissure (AC-PC) line, 0 mm below the
AC-PC plane, and 30% of the total AC-PC distance posterior to the
midpoint of AC-PC.
All patients underwent pre-operative structural and diffusion
tensor MRI, as well
as post-operative x-ray computed tomography (CT) scans to
localize electrode placement. Patients undergoing bilateral VIM DBS
were implanted with two quadripolar
electrodes (model 3389, Medtronic, Minneapolis, Minnesota) for a
total of eight contacts per patient. Contacts measure 1.5 mm in
diameter with a 0.5 mm tip before the most
distal contact and 0.5 mm spacing between contacts. Final
stereotaxic coordinates for
each electrode contact are provided in Table 1. Patient
demographics for both studies
-
13
included 4 males and 8 females with average age 65 15 years
(Table 2); two of the patients from the postmortem diffusion MRI
study were not used in the clinical study due
to CT slicing above 1mm in thickness.
Patients also underwent detailed preoperative, intraoperative
and postoperative
neurological testing for motor and tremor control, sensory and
cranial nerve testing,
muscle tone and spasticity, overall flexibility and reflexes, as
well as cognitive
assessment. The physical examinations were performed by two
senior clinicians
experienced with clinical care of patients with ET.
2.1.1 Pre-operative anatomic MRI2
Pre-operative MR imaging was performed on a 3 Tesla GE Discovery
MR750
scanner (Waukesha, WI). T1-weighted structural images were
obtained with a 3D fast
spoiled-gradient-recalled (FSPGR) pulse sequence (TR = 6.5 ms,
TE = 2.5 ms, = 12,
BW = 140 kHz), at 1 mm isotropic resolution. T2 fluid attenuated
inversion recovery (FLAIR) images were acquired with an
inversion-prepared spin echo pulse sequence (TR = 10,000 ms, TE =
148 ms, TI = 2,250 ms, BW = 781 Hz/pixel) at 1 x 1 mm in-plane
resolution with 1 mm slice thickness and 1 mm spacing between
slices.
2.1.2 Pre-operative diffusion tensor MRI
Diffusion tensor data were acquired with an echo-planar imaging
sequence (TR = 8,000 ms, TE = 84.9 ms, BW = 1562 Hz/pixel) using a
30-direction gradient encoding scheme at b = 1000 s/mm2 with 2
non-diffusion-weighted images. The acquisition
2 Sections 2.1.1 to 2.1.3 credit to Dr. Nandan Lad, Ms. Beth
Parente, Dr. Allen Song; Duke University
Medical Center Dept. of Surgery; Duke-UNC Brain Imaging and
Analysis Center
-
14
matrix was 128 x 128 over a 240 x 240 mm field of view (FOV)
with a slice thickness of 2 mm. Data were zero-filled in k-space to
a matrix size of 256 x 256 prior to
reconstruction for a final nominal voxel size of 0.94 x 0.94 x 2
mm.
2.1.3 Post-operative CT
CT images were acquired on a Siemens SOMATOM Definition Flash
scanner
with a spiral scan using a 512 x 512 Matrix over a 250 x 250 mm
FOV for an in-plane
resolution of 0.484 mm. Approximately 300 contiguous,
non-overlapping, 0.625 mm
thick slices were acquired covering the entire neurocranium,
Additional scan parameters
include MA setting = 250 and kVp = 120. The standard
reconstruction algorithm was
used.
2.1.4 Post-operative electrode testing and outcomes3
Post-operative electrode testing was performed on DBS patients
after initial
surgical recovery. Each contact was tested independently at
voltages ranging from 0.5
to 3 volts dependent on patient tolerance, with frequency
between 135-185 Hz and
pulse width between 60-90 microseconds. For each contact,
treatment efficacy was
recorded on a three level subjective scale that included no
effect, mild/moderate control, and good/excellent control. Patients
experiencing persistent undesired side
effects were recorded on as having significant side effects,
while patients that did not
experience any side effects, experienced transient side effects,
or only presented side
effects at very high testing voltages that were not used in the
final voltage setting were
recorded as no significant side effects. Typical side effects
included paresthesias,
3 Section credit to Dr. Patrick Hickey; Duke University Medical
Center Dept. of Neurology
-
15
worsening of tremor, and/or dysarthria. In some cases an
effective contact was
identified early in testing and the remaining contacts were not
tested to avoid
unnecessary patient discomfort. The corresponding data points
were recorded as n/a.
Outcome testing results are summarized in Table 3.
2.2 Fiber tractography
2.2.1 Deterministic fiber tractography in clinical diffusion
MRI
Deterministic fiber tracking was performed using iPlan
stereotaxy software
(Brainlab, Feldkirchen, Germany). Future research will consider
further validation with other techniques (Aganj et al, 2010; Aganj
et al., 2011; Lenglet et al., 2012). All imaging datasets were
individually loaded into Brainlab iPlan. After ensuring accurate
and
overlapping fusion of all dataset image pairs in the Image
Fusion function, regions of
interest (ROI) were mapped under Fiber Tracking using the FLAIR
MRI image set for clear structural contrast. Fractional anisotropy
threshold and minimum length were
standardized for all patients at 0.2 and 75 mm, while maximal
angle change of fibers
was set at the default value of 70 degrees.
In the case of ET patients, the putative mechanism of action of
DBS of the Vim
thalamus is the modulation of the underlying white matter fiber
tract termed the
dentatorubrothalamic (DRT) tract (Groppa et al., 2014). The DRT
tract is the primary fiber bundle forming the superior cerebellar
peduncle, which is one of the largest
efferent connections of the cerebellum and consists of axon
fibers arising from cells
located in the dentate, emboliform, and globose nuclei. These
fibers then project to the thalamus and terminate in the ventral
lateral and ventral posterolateral thalamic nuclei,
which go on to project to the primary motor cortex (Habas and
Cabanis, 2007; Kwon et
-
16
al., 2011). As such, this system is functionally relevant for
somatomotor coordination and tremor control. Three cubic ROIs were
marked for the DRT tract (Figure 1): 1) in axial section, ROI
around the red nucleus at its widest diameter, 2) in axial section,
ROI around resulting fibers which reached the precentral gyrus
containing the primary motor
cortex (M1), 3) in axial section, end region ROI around the
ipsilateral dentate nucleus to include resulting fibers from the
first ROI which traveled along the superior cerebellar
peduncle (Yousry et al., 1997). In addition, mapping of the most
distal DBS active contact (contact 0) and its relation to the DRT
tract was performed by establishing two cubic ROIs: 1) in axial
cut, the cubic ROI generated for the most distal DBS active contact
(contact 0) and 2) in axial cut, ROI around the precentral gyrus to
include resulting fibers from the first ROI which traveled along
the DRT tract.
2.2.2 Analysis of DBS contact proximity with clinical DRT
tractography
The relevant active contact as determined by post-operative
testing outcomes
was established as the sole ROI for fiber tracking. Tractography
was performed on 10
ET patients using the FLAIR MRI image set for clear structural
contrast at fractional
anisotropy threshold of 0.2 and minimum length 75 mm, while
maximal angle change of
fibers was set at the default value of 70 degrees.
2.2.3 Probabilistic fiber tractography in clinical diffusion
MRI4
Tractography regions of interest (ROIs) were manually segmented
from both anatomic and tensor-derived image data, using a
histology-based human brainstem
4Section credit to Dr. Evan Calabrese; Duke University Medical
Center Dept. of Radiology CIVM
-
17
atlas for reference (Paxinos and Huang, 1995). ROIs for tracking
the DRT included the superior cerebellar peduncles, the red nuclei,
and Vim nuclei of the thalamus.
Tractography fiber data were reconstructed using FSLs BedpostX,
a direct,
multi-fiber orientation estimation algorithm that provides
estimates of fiber distribution
error for probabilistic tractography (Behrens et al., 2007).
These data were used for probabilistic fiber tractography using
FSLs ProbtrackX. Tractography of the bilateral
DRTs was generated using the superior cerebellar peduncle as a
seed region, and the
contralateral red nucleus and Vim nucleus of the thalamus as
waypoints. Tracking
parameters included 5000 seeds per voxel, a step size of 100 m,
and a curvature
threshold of 45 per voxel. Resulting tractography data were
thresholded at > 2000
tracks per voxel, or roughly 1%, consistent with previous work
(Jbabdi et al., 2013).
2.2.4 Analysis of DBS contact proximity with postmortem DRT
tractography
In order to validate the anatomic accuracy of postmortem
tractography results,
we assessed the proximity of each DBS lead to
dentatorubrothalamic tract tractography
from postmortem brainstem datasets after spatial transformation
into the anatomic
space of each patient dataset. Individual DBS leads were clearly
differentiated and
identified on the post-operative CT scan, and the proximity of
each spherical lead to the
postmortem dentatorubrothalamic tract tractography were
classified in a four rank
system. The eight leads per patient were analyzed as separate
data points because
each lead was tested independently for treatment efficacy. DBS
leads directly inside
the tract were ranked as inside, leads contacting the surface of
the rendered tract
were ranked as touching, leads in close proximity to the tract
but not contacting it were
ranked as close, and leads not in close vicinity of the tract
were ranked as not
-
18
touching. The shape and size of the dentatorubrothalamic tract
was highly dependent
on the threshold used for tractography. We determined the area
inside the DRT tract to
be at a threshold of greater than 2000 fibers per voxel, which
is consistent with optimal
threshold recommendations for probabilitstic tractography. The
rank system for DBS
lead proximity was evaluated independently by two observers and
then compiled into a
single numerically ranked datasheet and evaluated with a
Spearman rank correlation.
2.2.5 Analysis of DBS contact proximity with postmortem
rendering of Vim
We also assessed the proximity of each DBS lead to a rendering
of the Vim from
the postmortem brainstem datasets to gauge the correlation
between patient treatment
outcomes and stimulation of the Vim of the thalamus. Vim was
segmented bilaterally
for the postmortem brainstem template referencing anatomy from
the Mai-Paxinos
Human Nervous System atlas. The high contrast resolution of the
postmortem
brainstem datasets allowed for accurate definition of the Vim
apart from the surrounding
thalamic nuclei. Vim segmentation was transformed into each
individual patients
anatomic space using the spatial transformations detailed
earlier. A similar four rank
system as the analysis of DRT tract proximity was used
respective to the Vim, after
which correlation was evaluated with a Spearman statistic
test.
3. Results
Postoperative lead visualization and overlapping of contacts
with their underlying
DTI target has been done in a limited fashion to date. We
utilized the currently
commercially available DBS lead 3389 (Medtronic, Inc.,
Minneapolis, MN) to examine the underlying fiber connectivity and
putative circuits being modulated by stimulation.
-
19
The function of these circuits has been extensively studied in
preclinical models,
however visualization in humans has been limited due to
inability to reliably visualize the
lead and its contacts and their relation to associated fiber
tractography in stereotactic
space.
Furthermore, the ability to visualize the underlying circuit
being modulated and
regional neuroanatomy is critical for optimal clinical results
(Coenen et al., 2011; Henderson, 2012; Hyam et al., 2012).
Specifically, it was found that the DRT tract was reproducibly
visualized passing through the active DBS contacts of 10 ET
patients
undergoing Vim stimulation. These findings are discussed in
detail as follows.
3.1. DBS Lead and Contact segmentation and modeling
Leads and contacts were segmented and modeled using the Object
Creation function after alignment in the View and Adjustment
function in Brainlab software. The post-operative CT image set was
used to align the length of the lead parallel to the
vertical axis of the screen, with the bottom of the lead (where
the contacts are located) centered right on the horizontal
axis.
The lead can then be made into an object using auto segmentation
in Object Creation; a histogram is provided in this function so
that the lead can be outlined by its
difference in radiopacity. This provides a standardized way to
accurately depict the
leads actual shape and slight curvature in its final location
instead of relying on a linear
model as has been done previously (Coenen et al., 2011).
Contacts are 1.5 mm in diameter, and created as new objects
standardized using the brush function; this function allows users
to adjust size of the brush diameter to highlight and create an
object. To correctly model the spacing of the contacts on the
-
20
lead in relation to each other, a void object was also made
using brush size 0.5 mm; this object is used in the sagittal view
to fill out the 0.5 mm dead space at the tip of the lead as well as
adjusted to fill out space between contacts. The electrode pattern
used for patients had between contact spacing fixed at 0.5 mm for
typical Vim DBS leads
(Model 3389). The standardized brush sizes used for contact and
spacing coupled with the exact outline of the lead by density
allows for correct positioning of the dead space
at the lead tip and subsequent contacts/contact spacing in the
simulation. To model
each contact as consistently as possible to their actual shape,
the brush function set at
1.5 mm was used in the sagittal, coronal, and axial planes to
create an approximate
sphere of volume 0.002 cm,verified in the Plan Content
section.
3.2. Diffusion Tensor Imaging and Fiber tractography
To visualize relevant fiber tracts that cross through the
individual DBS lead
contacts, the Fiber Tracking function was utilized to make a new
region of interest from
the previously created contact using the Existing 3D Object
selection. It is then possible to track fibers that run through
this contact to brain structures of interest (Figure 2A).
Figure 2B shows a sample ET patient treated with bilateral Vim
DBS using
modeled electrodes created as regions of interest in the Fiber
Tracking function. To
account for the specificity and small size of this ROI, minimum
fiber length was adjusted to 30 mm, while fractional anisotropy
threshold and maximal angle change of fibers
were held constant at 0.2 and 70 degrees, respectively.
3.3 Deterministic tractography
Deterministic tractography through the active DBS contact set as
the sole
region of interest yielded consistent fiber tracking patterns
across the patient datasets
-
21
(Figure 3). It can be observed that these fiber tracts directly
contacting the DBS electrodes closely follow the DRT tract modeled
previously, and this was consistently
seen throughout the 10 ET patients. As the DBS electrode itself
is set as region of
interest, there is no need for a correlation analysis of the
results: the fiber tracts shown
are in overlapping voxels as the active contact that is used for
treatment. Overlay of
fiber tracts contacting the DBS electrodes with the target DRT
tract can provide analysis
on correlation of electrode positioning with treatment outcomes
for each patient.
3.4 Probabilistic tractography5
Probabilistic tractography with multiple fiber orientations was
used to reconstruct
DRT probability maps in the postmortem brainstem dataset. The
DRT courses from the
dentate nucleus of the cerebellum through the superior
cerebellar peduncle, crosses the
midline in the mid pons, passes through and around the
contralateral red nucleus, and
finally relays in the Vim nucleus of the thalamus before
continuing to cortical motor
areas (e.g. Figure 4A). Importantly, routine clinical diffusion
tractography data are poorly suited for accurately representing
crossing fibers, such as those present in the
midline crossing of the DRT between the dentate nucleus and the
red nucleus.
Probabilistic tractography of the postmortem template correctly
represented the midline
crossing of the DRT as well as its connections to the
contralateral red nuclei and Vim
nuclei (Figure 4AC). Probabilistic tractography also revealed
minor branches of the DRT coursing along the medial aspect of the
red nuclei (Figure 4B). These medial pathways have been observed in
histology studies of the human brainstem (Massion,
5 Sections 3.4 to 3.6 referenced from manuscript submitted to
Human Brain Mapping co-authored
under Dr. Evan Calabrese
-
22
1967), but have not previously been demonstrated with diffusion
tractography. Visualization of probabilistic DRT tractography with
surface renderings of the Vim nuclei
and red nuclei clearly demonstrates the close relationship
between these structures in
the human brainstem, and highlights the potential difficulty in
accurately targeting DBS
electrodes in this complex brain region (Figure 4C).
3.5 Registration of postmortem MRI to clinical datasets
Postmortem brainstem data were non-linearly registered to 12
patient datasets.
We observed considerable variation in patient brain anatomy,
particularly with regard to
ventricle size. Nonetheless, registration of postmortem
brainstem data to patient
datasets yielded good visual alignment (Figure 5). Major
borders, such as the anterior surface of the pons, the dorsal
surface of the thalamus, and the posterior surface of the
tectum showed strong agreement between patient datasets and
registered postmortem
data (Figure 5AB). Smaller features, such as the optic chiasm,
were also very closely aligned after registration (Figure 5BC).
Additional assets from the postmortem dataset, including
probabilistic
tractography and 3D segmentations of the Vim and red nuclei,
were also transformed
into patient image space. These data, combined with
post-operative CT data for
electrode localization, allowed 3D visualization of the complex
spatial relationships
between DBS contacts and the relevant nuclei and white matter
tracts (Figure 6AD). In most clinical datasets, DBS contacts were
clearly visible in post-operative CT data as
four discrete bulges at the distal end of the electrode (e.g.
Figure 6B). In cases where contacts were not clearly visible in CT
data, they could be inferred based on the specific
geometry of the electrodes used (Coenen et al., 2011b). Using
these data, we
-
23
measured the position of each electrode contact with respect to
the DRT and Vim from
the postmortem template. Across all patient datasets, 22/24
electrodes had a least one
contact directly touching the Vim nucleus segmented from the
postmortem template
(e.g. Figure 6AB). In contrast, only 18/24 electrodes had at
least one contact directly touching the thresholded DRT model
(Figure 7).
3.6 Statistical analysis of DBS contact position and clinical
efficacy
In order to assess the accuracy of our postmortem template, we
tested for
statistically significant correlation between electrode
proximity to the DRT and Vim, and
clinical outcomes including treatment efficacy and the presence
of side effects. The
non-parametric Spearman rank correlation was used because
clinical outcomes were
assessed as ordinal variables. The calculated p-values and
R-values (i.e. Spearman correlation coefficients) for each
comparison are presented in Table 4. We observed no significant
correlation between contact proximity to the DRT or Vim and the
presence of
side effects. Despite the fact that the Vim nucleus of the
thalamus was the explicit target
for DBS electrodes, we did not detect a statistically
significant correlation between
treatment outcome and contact proximity to the Vim from the
postmortem template. We
did, however, detect a highly significant, yet weak, positive
correlation between
treatment efficacy and contact proximity to the DRT from the
postmortem template (p = 0.005, R = 0.336). This correlation
remained significant after Bonferroni correction for multiple
comparisons (pcorrected = 0.02). This correlation suggests that our
postmortem DRT model has at least some degree of anatomic relevance
for DBS electrode targeting
in ET patients.
-
24
4. Discussion
Our results describe new protocols for both in vivo and ex vivo
MRI to
consistently model DBS electrode position in a standardized
fashion relative to
visualizations of relevant pathways mapped by diffusion tensor
imaging and fiber
tractography. Reliable fiber tracts and modulation of the
associated sensorimotor
networks in patients with ET was seen, a finding that is
consistent with prior studies
(Benabid et al., 1996; Coenen et al., 2011). The most striking
findings in our analysis are the intimate positioning of DBS leads
with surrounding fiber tracts previously shown
to respond to neuromodulation of the sensorimotor system by
modulating distinct white
matter circuits in ET. In particular, 1) we demonstrate using
individual patient MRI that fiber tracts approximating the DRT
tract in ET patients are stimulated by their active
DBS contact in the thalamus, and 2) we analyze the correlation
between patient DBS contact positioning to probabilistic DRT
tractography and 3D modeling of the Vim in a
high resolution postmortem brainstem and thalamus template.
4.1 In vivo patient dataset analysis
Seeding of the DRT pathway in patients undergoing Vim DBS for
tremor showed
connections from the primary motor cortex (M1), ipsilateral red
nucleus and cerebellum, which is consistent with the published
literature (Coenen et al., 2011). The Vim is generally accepted to
be the cerebellar receiving area of the thalamus before the
fibers
are projected to the primary motor cortex (Hasslet et al., 1979;
Carpenter, 1991). Retrograde tracer studies of herpes simplex
virus-1 from M1 injections in macaques stains both the cerebellum
and globus pallidus (Hyam et al., 2012).
-
25
In the in vivo portion of this study, we used physical
localization of contacts from
individual cases to generate consistent fiber tractography of
tracts directly touching the
contact. The tracts were then visually assessed to approximate
the shape and position
of the DRT tract. Benefits of this approach include bypassing
the statistic correlation
analysis of tract proximity to contact positioning, as the
tracts shown are mapped
through the active contact as ROI. However, this approach can be
greatly refined with 1) overlaying the DRT tract generated through
its three relevant ROIs (red nuclei, dentate nucleus, precentral
gyrus) on top of the active contact generated tractography to
determine if they indeed have significant overlap, 2) quantizing
this overlap numerically with fiber counts or another thresholded
method, and 3) analyzing this degree of overlap for each patient
with their patient outcomes. Generation of connectomic maps of
cortical connectivity by selecting the entire thalamus or entire
cortex relative to the DBS
contact of interest is an alternative to the approach we have
used (Henderson, 2012). Further study is required to test these
methods in other regions of surgical interest such
as the sensory and anterior thalamic nuclei for pain and
epilepsy surgery, respectively
(Bittar et al., 2005; Lega et al., 2010). By combining patient
dataset models with intraoperative and postoperative
clinical results, detailed analyses can be made for each patient
to pinpoint correlations
in electrode placement and treatment outcome that can serve as a
guide for
programming as well as creation of patient-specific models
(Hagmann et al., 2010; Lenglet et al., 2012). Further studies
examining electrical fields of stimulation and surrounding pathways
will be critical to dissect therapeutic stimulation from
potential
stimulation side effects. The ability to visualize the fiber
target of interest in relation to
-
26
the surgical implant allows not only for more accurate
sub-millimiter lead placement, but
also postoperative programming, current steering and novel
stimulation algorithms. DTI
fiber tractography will likely play an increasingly important
role for mapping proposed
white matter DBS targets for large neurological disorders
including Alzheimers Disease
(Fornix stimulation) and Depression (Cingulate Cg25
stimulation).
4.2 Ex vivo patient dataset analysis6
In the ex vivo portion of this study, we present a high spatial
and angular
resolution diffusion MRI template of the postmortem human
brainstem and thalamus
with 3D reconstructions of the nuclei and white matter tracts
involved in ET circuitry. We
demonstrate accurate registration of these data to in vivo,
clinical images from patients
receiving DBS therapy, and correlate electrode proximity to
tractography of the DRT
with improvement of ET symptoms. This serves as a proof of
concept for using high-
resolution postmortem diffusion MRI reference atlases for DBS
targeting. Our results
show that; 1) postmortem diffusion MRI can be used to create a
high-quality, high-resolution template of the human brainstem and
thalamus; 2) these data can be accurately aligned to patient
datasets using automated image registration; and 3) that electrode
position within the registered template has a significant
correlation with
treatment efficacy.
Postmortem diffusion tractography can provide increased anatomic
accuracy
through improved image quality, increased spatial and angular
(diffusion) resolution,
6 Section referenced from manuscript submitted to Human Brain
Mapping co-authored under Dr.
Evan Calabrese
-
27
and reduced image artifacts. High angular resolution diffusion
datasets also allow
advanced tractography methods including probabilistic
tractography with multiple fiber
orientations, which may be more sensitive than standard
deterministic tractography
(Behrens et al., 2007). We were able to achieve good
registration between our postmortem template and 12 patient
datasets, particularly with regard to ventricular
and/or exterior borders of the brainstem and thalamus. We were
able to show a
statistically significant correlation between treatment efficacy
and contact proximity to
the DRT of our postmortem template. Although this correlation
was highly significant, it
was relatively weak, suggesting that other factors play a role
in treatment efficacy. One
interesting result of our study was the lack of a significant
correlation between treatment
efficacy and contact proximity to the Vim, which was the
intended electrode target.
There is increasing evidence that the anti-tremor effects of Vim
DBS are related to
modulation of the DRT rather than the Vim itself. DRT fibers
pass through a portion of
the Vim, and it is possible that stimulation of this area is
principally responsible for
tremor control.
High-resolution MRI-based reference atlases, like the postmortem
data
presented here, could improve on conventional histology-based
atlases by incorporating
accurate volumetric imaging and 3D connectivity mapping from
diffusion tractography.
Usage of such postmortem atlases in conjunction with individual
deterministic tractography of in vivo MRI datasets can serve as
valuable tools to evaluating and
improving patient treatment efficacy.
-
28
References
Aganj, I., Lenglet, C., Sapiro, G., Yacoub, E., Ugurbil, K.,
Harel, N., 2010. Reconstruction of the orientation distribution
function in single and multiple shell q-ball imaging within
constant solid angle, Magnetic Resonance in Medicine 64,
554-566.
Aganj, I., Lenglet, C., Jahanshad, N., Yacoub, E., Harel, N.,
Thompson, P.M., Sapiro, G., 2011. A Hough transform global
probabilistic approach to multiple-subject diffusion MRI
tractography, Medical Image Analysis 15, 414-425.
Barkhoudarian, G., Klochkov, T., Sedrak, M., Frew, A., Gorgulho,
A., Behnke, E., De Salles, A., 2010. A role of diffusion tensor
imaging in movement disorder surgery. Acta Neurochir (Wien) 152,
2089-2095.
Basser, P.J., Pajevic, S., Pierpaoli, C., Duda, J., Aldroubi,
A., 2000. In vivo fiber tractography using DT-MRI data. Magn Reson
Med 44, 625-632.
Benabid, A.L., Pollak, P., Gao, D., Hoffmann, D., Limousin, P.,
Gay, E., Payen, I., Benazzouz, A., 1996. Chronic electrical
stimulation of the ventralis intermedius nucleus of the thalamus as
a treatment of movement disorders. J Neurosurg 84, 203-214.
Bittar, R.G., Kar-Purkayastha, I., Owen, S.L., Bear, R.E.,
Green, A., Wang, S., Aziz, T.Z., 2005. Deep brain stimulation for
pain relief: a meta-analysis. J Clin Neurosci 12, 515-519.
Cho, S.H., Kim, S.H., Choi, B.Y., Cho, S.H., Kang, J.H., Lee,
C.H., Byun, W.M., Jang, S.H., 2007. Motor outcome according to
diffusion tensor tractography findings in the early stage of
intracerebral hemorrhage. Neurosci Lett 421, 142-146.
Coenen, V.A., Allert, N., Madler, B., 2011. A role of diffusion
tensor imaging fiber tracking in deep brain stimulation surgery:
DBS of the dentato-rubro-thalamic tract (drt) for the treatment of
therapy-refractory tremor. Acta Neurochir (Wien) 153, 1579-1585;
discussion 1585.
-
29
De Salles, A.A., Frighetto, L., Behnke, E., Sinha, S., Tseng,
L., Torres, R., Lee, M., Cabatan-Awang, C., Frysinger, R., 2004.
Functional neurosurgery in the MRI environment. Minim Invasive
Neurosurg 47, 284-289.
Deuschl, G., Bain, P., Brin, M., 1998. Consensus statement of
the Movement Disorder Society on Tremor. Ad Hoc Scientific
Committee. Mov Disord 13 Suppl 3, 2-23.
Deuschl, G., Raethjen, J., Hellriegel, H., Elble, R., 2011.
Treatment of patients with essential tremor. Lancet Neurol 10,
148-161.
Duchin, Y., Abosch, A., Yacoob, E., Sapiro, G., Harel, n., 2012.
Feasibility of using ultra-high field (7T) MRI for clinical
surgical targeting, PLoS One.
Elble, R.J., Deuschl, G., 2009. An update on essential tremor.
Curr Neurol Neurosci Rep 9, 273-277.
Flora, E.D., Perera, C.L., Cameron, A.L., Maddern, G.J., 2010.
Deep brain stimulation for essential tremor: a systematic review.
Mov Disord 25, 1550-1559.
Gutman, D.A., Holtzheimer, P.E., Behrens, T.E., Johansen-Berg,
H., Mayberg, H.S., 2009. A tractography analysis of two deep brain
stimulation white matter targets for depression. Biol Psychiatry
65, 276-282.
Habas, C., Cabanis, E.A., 2007. Cortical projection to the human
red nucleus: complementary results with probabilistic tractography
at 3 T. Neuroradiology 49, 777-784.
Hagmann, P., Cammoun, L., Gigandet, X., Gerhard, S., Grant,
P.E., Wedeen, V., Meuli, R., Thiran, J.P., Honey, C.J., Sporns, O.,
2010. MR connectomics: Principles and challenges. J Neurosci
Methods 194, 34-45.
Henderson, J.M., 2012. "Connectomic surgery": diffusion tensor
imaging (DTI) tractography as a targeting modality for surgical
modulation of neural networks. Front Integr Neurosci 6, 15.
-
30
Hyam, J.A., Owen, S.L., Kringelbach, M.L., Jenkinson, N., Stein,
J.F., Green, A.L., Aziz, T.Z., 2012. Contrasting connectivity of
the ventralis intermedius and ventralis oralis posterior nuclei of
the motor thalamus demonstrated by probabilistic tractography.
Neurosurgery 70, 162-169; discussion 169.
Jang, S.H., Bai, D., Son, S.M., Lee, J., Kim, D.S., Sakong, J.,
Kim, D.G., Yang, D.S., 2008. Motor outcome prediction using
diffusion tensor tractography in pontine infarct. Ann Neurol 64,
460-465.
Kim, J., Lenglet, C., Duchin, Y., Sapiro, G., Harel, N., 2014.
Semi-automatic segmentation of brain subcortical structures from
high-field MRI, IEEE Transactions on Information Technology in
Biomedicine.
Kumar, R., Lozano, A.M., Sime, E., Lang, A.E., 2003. Long-term
follow-up of thalamic deep brain stimulation for essential and
parkinsonian tremor. Neurology 61, 1601-1604.
Kwon, H.G., Hong, J.H., Hong, C.P., Lee, D.H., Ahn, S.H., Jang,
S.H., 2011. Dentatorubrothalamic tract in human brain: diffusion
tensor tractography study. Neuroradiology 53, 787-791.
Laxton, A.W., Tang-Wai, D.F., McAndrews, M.P., Zumsteg, D.,
Wennberg, R., Keren, R., Wherrett, J., Naglie, G., Hamani, C.,
Smith, G.S., Lozano, A.M., 2010. A phase I trial of deep brain
stimulation of memory circuits in Alzheimer's disease. Ann Neurol
68, 521-534.
Lee, M.W., De Salles, A.A., Frighetto, L., Torres, R., Behnke,
E., Bronstein, J.M., 2005. Deep brain stimulation in intraoperative
MRI environment - comparison of imaging techniques and electrode
fixation methods. Minim Invasive Neurosurg 48, 1-6.
Lega, B.C., Halpern, C.H., Jaggi, J.L., Baltuch, G.H., 2010.
Deep brain stimulation in the treatment of refractory epilepsy:
update on current data and future directions. Neurobiol Dis 38,
354-360.
-
31
Lehericy, S., Grand, S., Pollak, P., Poupon, F., Le Bas, J.F.,
Limousin, P., Jedynak, P., Marsault, C., Agid, Y., Vidailhet, M.,
2001. Clinical characteristics and topography of lesions in
movement disorders due to thalamic lesions. Neurology 57,
1055-1066.
Lehman, J.F., Greenberg, B.D., McIntyre, C.C., Rasmussen, S.A.,
Haber, S.N., 2011. Rules ventral prefrontal cortical axons use to
reach their targets: implications for diffusion tensor imaging
tractography and deep brain stimulation for psychiatric illness. J
Neurosci 31, 10392-10402.
Lenglet, C., Abosch, A., Yacoub, E., Sapiro, S., Harel, N.,
2012. Comprehensive in vivo mapping of the human basal ganglia and
thalamic connectome in individuals using 7T MRI, PLoS One 7:1.
Lozano, A.M., Giacobbe, P., Hamani, C., Rizvi, S.J., Kennedy,
S.H., Kolivakis, T.T., Debonnel, G., Sadikot, A.F., Lam, R.W.,
Howard, A.K., Ilcewicz-Klimek, M., Honey, C.R., Mayberg, H.S.,
2012. A multicenter pilot study of subcallosal cingulate area deep
brain stimulation for treatment-resistant depression. J Neurosurg
116, 315-322.
Mayberg, H.S., Lozano, A.M., Voon, V., McNeely, H.E.,
Seminowicz, D., Hamani, C., Schwalb, J.M., Kennedy, S.H., 2005.
Deep brain stimulation for treatment-resistant depression. Neuron
45, 651-660.
Mori, S., Crain, B.J., Chacko, V.P., van Zijl, P.C., 1999.
Three-dimensional tracking of axonal projections in the brain by
magnetic resonance imaging. Ann Neurol 45, 265-269.
Poupon, C., Clark, C.A., Frouin, V., Regis, J., Bloch, I., Le
Bihan, D., Mangin, J., 2000. Regularization of diffusion-based
direction maps for the tracking of brain white matter fascicles.
Neuroimage 12, 184-195.
Schuurman, P.R., Bosch, D.A., Merkus, M.P., Speelman, J.D.,
2008. Long-term follow-up of thalamic stimulation versus
thalamotomy for tremor suppression. Mov Disord 23, 1146-1153.
-
32
Sedrak, M., Gorgulho, A., De Salles, A.F., Frew, A., Behnke, E.,
Ishida, W., Klochkov, T., Malkasian, D., 2008. The role of modern
imaging modalities on deep brain stimulation targeting for mental
illness. Acta Neurochir Suppl 101, 3-7.
Yousry, T.A., Schmid, U.D., Alkadhi, H., Schmidt, D., Peraud,
A., Buettner, A., Winkler, P., 1997. Localization of the motor hand
area to a knob on the precentral gyrus. A new landmark. Brain 120 (
Pt 1), 141-157.
-
33
Figure 1. A) The dentatorubrothalamic tract (DRT) was mapped
through three cubic regions of interest (ROIs) in the FLAIR MRI
image set; B) first ROI around the red nucleus in axial view; C)
the second ROI around the precentral gyrus; D) the third ROI around
the ipsilateral dentate nucleus.
-
34
Figure 2. Each contact object was standardized using the Axial
Custom View in Object Creation with brush size 1.5 mm to
consistently model contacts as spheres with volume 0.002 cm.
Contact objects were set as ROIs in the Fiber Tracking section;
this enables tracking of fibers running through each contact to
target brain structures. A) Fibers through contacts in sampled ET
patients closely follow the relevant DRT pathway. B) Close up of
fibers directly passing through contacts 0 and 1 are shown.
-
35
Figure 3. Relevant active contact as determined by
post-operative testing outcomes was established as the sole region
of interest for fiber tracking in each patient FLAIR MRI for 10 ET
patients; visualization of fiber tractography with the DBS
electrodes (red).
-
36
Figure 4. Probabilistic tractography of the DRT from postmortem
diffusion MRI of the brainstem and thalamus. A) A schematic of the
DRT is shown for reference. B) Probabilistic tractography shows the
course of the DRT, including minor branches that pass medially to
the red nucleus (arrowheads). C) Visualization of the spatial
relationships between the DRT, the red nucleus (red), and Vim
nucleus of the thalamus (blue).
-
37
Figure 5. Representative data showing registration between in
vivo and postmortem MRI datasets. A) A parasagittal in vivo image
with a surface rendering of the registered postmortem dataset
superimposed. BD) Sagittal, axial and coronal slices, respectively,
of an in vivo dataset with the corresponding slices from the
registered postmortem anatomic image superimposed. Arrowheads
indicate major borders that demonstrate the accuracy of
registration.
-
38
Figure 6. Visuospatial integration of image data including in
vivo MRI, CT, and registered postmortem data. In each panel, a
single slice of the pre-operative FSPGR image is shown with the
corresponding slice of the registered postmortem dataset
superimposed. Probabilistic tractography of the DRT (orange) is
shown along with surface renderings of the red nuclei (red), the
Vim nuclei of the thalamus (blue), and the implanted DBS leads
derived from post-operative CT data (green). AB) Posterior and
anterior views, respectively, of a coronal slice through the red
nucleus. C) Mid-sagittal slice. D) Oblique view of an axial slice
through the red nucleus.
-
39
Figure 7. Visualization of DBS electrode position relative to
the DRT from the registered postmortem template for all 12 patient
datasets examined in this study. For each patient, we show a single
oblique slice, roughly corresponding to the diagram in the top
left, along with probabilistic tractography of the DRT derived from
the registered postmortem template (orange), and a surface
rendering of the implanted DBS electrodes derived from
post-operative CT data (green).
-
40
Table 1: Stereotaxic coordinates of each electrode contact for
each patient included in the study. Stereotaxic coordinates are
provided in millimeters from the mid-commissural point (MCP).
Distance from the third ventricle (3V) in millimeters is also
included. n/a indicates missing data.
MCP Coordinates (mm) Electrode Contact
Lateral (X) AP (Y) Vertical (Z) Distance
From 3V
Patient 01 Contact 0 12.88 -5.63 -1.98 12.3 Patient 01 Contact 1
13.57 -4.68 0.00 12.0 Patient 01 Contact 2 14.94 -3.30 1.59 13.0
Patient 01 Contact 3 15.07 -2.80 3.97 13.0 Patient 01 Contact 8
11.56 -6.91 0.49 9.5 Patient 01 Contact 9 12.79 -5.93 2.51 10.0
Patient 01 Contact 10 13.04 -4.18 4.63 11.6 Patient 01 Contact 11
13.79 -3.81 6.51 12.6 Patient 02 Contact 0 13.66 -5.79 -0.80 11.1
Patient 02 Contact 1 14.46 -4.99 1.80 11.0 Patient 02 Contact 2
14.66 -5.06 2.61 11.8 Patient 02 Contact 3 -15.54 -3.97 4.30 13.4
Patient 02 Contact 8 15.77 -3.89 1.77 11.6 Patient 02 Contact 9
16.33 -3.08 3.67 14.2 Patient 02 Contact 10 17.13 -2.08 6.07 15.5
Patient 02 Contact 11 17.30 -1.08 7.97 15.7 Patient 03 Contact 0
13.85 -8.09 -3.43 12.1 Patient 03 Contact 1 14.48 -6.76 -1.33 11.5
Patient 03 Contact 2 14.48 -5.94 -0.06 10.2 Patient 03 Contact 3
15.19 -5.15 1.12 11.8 Patient 03 Contact 8 12.68 -8.42 -0.89 8.1
Patient 03 Contact 9 13.13 -7.79 0.56 9.0 Patient 03 Contact 10
13.67 -6.66 1.93 10.6 Patient 03 Contact 11 14.49 -5.66 2.84 10.5
Patient 04 Contact 0 13.37 -8.47 3.14 9.3 Patient 04 Contact 1
13.66 -7.19 5.00 9.9 Patient 04 Contact 2 13.52 -6.33 7.00 11.4
Patient 04 Contact 3 14.66 -5.05 8.85 12.7 Patient 04 Contact 8
15.75 -3.44 2.69 10.7 Patient 04 Contact 9 16.60 -3.21 4.83 12.5
Patient 04 Contact 10 16.60 -2.07 5.97 13.0 Patient 04 Contact 11
18.32 -1.78 7.97 13.8 Patient 05 Contact 0 13.39 -7.39 -1.59 11.1
Patient 05 Contact 1 13.89 -6.13 -0.04 11.0 Patient 05 Contact 2
14.16 -5.02 2.63 10.6 Patient 05 Contact 3 15.05 -3.57 4.63 13.7
Patient 05 Contact 8 12.20 -5.56 -1.30 11.0 Patient 05 Contact 9
13.22 -4.62 0.25 10.6 Patient 05 Contact 10 13.22 -3.17 2.37 12.7
Patient 05 Contact 11 15.43 -1.91 4.37 13.1 Patient 06 Contact 0
10.40 -5.54 -0.27 9.0
-
41
Patient 06 Contact 1 11.35 -5.22 1.18 8.3 Patient 06 Contact 2
11.79 -4.35 3.07 10.6 Patient 06 Contact 3 12.32 -3.17 5.83 11.3
Patient 06 Contact 8 14.19 -4.68 -1.69 12.2 Patient 06 Contact 9
14.09 -4.14 -0.44 12.2 Patient 06 Contact 10 15.12 -2.57 2.32 13.5
Patient 06 Contact 11 16.44 -1.53 4.99 14.1 Patient 07 Contact 0
13.73 -10.95 -2.66 11.7 Patient 07 Contact 1 13.51 -10.73 1.21 10.8
Patient 07 Contact 2 13.68 -9.81 0.79 9.7 Patient 07 Contact 3
14.01 -8.14 3.23 10.0 Patient 07 Contact 8 11.09 -8.79 -1.73 9.4
Patient 07 Contact 9 12.20 -7.75 0.82 8.6 Patient 07 Contact 10
12.12 -6.53 2.16 8.0 Patient 07 Contact 11 12.78 -6.09 4.16 8.9
Patient 08 Contact 0 n/a n/a n/a n/a Patient 08 Contact 1 n/a n/a
n/a n/a Patient 08 Contact 2 n/a n/a n/a n/a Patient 08 Contact 3
n/a n/a n/a n/a Patient 08 Contact 8 n/a n/a n/a n/a Patient 08
Contact 9 n/a n/a n/a n/a Patient 08 Contact 10 n/a n/a n/a n/a
Patient 08 Contact 11 n/a n/a n/a n/a Patient 09 Contact 0 13.32
-8.37 -3.41 11.3 Patient 09 Contact 1 13.83 -6.86 0.70 10.2 Patient
09 Contact 2 14.96 -5.29 2.22 11.4 Patient 09 Contact 3 15.56 -4.16
4.93 12.4 Patient 09 Contact 8 12.59 -8.23 -2.58 10.7 Patient 09
Contact 9 13.16 -7.11 -0.77 10.0 Patient 09 Contact 10 13.72 -5.98
1.15 10.5 Patient 09 Contact 11 14.40 -4.62 3.63 12.2 Patient 10
Contact 0 13.64 -5.82 -3.34 12.0 Patient 10 Contact 1 14.42 -4.91
-0.44 10.6 Patient 10 Contact 2 15.15 -3.60 -1.74 12.2 Patient 10
Contact 3 16.32 -1.94 4.21 13.6 Patient 10 Contact 8 11.96 -6.01
-2.03 10.3 Patient 10 Contact 9 11.81 -5.49 -0.87 9.3 Patient 10
Contact 10 12.74 -4.11 2.03 10.7 Patient 10 Contact 11 14.10 -2.59
2.76 13.3 Patient 11 Contact 0 16.35 -7.32 -5.18 15.1 Patient 11
Contact 1 16.41 -7.09 -4.00 14.9 Patient 11 Contact 2 16.55 -6.38
-2.05 14.5 Patient 11 Contact 3 16.82 -5.42 0.73 14.7 Patient 11
Contact 8 9.49 -8.81 -3.68 8.4 Patient 11 Contact 9 9.86 -7.97
-1.64 8.0 Patient 11 Contact 10 10.23 -7.14 0.29 8.6 Patient 11
Contact 11 10.73 -6.46 2.11 9.5 Patient 12 Contact 0 14.34 -7.58
-1.38 12.4 Patient 12 Contact 1 14.51 -6.91 -0.03 11.0
-
42
Patient 12 Contact 2 14.85 -6.06 2.18 11.7 Patient 12 Contact 3
15.18 -5.04 4.38 12.9 Patient 12 Contact 8 12.17 -9.66 -2.06 10.7
Patient 12 Contact 9 12.66 -8.94 -0.40 9.6 Patient 12 Contact 10
13.38 -7.58 1.97 10.3 Patient 12 Contact 11 14.40 -6.56 3.32
12.0
-
43
Table 2: Demographic information for the 12 patients included in
this study.
Patient Gender Age Patient 01 F 32 Patient 02 F 75 Patient 03 M
74 Patient 04 F 39 Patient 05 F 61 Patient 06 F 68 Patient 07 F 71
Patient 08 F 84 Patient 09 M 72 Patient 10 F 65 Patient 11 M 72
Patient 12 M 69
-
44
Table 3: Treatment efficacy, side effects, and proximity to the
DRT and Vim from the postmortem model for each DBS contact (eight
per patient). + indicates mild/moderate tremor reduction and the
presence of side effects. ++ indicates good/excellent tremor
reduction. - indicates no tremor reduction and the absence of side
effects. n/a denotes missing data points.
Electrode Contact Efficacy Side Effects DRT Proximity Vim
Proximity Patient 01 Contact 0 ++ + 0 mm 0 mm Patient 01 Contact
1 ++ + 0 mm 0 mm Patient 01 Contact 2 + + 0 mm >1 mm Patient 01
Contact 3 + + 0 mm 0 mm Patient 01 Contact 8 ++ + 1 mm Patient 01
Contact 11 + + >1 mm 0 mm Patient 02 Contact 0 ++ + 0 mm 0 mm
Patient 02 Contact 1 ++ + 0 mm 0 mm Patient 02 Contact 2 ++ + >1
mm >1 mm Patient 02 Contact 3 + + 1 mm 0 mm Patient 03 Contact 2
- + >1 mm >1 mm Patient 03 Contact 3 - + >1 mm 1 mm
Patient 04 Contact 3 + + 1 mm Patient 04 Contact 8 ++ - 1 mm 0 mm
Patient 05 Contact 8 ++ - 1 mm >1 mm Patient 06 Contact 0 + -
>1 mm 0 mm
-
45
Patient 06 Contact 1 ++ - 0 mm 0 mm Patient 06 Contact 2 n/a n/a
0 mm >1 mm Patient 06 Contact 3 n/a n/a >1 mm >1 mm
Patient 06 Contact 8 + - 0 mm 1 mm 1 mm 0 mm Patient 07 Contact 1
++ - >1 mm 0 mm Patient 07 Contact 2 n/a n/a >1 mm >1 mm
Patient 07 Contact 3 n/a n/a >1 mm 0 mm Patient 07 Contact 8 ++
- >1 mm 0 mm Patient 07 Contact 9 + - >1 mm 0 mm Patient 07
Contact 10 n/a n/a >1 mm 0 mm Patient 07 Contact 11 >1 mm 0
mm Patient 08 Contact 0 ++ - 0 mm >1 mm Patient 08 Contact 1 ++
- 0 mm 0 mm Patient 08 Contact 2 n/a n/a 1 mm Patient 08 Contact 9
- - >1 mm 0 mm Patient 08 Contact 10 n/a n/a >1 mm 0 mm
Patient 08 Contact 11 n/a n/a >1 mm 0 mm Patient 09 Contact 0 ++
- 1 mm Patient 09 Contact 1 + - 0 mm 1 mm 0 mm Patient 11 Contact 8
++ + >1 mm 0 mm Patient 11 Contact 9 ++ + >1 mm 0 mm Patient
11 Contact 10 n/a n/a >1 mm 0 mm Patient 11 Contact 11 n/a n/a 1
mm 0 mm
-
46
Patient 12 Contact 2 n/a n/a >1 mm 0 mm Patient 12 Contact 3
n/a n/a >1 mm 0 mm Patient 12 Contact 8 - - >1 mm >1 mm
Patient 12 Contact 9 ++ - 0 mm >1 mm Patient 12 Contact 10 n/a
n/a 1 mm 0 mm
-
47
Table 4: Results of Spearman rank correlation. * denotes
statistical significance.
Spearman Correlation Correlation Coefficient p-value Contact
Proximity to Vim vs Side Effects 0.015 0.902 Contact Proximity to
DRT vs Side Effects 0.029 0.815 Contact Proximity to Vim vs
Efficacy 0.043 0.726 Contact Proximity to DRT vs Efficacy 0.336
0.005*