TECHNICAL NOTE

Satoru Shimizu, M.D. ANATOMIC DISSECTION AND CLASSIC THREE-

Department of Neurosurgery,
Kitasato University
School of Medicine,
Kanagawa, Japan
Ryusui Tanaka, M.D.
Department of Neurosurgery,
Kitasato University
School of Medicine,
Kanagawa, Japan
Albert L. Rhoton, Jr., M.D.
Department of Neurological Surgery,
University of Florida,
Gainesville, Florida
Yutaka Fukushima, M.D.
Department of Neurological Surgery,
University of Florida,
Gainesville, Florida
Shigeyuki Osawa, M.D.
Department of Neurosurgery,
Kitasato University
School of Medicine,
Kanagawa, Japan
DIMENSIONAL DOCUMENTATION: A UNIT OF
EDUCATION FOR NEUROSURGICAL ANATOMY REVISITED
OBJECTIVE AND IMPORTANCE: Despite development of computer-assisted neuro-
surgical navigation, learning by dissecting anatomic specimens is still important.
CLINICAL PRESENTATION: We describe the processes from preparation of specimens
for cranial dissection to documentation of three-dimensional (3-D) stereoscopic pic-
tures, particularly focusing on the latter, which has been initiated in the Microneuro-
anatomy Laboratory, Department of Neurological Surgery, University of Florida.
INTERVENTION: Preparation consists of irrigation of the major vessels and injection
of colored silicone. The 3-D documentation, obtaining two pictures corresponding to
each eye’s view, is obtained by the shoot-shift-shoot method using a single camera
mounted on a slide bar. The key of this method is correct shifting of the camera without
alignment error to get exact 3-D effects. Observation of 3-D image can be made with
free viewing, a 3-D viewer, or projection. Tips concerning all of the processes involved
are described.
CONCLUSION: The presented method of dissection and obtaining 3-D images is
beneficial for accomplishing studies of anatomy and for providing teaching method.
KEY WORDS: Anatomic dissection, Education, Three-dimensional imaging

Neurosurgery 58:1000-1002, 2006 DOI: 10.1227/01.NEU.0000210247.37628.43 www.neurosurgery-online.com

Masatou Kawashima, M.D., Ph.D.
Department of Neurosurgery,
Kitasato University

School of Medicine,
Kanagawa, Japan
Hidehiro Oka, M.D.
Department of Neurosurgery,
Kitasato University
School of Medicine,
Kanagawa, Japan
Kiyotaka Fujii, M.D.
Department of Neurosurgery,
Kitasato University
School of Medicine,
Kanagawa, Japan
Reprint requests:
Satoru Shimizu, M.D.,
Department of Neurosurgery,
Kitasato University
School of Medicine,
1-15-1 Kitasato, Sagamihara,
Kanagawa,
228-8555 Japan.
Email: Satoru4756@aol.com
Received, October 3, 2005.
Accepted, January 3, 2006.
C TECHNIQUE
urrently, technologies of surgical nav-
igation using virtual reality are avail-
able in the neurosurgical field, but
knowledge of surgical anatomy continues to
be essential, as is learning the anatomic
structures and simulating surgical proce-
dures using cadavers. Combinations of lec-
tures using three-dimensional (3-D) stereo-
scopic presentation and anatomic dissection
facilitate understanding of spatial relation-
ships between complex structures. Educa-
tional effects have been experienced by a
large number of participants in recent mi-
crosurgical dissection courses held by the
senior author (ALR) (4). In this report, we
present details for conventional processes,
from preparation of specimens to generation
of 3-D stereoscopic documentation, initiated
at the Microneuroanatomy Laboratory, De-
partment of Neurological Surgery, Univer-
sity of Florida. Descriptions of techniques
are for those applied in 2003 when the first
author (SS) worked in the laboratory.
Silicone Injection and Preservation of
Specimens
Details of this process have been well-
documented previously (5), so only important
tips are mentioned here:
1) A cadaver, usually fixed in formalde-
hyde, is prepared, and the head is separated at
the neck to provide good isolation of blood
vessels for cannulation. When one needs to
study a lower cervical level, separation is
made like a wedge in the chest if the neck is
large and thick.
2) Dissect and isolate the following major
blood vessels for infusion: jugular veins, ca-
rotid arteries, and vertebral arteries. Most
specimens will have the common carotid ar-
tery, but, occasionally, the artery is cut just
below the carotid bifurcation. In such cases,
emplacement of silastic tubing should be con-
ducted for both internal and external carotid
arteries separately to obtain sufficient fixation.

E1000 | VOLUME 58 | NUMBER 5 | MAY 2006 www.neurosurgery-online.com

 NEUROANATOMY
For vertebral arteries, the transverse process of the vertebra
may be removed to achieve a longer length of the vertebral
artery stump if observation of the lowest vertebra is not
planned. Excessive dissection results in migration of the sili-
con into the carotid sheath or paravertebral space and should
be avoided because such contamination interferes with subse-
quent dissection and observation of the neck.
3) Insert appropriately sized bevel-cut Tygon silastic tubing
(Fisher Scientific, Pittsburgh, PA) into each isolated blood
vessel as far as possible (Fig. 1A).
4) Tie the vessels’ stump with 2–0 silk suture. The tightness
of the ligation must be enough to both fix the tubes and
maintain patency of the tubing.
5) Connect disposable polypropylene Fisher pipette tips
(Fisher Scientific) to the silastic tubing for each blood vessel
and clamp the end of the tubing using polypropylene tube
FIGURE 1. A, stump of neck. Silastic tubing is put in each isolated blood
vessel and tied. In this specimen, tubing has been inserted into both inter-
nal and external carotid arteries because of insufficient length of common
carotid artery. B, injection of colored silicone mixture. All tubes except for
injection are clamped with polypropylene clamps.
NEUROSURGERY
AND THREE-DIMENSIONAL DOCUMENTATION
clamps (Fisher Scientific). The connector tips are cut to the size
of each silastic tube.
6) Flush all cannulated vessels repeatedly with tap water
with a 60-ml syringe to remove clots and insure smooth flow.
Low resistance to injection and immediate drainage of the
water injected through other vessels suggests that the speci-
men has few clots so that satisfactory injection of silicone can
be expected.
7) Ligate or clamp the stumps of the vessels that excessively
leak irrigation water.
8) After repeated irrigation, clamp each silastic tube with
plastic clamps to keep the water inside and preserve the
specimen in a container with 75% ethyl alcohol. The above
irrigation process should be repeated until no visible clots
appear, this usually being performed three times a day for 3
days until completion.
9) Prepare silicone (3110 RTV silicone rubber; Dow Corning,
Midland, MI) and thinner (polydimethylsiloxane: 200 fluid 20
CST; Dow Corning) mixtures at the following volume ratios of
the two materials: for arteries, thinner 2:silicone 1; for veins,
thinner 1:silicone 1
10) Add pigment (Crayola Powder Paint; Binney & Smith,
Easton, PA) of the desired color: red or blue. A higher con-
centration of pigment provides the vasculature with vivid
color. Approximate concentration of the pigment is an amount
that is enough to eliminate the transparency of the cup.
11) When ready, add catalyst (Dow Corning) for each solu-
tion immediately before the injection (1/3–1/2 tube for 100–
150 ml of solution). More catalyst means faster curing of the
silicone.
12) Inject the colored silicone manually using a 60-ml sy-
ringe according to the following guidelines: for arteries, (com-
mon) carotid, 20 ml each ⫻ 2 ⫽ 40 ml; vertebral, 10 ml each ⫻
2 ⫽ 20 ml; and for veins, jugular 50 ml each ⫻ 2 ⫽ 100 ml.
Begin with injection into the vessel that has the best patency
(Fig. 1B), arteries first (carotid and then vertebral) followed by
veins. Clamp vessels after each injection. Injections should be
conducted under constant pressure. For narrower arteries, this
may be difficult, but pressure should be maintained for a
sufficient period to perfuse tiny branches. When resistance
disappears suddenly during injection, stop the procedure im-
mediately because rupture of the vessel can be concluded. To
prevent migration of the leaked silicone through the vessel
sections into the spinal canal, packing with soft paper may be
used.
13) Store the specimen in a plastic bag in a bucket in an
inverted position for 48 hours without ethanol soaking be-
cause it interferes with the silicone curing process. Hardening
of silicone left in cups can be used to indicate degree of curing
in the injected specimen.
14) Cut each silastic tubing at the point of entry into the
vessels after confirmation of the curing process. Do not pull
out the tubing in such a way as to disturb the injected silicone
in its correct position.
Hereafter, preserve the specimen in a container filled with
75% ethyl alcohol. Ethyl alcohol is an appropriate solution to
VOLUME 58 | NUMBER 5 | MAY 2006 | E1000
SHIMIZU ET AL.

preserve the specimen without it becoming hard while avoid- before taking any pictures. Important check points are as
ing an offensive odor, which is a major disadvantage of form- follows: stability (soft flexible tissues, such as nerves, muscles,
aldehyde. or fat are readily moved by gravity or their own elasticity);
presence of unnecessary tissues or dust; and moistness of the
Dissection of Specimens specimen, including utensils (dissectors, spatulas, background
Put the specimen on an observation stand in the desired cloths). Excessively soaked surfaces and drops of water make
position to perform dissections. Small rubber pads may be focus dull or cause unwanted reflection of flash lights, so
useful to secure irregularly shaped objects. Specimens under removal with suckers or pellets of cotton is recommended.
heated microscope illumination should be kept moist during 3) After loading a 35-mm format film cartridge (Fujichrome
dissection by using a spray filled with water or ethyl alcohol. Sensia II, ISO 100/21), set the camera. We use a Nikon F3
Membranous tissues may easily become dry and change color (Nikon, Garden City, NY) with a 105 mm 1:2.8D Micro-Nikkor
(e.g., the color of the dura mater changes from grayish white lens (Nikon) on a slide bar mounted on a tripod. Choose the
to khaki on drying). Water can remain on the surface longer film carefully because expression of colors may differ some-
than ethyl alcohol because it evaporates less rapidly. what between film manufacturers. The slide bar is used to
Dissection is made of the area of interest as in anatomic obtain pictures from two viewpoints corresponding to each
demonstration or actual surgery. Use sharp scalpels or scissors eye, the so-called stereo pair. A firm tripod without any loose-
to obtain smooth cuts, especially for parenchymal tissues of ness is ideal to obtain accurate stereo pairs. The rail of the slide
the brain and spinal cord. bar is put parallel to the object because excessive differences in
When vascular wall injury occurs in the process of dissec- the distance to the object between two viewpoints results in
tion, do not manipulate the area roughly or the colored sili- objects of different size in individual pictures.
cone will extrude outside of the vessel lumen. Silicone dye, 4) The vertical mounting must be adapted on the sliding bar
once extruded, is difficult to return to the vessel. The venous for vertical pictures. Ensure that the screws of the slide bar are
wall is easily lacerated, so in-
jured vascular wall may be
removed. However, remem- TABLE 1. Appropriate flash settings (underlined) with a 105-mm lens for each distance
ber that silicone casts are Distancea Flash settings
sometimes insufficient to 1:2–2.1 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
make forms of the veins 1:2.3–2.5 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
when clots are present. In 1:2.5–2.7 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
step-by-step dissection using 1:3.5 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
photo documentation, the 1:3.8 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
next step should be post- 1:3.9 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
poned until confirmation of 1:4.9 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
the quality of the pictures is 1:7 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
made. Once resected, tissue a
Expressed as magnification ratios indicated in the lens.
cannot be put back in place.

Taking 3-D Pictures

The 3-D equipment pre-
sented can be purchased TABLE 2. Appropriate flash settings (underlined) with a 105-mm lens and teleconverter in each
from Reel 3-D Enterprises, distance
Inc. (Culver City, CA), acces- Distancea Flash settings
sible via their web site (ste- 1:1.15 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
reoscopy.com), except for 1:1.2 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
notes, for which the reader 1:1.8 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
interested in principles of 1:2.1 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
3-D presentation is referred 1:2.2 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
to Ferwerda (1). 1:2.5 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
1) Cover the background of 1:3 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
the specimen fully with 1:3.1 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
soaked, dark-colored cloths 1:3.9 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
(usually blue or black) to ob- 1:5 Full OO 1/2 OO 1/4 OO 1/8 OO 1/16 OO 1/32 OO 1/64
tain vivid outlines in pictures. a
Expressed as magnification ratios indicated in the teleconverter.
2) Make a final check of the
appearance of the specimen

E1000 | VOLUME 58 | NUMBER 5 | MAY 2006 www.neurosurgery-online.com

 NEUROANATOMY
tight enough to allow for both stability and mobility for
swinging. The lens and shutter speed should be adjusted to
f32 (the smallest of lens aperture openings) and 1/60 seconds,
respectively, these providing sharp outlines for both shallow
and deep sites of the object. Use teleconverters (Nikon Tele-
converter TC-201 2X, Nikon) for small objects.
5) Illuminate the object with a microscope or other electric
lamp to help identify structures and adjust the focus in the
viewfinder.
6) Decide the layout in the viewfinder by choosing reference
points to ensure the same layout after moving to the opposite
side. The decision as to layout should be made with examina-
tion from both viewpoints before taking the picture to avoid
unexpected differences, especially in the background and in-
cluding unexpected objects. The reference circle, margins, and
angles of the viewfinder are useful for this purpose. When no
identifiable structures can be found as reference points, small
objects may be put beside or on the specimen as markers and
removed after making decisions as to layout. Take care not to
move the specimen and surrounding items in this process (i.e.,
spatula and its holder, cloths). Careful adjustment of align-
ment is most important to get exact 3-D images because align-
ment errors (which introduce discrepancy at the center of the
object between viewpoints, especially vertical discrepancy and
torsion) results in poor 3-D images.
7) Adjust the focus for that part of the object that is of the
most interest and bear in mind that it is difficult to obtain
sharp focus for different levels equally in a specimen with
much depth.
8) Select the amount of light for the flash (we use a Sunpak
Thyristor auto 544) with 19 steps, full, 1/2, 1/4, 1/8, 1/16,
1/32, 1/64, and two steps among those amounts. Appropriate
amounts of light to provide correct exposure at different dis-
tances are listed in Tables 1 and 2. For close views, the amount
of light should be diminished to avoid halation. In general,
lower intensities are more appropriate for dry skulls to avoid
excessive whitening caused by reflection. At times, trial and
error is needed to obtain pictures of sufficient quality. The
usual direction of the flash is parallel to the axis of the lens, but
modification of the angle may be necessary for creating
greater contrast in the region of interest. Shadows can enhance
a sense of perspective. The same direction of shadow is
needed for both stereo pairs. During repeat of shutter release
and film advance, occasional confirmation of the viewfinder
position is recommended to preclude movement of the cam-
era.
9) After taking pictures on one side, slide the camera to the
opposite side along the slide bar and adjust the reference point
to be the same as in the first picture (shoot-shift-shoot method)
(Fig. 2). Shift of the camera may change the distance to the
object, but adjustment of the focus is not appropriate because
it changes the size of the object in the picture, and thus the
picture does not work as a shadow contour of the first picture.
An example of a stereo pair obtained with this method is
shown in Figure 3. The distance between the two viewpoints
has to be within normal human interpupillary distance, 65
NEUROSURGERY
AND THREE-DIMENSIONAL DOCUMENTATION
FIGURE 2. Schema of method for taking 3-D pictures using single cam-
era, shoot-shift-shoot method. Spatial relationships between position of
camera and object in viewfinder depicted. After taking pictures on the
right side (top), the camera is moved to the left-side view point using a
slide bar (center) and is swung to adjust the reference point to match the
first picture (bottom). In this sample, a cortical branch of the middle cere-
bral artery in the center of reference circle is the reference point. Crosshair
in left row shows the center of the viewfinder. Thus, the picture on the
left is ready to take.
mm on average (1). A longer distance between the two view-
points makes for deeper 3-D images. Objects situated proximal
to distal to the camera, such as spatulas or dissectors, may be
perceptible as double contours when the sliding distance is
longer than that which allows for sensory fusion. Once the
taking pictures has begun, the specimen has to be kept in its
initial position. Do not touch anything except the film advance
lever, shutter release switch, and flash.
Film Development
In the process of development, the images must be correctly
mounted on slide holders. This process is as important as
taking pictures to provide accurate and clear 3-D images.
Observation of 3-D Images
Three methods of observation of the 3-D image are avail-
able. We mostly use the latter two described below. Although
the free viewing method requires no special optical aid, it
requires experience, and not everyone can perform stereopsis
(1). Two approaches of free viewing are available. In parallel-
eyed viewing, the stereo pair, situated on an illuminated stand
for sorting and viewing of slides, is held directly in front of an
observer who looks with the left eye at the left slide and with
the right eye at the right side. In cross-eyed viewing, the
observer has to view the sample cross-eyed, that is, viewing
the slide on the right with the left eye and that on the left with
the right eye. In both methods, the observer can see three
VOLUME 58 | NUMBER 5 | MAY 2006 | E1000
SHIMIZU ET AL.
FIGURE 3. A stereo pair, left view (A) and right view (B), with right
pterional approach. These pictures appear almost the same, but differ as
shown by the angle of the spatulas and the overlapping of structures
images next to one another. Attention should only be paid to
that in the center, the stereoscopic image.
Using a 3-D viewer is an easier way of viewing than free
viewing during individual use. Optical viewers, such as the
Pinsharp Adjustable Twin 35 mm Slide Viewer (Fig. 4A),
which has two independent slide chambers for each stereo
pair and which can be illuminated by room light through a
semitransparent membrane, can be used. The 3-D images can
FIGURE 4. Equipment needed for 3-D observation. 3-D viewer (A) for
personal use and a 3-D glass, cardboard frame type (B), a silver screen
(C), and two slide projectors for each stereo pair with polarizing lens fil-
ters (horizontal type, inset, D) for projection.
E1000 | VOLUME 58 | NUMBER 5 | MAY 2006
(between the olfactory and optic nerves, the oculomotor nerve, and the
superficial sylvian vein). The center (reference point) is the internal
carotid artery, and no alignment error is evident with this pair.
be viewed through binocular lenses. The slides for both view-
points are inserted into the corresponding chambers of the
viewer to give visual perception to both eyes individually.
When the slides corresponding to the viewpoints are in-
serted in reverse (i.e., the picture for the right/left view point
is put into the left/right chamber), the 3-D picture will appear
inside out, a strange image contrary with the common sense of
anatomy.
The screen projection method is suitable to make presenta-
tions for a number of observers (1). Two slide projectors with
polarizing lens filters of a horizontal type (Cokin, Rungis,
France) for each stereo pair projection, a screen painted with
aluminum (silver screen), and 3-D glasses are needed (Fig. 4,
B–D). The 3-D glasses we use have gray polarized lenses and
should not be confused with anaglyph glass (red/blue, red/
cyan, or red/green). Polarizing filters have a molecular struc-
ture that is similar to an invisible picket fence, with all of the
pickets going in one direction, which coordinates the direction
of vibration of light. A ray of light, a straight line, vibrates in
all directions at right angles to its line of transmission. Polar-
ization (i.e., causing the vibration to be in a single direction)
can be achieved by transmitting light through the filter. The
silver screen keeps the reflected light polarized, and the sec-
ond filter (3-D glass) with the polarizing direction at exactly
right angles to the first polarized light provides darkness to
the corresponding eye. Thus, extinguishing effects of the con-
tralateral views in each eye provide stereopsis (Fig. 5A).
Set up the slide projectors, one of them corresponding to the
right side and the other to the left side view, vertically or
horizontally on a stable base (Fig. 4D). Pairs of slides for the
same picture are set in each projector and projected to confirm
screen adjustment. Usually, we use so-called target slides for
this purpose, adjusting the focus of each slide and then the
lines and circles horizontally and vertically to get exact over-
lap. Great care should be taken in this adjustment process
because the projection method may be more sensitive to align-
www.neurosurgery-online.com
 NEUROANATOMY AND THREE-DIMENSIONAL DOCUMENTATION

FIGURE 6. Examples of appropriate (A) and inappropriate (B) projection

of 3-D pictures on a screen. Views observed without 3-D glasses are simu-
lated. In the former, contours of each view are arranged horizontally,
whereas the latter has alignment errors, with tortuous arrangement of con-
tours. Alignment errors can be caused in the taking of 3-D pictures, inap-
propriate mounting on slide holders, or insufficient adjustment of projec-
tors.

ment errors between the pictures than with the 3-D viewer
method, and thus spectators may experience the 3-D picture as
not focused (Fig. 6).
In practice, one should put on 3-D glasses and turn off one
projector, rotating the polarizing lens filter in one viewpoint’s
projector to mask the view on the other side (e.g., the slide on
the left side has to be masked for the right eye view) (Fig. 5, B
and C). The same maneuver is then performed for the other
side.
After the setting of 3-D projection for observation through
3-D glasses, spectators should be situated in front of the screen
or as close as possible to the front of the screen for receiving
FIGURE 5. A, principles of 3-D projection. A ray of light, the image of a
slide, vibrates in all directions at right angles to its line of transmission.
Polarization (i.e., making vibration with a single direction) can be
achieved by transmitting light through polarizing filter (bidirectional
arrow on filter shows direction of polarization). Silver screen keeps
reflected light polarized, and second filter (lenses in 3-D glass) with polar-
izing direction at exact right angles to first polarized light provides dark-
ness (dotted arrows) to corresponding eye. Extinguishing effects of con-
tralateral views in each eye for stereopsis can be made when observing
polarized light coordinated by polarizing filters arranged in the V-position
through glasses with polarizing direction arranged in the same position
(polarization direction for left filter is from top left to bottom right and for
right filter from top right to bottom left, both directions form an angle of
45 degrees with a vertical line, the standard fashion). B and C, adjustment
of direction of polarizing filter. After adjustment of alignment, turn off
projector on one side while putting on 3-D glasses. When polarizing direc-
tion defined by polarizing filter is not at exact right angles to direction in
3-D glass, light can be identified, so rotate filter until masking view on
opposite side (B). This adjustment of filter on left side masks right eye’s
view (C). Insets show observation of a slide through 3-D glass and polar-
izing filter. Slide is visible when polarization directions between glass and
filter are not at right angles (B). After rotation of filter to make right
angle between two polarization directions, slide is not visible, and a
masked vision has been made (C). Do the same maneuver on other side
and finally project both views simultaneously. Thus, 3-D perception is
available.

NEUROSURGERY VOLUME 58 | NUMBER 5 | MAY 2006 | E1000

SHIMIZU ET AL.
polarized light correctly. This ideal position is termed the
ortho-stereo position. Observation of correct 3-D images is
otherwise difficult. Although there is no definite angle for the
ortho-stereo position, an approximate value is within 60 de-
grees of the screen.
DISCUSSION
Stereopsis (i.e., depth perception) is based on binocular
parallax, a principle that human eyes are placed some distance
apart and thus see different images. Stereoscopic photography
using this principle was invented less than 10 years after the
invention of photography, around 1838, and became a form of
home entertainment by the early 20th century (1).
Modalities for 3-D presentation, such as free or 3-D optical
viewing, combination of polarized images and 3-D glasses,
and anaglyphic 3-D stereoscopic printing (2), have also been
applied to the medical field (1, 2). The anaglyphic method is
convenient after once completing printing, but observation
with red and blue filter glasses to extinguish superimposed
red and blue dyed images may interfere with identification of
original colors. Thus, we prefer the conventional 3-D observa-
tion approach using original stereo pair slides.
The presented consecutive methods of facilitating the teach-
ing of neurosurgical anatomy have been initiated by previous
workers serving in the Microneuroanatomy Laboratory, De-
partment of Neurological Surgery, University of Florida, and
the anatomic dissection and pictures obtained using the
present method have provided anatomic details for various
regions (3). However, it is a fact that the method is laborious,
and it takes time to accomplish a series of dissection and 3-D
documentation (including the period for mastering the 3-D
technique), whereas recently developed computer assisted
3-D educational tools can readily show the anatomy of inter-
est. According to the authors’ experience, mastering of taking
3-D pictures requires much practice, and completion of a
series of a dissection may take weeks (of course the duration
depends on the meticulousness of each dissection). Despite
these disadvantageous aspects for busy clinicians, we believe
that dissection of anatomic specimens is a beneficial experi-
ence, and the educational effect is superior to other teaching
materials. Furthermore, this type of dissection provides not
only knowledge of basic anatomy but also stimulates ideas to
develop new approaches, including application of new tech-
nologies. As well, 3-D pictures become valuable properties
that allow sharing of the virtual experience of the dissection.
We hope that this paper will encourage learning of neurosur-
gical anatomy from such dissection and will facilitate effective
3-D presentations so that a larger number of persons who
work in neurosurgery may become familiar with this ap-
proach.
REFERENCES
1. Ferwerda JG: The World of 3-D. Borger, 3-D Book Productions, 2003, ed 2, pp
210–221.
E1000 | VOLUME 58 | NUMBER 5 | MAY 2006
2. Ribas GC, Bento RF, Rodrigues AJ Jr: Anaglyphic three-dimensional stereo-
scopic printing: Revival of an old method for anatomical and surgical teach-
ing and reporting. Technical note. J Neurosurg 95:1057–1066, 2001.
3. Rhoton AL Jr: Cranial Anatomy and Surgical Approaches. Schaumburg, Lip-
pincott Williams & Wilkins, 2003.
4. Rhoton AL Jr: Operative techniques and instrumentation for neurosurgery.
Neurosurgery 53:907–934, 2003.
5. Sanan A, Aziz KM, Janjua RM, van Loveren HR, Keller JT: Colored silicone
injection for use in neurosurgical dissections: Anatomic technical note.
Neurosurgery 45:1267–1271, 1999.
Acknowledgments
We thank Ronald Smith, M.S., and other research fellows working at the
Microneuroanatomy Laboratory, Department of Neurological Surgery, Univer-
sity of Florida for their constant support and teaching. Drs. Shimizu, Tanaka,
Kawashima, and Fujii have served as research fellows in the Microneuro-
anatomy Laboratory in the Department of Neurological Surgery at the Univer-
sity of Florida.
COMMENTS
I n this well-written report, Shimizu et al. provide a detailed descrip-
tion of the steps required for the production of three-dimensional
(3D) stereoscopic pictures of anatomic specimens. The significant
teaching potential of such 3D stereoscopic anatomic demonstrations
has been well-documented by Dr. Rhoton during his highly recog-
nized lectures that have influenced neurosurgical anatomic education.
We support the concept of further emphasis on cadaveric anatomic
microsurgical dissections as an integral model during neurosurgical
residency training as well as in post-training in continued medical-
neurosurgical education.
Saleem I. Abdulrauf
St. Louis, Missouri
M. Gazi Yaşargil
Little Rock, Arkansas
I n 1995, we initiated the acquisition of 3D stereoscopic images at the
University of São Paulo Medical School for the purpose of neuro-
surgical anatomy teaching, initially performed through the projection
of stereopairs of slides. In addition to the use of this method in regular
courses for medical students and in many courses organized in neu-
rosurgical meetings and centers, we also had the opportunity to begin
to give regular lectures on neurosurgical anatomy to the residents of
the Department of Neurosurgery of the University of Virginia on a
yearly basis beginning in 1997. Also in 1997, during a neurosurgical
Brazilian meeting in Salvador, Bahia, we had the opportunity to show
this method to Dr. Rhoton and soon we had the satisfaction of starting
to see his beautiful presentations in stereoscopic 3D. In December
2001, we reported the method of making stereoscopic 3D prints for
neurosurgical anatomy documentation through the old anaglyphic
technique (2), which still constitutes the simplest way of making 3D
prints. Since 2002, we have been making our 3D stereoscopic presen-
tations digitally.
The article by Shimizu et al. is a very important publication because
it explains, in fine detail, both the preparation and dissection of
specimen techniques and the stereoscopic 3D technique that is so
helpful for neurosurgical anatomy teaching. In regard to the technique
of shooting the pictures to obtain the stereopair, we prefer only to
slide the camera along the sliding bar instead of swinging the camera
toward a reference point, as done by the authors, because this shoot-
shift-shoot method causes a convergence that generates the so-called
www.neurosurgery-online.com
 NEUROANATOMY
keystone effect, as emphasized by Ferwerda (1) and is also cited by
Shimizu et al. With this technique, when the camera is turned a
fraction, the previous rectangular image will be transformed into a
new trapezium or another oblique quadrangle (1), which characterizes
the so-called keystone effect that causes a distorted final stereoscopic
image, a distortion that is minimized in close-ups, but is enhanced
with mid-distance shooting. The final stereoscopic images of a cra-
nium pictured along its anteroposterior axis by this technique, for
example, will make the cranium appear elongated, as if it were dis-
closing a sagittal synostosis. On the other hand, this technique allows
both images to have final similar actual sizes, which eases their
mounting process. The sliding bar technique generates two final slides
with different left and right sizes, which should then be cropped
during the mounting process so only the common images of the slides
are projected.
The possibility of dealing with digital images avoids all the diffi-
culties and limitations with slide mounting and optimizes image
manipulation and projection with a similar, but digital, set-up. The
digital dual projection is performed from a notebook computer with
the aid of software that sends each image of the stereopair to a
different projector, through two different computer video outputs.
Guilherme C. Ribas
São Paulo, Brazil
NEUROSURGERY
AND THREE-DIMENSIONAL DOCUMENTATION
1. Ferwerda JG: The World of 3D: A Practical Guide to Stereo Photography, Borger,
3D Book Productions, 1990.
2. Ribas GC, Bento RF, Rodrigues AJ Jr: Anaglyphic three-dimensional stereo-
scopic printing: Revival of an old method for anatomical and surgical teach-
ing and reporting. J Neurosurg 95:1057–1066, 2001.
T his is a very detailed description of the techniques of preparation
of specimens for cranial dissection and of taking 3D stereoscopic
pictures. In fact, I wonder how many neurosurgeons are actually
involved in these processes and how many in the world have the
practical facilities to reproduce these techniques. I know of only a few
places where people can prepare cadavers in as good conditions as Dr.
Rhoton’s team does. There is a group of neurosurgeons in the western
part of France from Angers, Rennes, and Tours. There is another one
in Alicante, Spain, where I used to organize a course on vertebral
artery surgery. In general, in each place, techniques of cadaver prep-
aration are particular and are considered a secret. At least they can
now compare their techniques with Dr. Rhoton’s. His beautiful prep-
arations and presentations can certainly be used as references for
other teams.
Bernard George
Paris, France
VOLUME 58 | NUMBER 5 | MAY 2006 | E1000