Content uploaded by Jeong-Wook Seo
Author content
All content in this area was uploaded by Jeong-Wook Seo on Jan 10, 2022
Content may be subject to copyright.
Int J Clin Exp Pathol 2020;13(12):3013-3020
www.ijcep.com /ISSN:1936-2625/IJCEP0119612
Original Article
Visualization of left ventricular Purkinje ber
distribution using wideeld optical coherence microscopy
Myung-Jin Cha1, Jeong-Wook Seo2, Hongki Kim3, Minkyu Kim3, Jeonghun Choi3, Duck-Hoon Kang3, Seil Oh1
Departments of 1Internal Medicine, 2Pathology, Seoul National University Hospital, Seoul, South Korea; 3Kohyoung
Technology, Inc, Seoul, South Korea
Received August 5, 2020; Accepted October 31, 2020; Epub December 1, 2020; Published December 15, 2020
Abstract: Background: The distribution and connection of ventricular Purkinje bers are known to be associated
with idiopathic left ventricular arrhythmias. Unusual anatomy is one of the important factors associated with cath-
eter ablation success rate. With the wideeld high-speed, swept-source optical coherence microscopy (OCM) and
light microscope, we visualized the left ventricular Purkinje ber distribution. Methods: Left ventricular walls of ve
adult ovine hearts were incised from the mitral annulus to the apex. Using the wideeld OCM technique and light
microscopy, we observed the distribution, direction, depth, and dividing patterns of the Purkinje network with mul-
tiple tangential angles and without tissue destruction. Results: Wideeld OCM was used to characterize the ovine
heart Purkinje network system in a 4 × 4 mm2 eld. Left ventricular Purkinje bers traveled in the sub-endocardial
area near the left-sided peri-membranous septal area and ran like a wide hair bundle. The distal branching bers
penetrated to the endocardium and connected to the contractile muscle. In this distal area, Purkinje bers were
connected to each other, forming multiple layers. Some Purkinje bers were directly connected within the false ten-
don between the papillary muscles or between the trabeculations. Some free-running Purkinje bers were directly
connected to the papillary muscle from the left bundle. Conclusion: Using wideeld OCM, we were able to observe
the left bundle and its branching patterns in ovine left ventricle without tissue destruction. This might be applied to
future cardiac ablation procedures.
Keywords: Purkinje bers, optical microscopy, heart conduction system, cardiac arrhythmia
Introduction
The His-Purkinje system (HPS) is responsible
for the rapid and efcient electrical conduction
of the ventricles [1]. Cardiac Purkinje cells form
a complex and dense network which cover the
ventricular myocardium [2]. Continuous efforts
have been made to visualize the structural and
functional capacity of the HPS using optical
mapping and dye injections, and the distribu-
tion of the HPS has been largely characterized
due to its results [3, 4]. Historically, techniques
of injecting dye into the sheath or immunoen-
zyme labelling have been widely used, and this
method can show the running distribution of
the ventricular conduction system [5]. However,
observations from light microscopy using vari-
ous staining methods were not sufcient to
visualize the entire system [3]. For instance, the
sheath of the Purkinje bers in humans is
excessively thin and cannot easily tolerate the
increasing perfusion pressure [6]. Imaging te-
chniques such as micro computed tomography
or bioprinting have been tried using a conduc-
tion system, but spatial resolution was limited
[7, 8]. Because the ungulate’s Purkinje bers
are piled up with thick sheaths, visualization is
relatively easier than human HPS. Therefore
the HPS was rst discovered in the sheep heart
[9].
This complex network has been known to be
related to the mechanism of ventricular tach-
yarrhythmias [10, 11]. The anatomic variation
of the HPS and its relationship with papillary
muscles have been associated with increased
incidence of ventricular tachyarrhythmias [12,
13]. Until now, there has been no way to
observe the distribution of the HPS during ca-
theter ablation for ventricular tachyarrhyth-
mias.
Optical coherence microscopy (OCM) is an
echo-based noninvasive non-destructive imag-
Purkinje network visualization
3014 Int J Clin Exp Pathol 2020;13(12):3013-3020
ing modality that measures the time-of-ight of
back-reected light using low-coherence inter-
ferometry [14]. With its high spatial resolution,
OCM is good for observing the HPS traveling in
the subendocardial area of the shallow layer. In
this study, we introduced the novel wideeld
high-speed, swept source OCM technique that
allows easy and precise observation of the
HPS located in the subendocardium. Unlike the
traditional OCM equipment, this new OCM
modality is suitable for observing widely tr-
aveling tissues such as HPS because it can
capture a wide area with high spatial resolution
from multiple different angles without tissue
destruction. In this study, we visualized ovine
HPS with the new OCM modality.
Using this new wideeld OCM modality, we
observed ovine Purkinje distribution and its
anatomic relationship with ventricular papillary
muscles.
Methods
Sheep heart preparation
Sheep hearts were commercially obtained from
the slaughterhouse (Choongju, South Korea).
IACUC (Institutional Animal Care and Use Com-
mittee) approval was not required when all pro-
cedures are being carried out on tissues ob-
tained from a slaughterhouse, The researchers
did not participate in the process of animal
slaughter.
Five frozen adult sheep hearts were thawed in
the laboratory for more than 12 hours at room
temperature of 25°C. Thawed hearts were xed
in formalin solution for 48 h after gross ana-
tomic evaluation. The lateral wall of the left
ventricle was linearly incised from the mitral
annulus to the apex so that the left ventricular
endocardial wall was exposed between two
papillary muscles. Then, the middle of the ante-
rior mitral leaet and the aortic annulus area
between the left and non-coronary cusps of
the aortic valve was incised so that the left-sid-
ed membranous septum was fully exposed
(Supplementary Figure 1). Left ventricle surfac-
es were divided into 14 regions based on tradi-
tional 16 segments. Each segment was num-
bered sequentially from the posteromedial
side. The left ventricular wall was cut out
with 3 mm depth and observed by OCM
(Supplementary Figure 2).
Optical coherence microscopy
A schematic of the built-in-house OCM system
is shown in Figure 1. Light from a wavelength-
Figure 1. OCM system. This new optical coherence microscopy had lateral resolution of 7.8 μm and axial resolution
of 11.4 μm in tissue. Field of view (FOV) from galvanometer scanning was 4 mm × 4 mm and wide FOV images up to
20 mm × 20 mm were generated by stitching and reconstructing the multiple galvanometer scanning FOV images
acquired by moving the two-axis translation stage. FC, ber coupler. C, circulator. PC, polarization controller. CL, col-
limation lens. GM, Galvanometer. RM, reference mirror. FL, focusing lens. OL, objective lens.
Purkinje network visualization
3015 Int J Clin Exp Pathol 2020;13(12):3013-3020
swept source (HSL-20, Santec, repetition rate =
100 kHz; central wavelength = 1310 nm; band-
width = 100 nm) was divided by a ber coupler
into reference and sample. Reected light from
the reference mirror was recombined with
reected light from the sample at the ber
coupler, and nally the interference signal
was collected by a balanced photo detector
(PDB470C, Thorlabs) which converts the light
into an analogue electrical signal. Then the
analogue electrical signal was converted to
digital signal by a digitizer (APX-5050, AVA-
LDATA). An objective lens (LSM03, Thorlabs; NA
= 0.036) was used to focus the light on the
specimen. The OCM system had lateral resolu-
tion of 7.8 μm and axial resolution of 11.4 μm
in tissue. The sensitivity of the system is mea-
sured to be 105 dB. Field of view acquired by
galvanometer scanning was 4 mm × 4 mm.
Wide eld of view images up to 20 mm × 20
mm were generated by stitching the multiple
galvanometer scanning elds of view images,
which could give a fast measurement time of
about 12 seconds but a small measurement
area limited by the small scanning angle of the
galvanometer. To make a larger measurement
area up to 20 mm × 20 mm, the wideeld OCM
method was used by moving the motorized
two-axis translation sample stage. A total of
25 moving steps of the sample stage could
achieve 20 mm × 20 mm measurement area.
Supplementary Figure 3 shows photography of
a frontal view and modules of the OCM system.
Since OCM measures intensity of the scatter-
ing signal from the tissue, in the OCM image,
white color corresponds to the highest scatter-
ing signal and black color to the lowest scatter-
ing signal from the tissue.
Light microscopic histology
The distribution of Purkinje bers was obse-
rved by light microscopy after sectioning the
myocardial tissue for OCM. The myocardial tis-
sue was embedded en face such that the sur-
face of the septal wall was sectioned rst and
then step sections were made at 4 μm intervals
from the surface to the depth. Unstained sec-
tions were examined and then some of them
were selected and stained by hematoxylin and
eosin. Additional stains of Masson’s trichrome
and reticulin were applied to reveal histologic
details. Histologic sections were scanned by
Aperio CS2 (Leica Biosystems Nussloch, Ger-
many) and examined by Aperio Imagescope
(leica Biosystems Nussloch, Germany). Semi-
macroscopic distribution of Purkinje bers and
their histologic details were analyzed and com-
pared with the OCM data.
Results
Gross anatomy of sheep hearts
The gross anatomy of the sheep heart was
basically similar to that of the human heart
although the inferior vena cava was longer and
the aorta branched to the left and right. The
two groups of papillary muscles were leaning to
the wall without sub-papillary intertrabecular
spaces. There was a muscular wall continuing
to the subaortic part of the left ventricle
between the right coronary cusp and the non-
coronary cusp where a membranous septum
was expected in a human heart. The whitish
sheath traveled widely and was divided into
several complicated divisions. Some of the
narrow sheath bundles were free-running over
the surface, not attaching the subendocardi-
um, and met directly with the papillary muscle.
Most bundles ran on the endomyocardial sur-
face and traversed the crest of the trabecula-
tion, and the valleys had no whitish membra-
nous sheaths.
Optical coherence microscopic imaging of
Purkinje system
By comparing images from OCM and standard
light microscopy, the whitish sheaths were
conrmed to be enclosing sheaths of Purkinje
bers (Figure 2). With OCM, Purkinje bers were
seen as low density and reticular sheath
was observed as high density. Although the
nucleus was not denitely distinguishable,
OCM was enough to observe the distribution
and running direction of HPS.
Serial images were obtained in 5 μm incre-
ments from the surface without destruction of
the tissue. The serial images of one tissue
block are presented in Figure 3. As the depth
of observation becomes deeper, the eld
becomes darker as the transmission of light
decreases. HPS was located within 20 μm
(base)~50 μm (apex) from the surface. The
sites to be observed could be observed at dif-
ferent angle. Supplementary Figure 4 is an
OCM image of the same site taken at different
angles.
Purkinje network visualization
3016 Int J Clin Exp Pathol 2020;13(12):3013-3020
Figure 2. OCM and conventional light microscopic observations of HPS in the same spot. A. OCM image. Purkinje
bers can be seen as low density and reticular sheath is observed as high density. It is suitable for observing the
running direction of bers in a wide area. B. Light microscope image. Purkinje bers stained by hematoxylin and
eosin have fewer myobrils compared to other nearby myocardial tissues, have glycogen around the nucleus, and
are larger and longer than cardiac myocytes.
Figure 3. Serial optical coherence microscopic images. With optical coherence microscopy, the region of interest
can be observed at various desired depths without tissue destruction.
Purkinje network visualization
3017 Int J Clin Exp Pathol 2020;13(12):3013-3020
Subendocardial distribution of Purkinje bers
The left bundle, which was separated from the
His bundle, runs in the apical direction like a
waterfall emerging from the myocardial peri-
membranous septal wall (Figure 4A). The di-
rection of this bundle is different from that of
working myocardium in the endocardial area
(Supplementary Figure 5). In the bundle (Figure
4B), each ber splits into a thinner fascicle
according to its destination (Supplementary
Figure 6). Some of them are free-running and
lead directly to papillary muscles (Figure 5).
The Purkinje bers, which travel to the distal
free wall or apex, move away from the base,
and then either reverse direction or reach the
destination and connect with myocardium.
Figure 4. OCM images of left bundle branch and Purkinje bundles. A. Left bundle branch. The left bundle runs in the
apical direction like a waterfall emerging from the myocardial peri-membranous septal wall. B. Bundle of Purkinje
bers. The direction of this bundle is different from that of the working myocardium in the endocardial area. Each
ber splits into a thinner fascicle according to its destination.
Figure 5. Free-running Purkinje bers. A. Gross ndings. B. OCM ndings. C. Histology. Some Purkinje bundles are
free running without attaching with ventricular surface or directly connect between trabeculae and the papillary
muscle. The Purkinje bers are seen at most of the false tendon at the apical part of the septum. Only Purkinje cells
without cardiac myocytes are in these false tendons.
Purkinje network visualization
3018 Int J Clin Exp Pathol 2020;13(12):3013-3020
Besides running in the subendocardial area
piled on the sheath, free running Purkinje
bers could be found within the false tendon
connecting tissues. This enables efcient
transmission of electrical signals to distant
areas by using a fast-running free running struc-
ture without being affected by cardiac contrac-
tion. Especially, the Purkinje bers inside the
false tendon connected to the papillary mus-
cles can be observed.
Observing the apical area connected to myo-
cardium showed a more fractal appearance
than the connected structure of the proximal
part (Supplementary Figure 7). This might be
because the Purkinje bers in the distal area
end in connection with the multiple working
cardiomyocytes. Figure 6 is a schematic dia-
gram of left ventricular Purkinje system we
could observe with OCM.
Histologic ndings of the Purkinje system from
light microscopy
The Purkinje bers form an interlacing network
from the beginning until their transition into
the myocardial cells. They are surrounded by
loose connective and fatty tissue at the proxi-
mal end but lose their covering sheath when
transitioning into myocardial cells.
The Purkinje system starts at the level of the
His bundle and is divided into the anterior, mid-
dle, and posterior fascicles at the basal sep-
tum. The largest bundles become medium- or
small-sized bundles on their way to the mid- or
apical septum. At the basal septum, Purkinje
cells are seen as a large group of bers. while
they gradually become medium- to small-sized
by the mid and apical septum.
The cardiac Purkinje cells run on the crest of
the trabeculae, but the basal part of the groove
between trabeculae is free of the Purkinje cells
or less populated. Free-standing bundles of
Purkinje bers were observed at the false ten-
don as seen by OCM (Figure 5C). The most dis-
tal Purkinje cells merged with the myocardial
cells predominantly at the mid- and apical
myocardium but this transition was also found
at the basal part.
Discussion
In this study, we could observe the running
distribution and direction of Purkinje network
system in the left ventricle using a wideeld
OCM technology. The ovine left ventricular
Purkinje bers were 1) traveling in the suben-
docardial area, 2) some free running bers
were directly connected to papillary muscles
from left bundle branch, and 3) distally, Purkinje
bers become shorter and branched, forming
multiple layers.
Although it is difcult to apply this study direc-
tly to the analysis of mechanism of human
arrhythmia,s, we were able to get new informa-
tion and ideas using the new imaging tech-
nique. Free-running bers have the advantage
of delivering electrical signal from the top to a
remote location in the shortest time. Previous
gross anatomy studies identied free-running
bers in 55% to 62% of human hearts, of wh-
ich 60% carried conduction tissue bers [15].
Previous studies have also reported connec-
tions between papillary muscles through false
tendons, although Purkinje bers inside the
false tendon were not visualized [16]. In rela-
tion to the mechanism of arrhythmia, a circuit
connecting the distal and proximal parts may
be formed, or the abnormal electrical activity
of myocardium may be easily transferred to the
proximal. Although the papillary muscle plays
an important role in the reentrant mechanism
of ventricular arrhythmias, the connection
Figure 6. Schematic diagram of left ventricular Pur-
kinje system. A. The left bundle branch is divided into
two or four major fascicles. B. Main fascicles are di-
vided into branches, but some branches meet again.
C. Some bers in the main fascicle are separated
from the subendocardial area and connect directly to
the papillary muscle. D. There are some free-running
bers connecting branched fascicles. E. Free-running
bers also branch or meet each other.
Purkinje network visualization
3019 Int J Clin Exp Pathol 2020;13(12):3013-3020
between the papillary muscles and the elec-
trical conduction system is not well known [17].
The association of papillary muscle with free-
running Purkinje bers is important inform-
ation to understand idiopathic left ventricular
arrhythmias [18]. In the swine heart model, left
ventricular papillary muscles were related to
fatal ventricular arrhythmias [19]. In addition,
arrhythmia occurring in the papillary muscle
and arrhythmia associated with the fascicle
are difcult to evaluate because clinical symp-
toms are not signicantly different [20].
Recently, Rivera et al. reported a muscular
connection between papillary muscles, which
are related to the electrocardiographic pattern
in papillary muscle related ventricular tachyar-
rhythmias [12]. To ablate the focal area of fas-
cicular reentrant or Purkinje ber-related ven-
tricular tachycardia, precise anatomic and
electrical mapping is important [21].
We investigated the possibility of using a
new wideeld OCM for observation of the ovine
HPS. The advantage of the wideeld OCM is
that 1) it can be visualized at tangential multi-
sections and 3-dimensional reconstruction
without tissue destruction, 2) it is suitable to
observe a wide range of Purkinje bers in one
microscopic screen, and 3) it has a high spatial
resolution. Although the resolution of OCM is
inferior compared to the standard light micro-
scope in resolution, wideeld OCM is able to
observe Purkinje bers and free running bers
without tissue destruction, which might be
important for use in catheter-based proce-
dures.
In order to observe the tissue with a conven-
tional optical microscope, the cutting direction
of interest for slicing the tissue should be de-
termined before microscopic observation.
Once cut, the tissue cannot be observed again
in the other direction. However, with OCM, we
could observe the site of interest without
tissue destruction. The HPS was especially suit-
ed to observe with OCM because it spreads
widely in the very shallow layer from the sur-
face. Previously, there were efforts to observe
heart structures using OCM technique. Also,
OCM is already used in clinical examinations for
coronary atherosclerosis because of its high
spatial resolution and its ability to observe tis-
sue in real time in vivo [22]. The ability to apply
OCM technology to catheter-based imaging is
also a great advantage [22]. With advance-
ments in OCM technology, in-vivo catheter-
based observation of the HPS may be
possible.
There are several considerations to be taken
into account when interpreting this study. The
conguration of the Purkinje network can be
structurally different between animal species
[23, 24]. The observations in sheep cannot be
directly applied to humans, as there may be
inconsistencies in distribution of Purkinje -
bers. Second, OCM cannot clarify the morphol-
ogy of the membranous septum, because OCM
had lateral resolution of 7.8 μm and axial reso-
lution of 11.4 μm in tissue.
Acknowledgements
This work was supported by both a National
Research Foundation of Korea grant funded
by the Korean government (No. NRF-2020-
R1A2C1013832) and a grant KHRS 2018-2
from the Korean Heart Rhythm Society.
Disclosure of conict of interest
None.
Abbreviations
HPS, His-Purkinje system; OCM, optical coher-
ence microscopy.
Address correspondence to: Dr. Seil Oh, Internal
Medicine, Seoul National University College of
Medicine and Seoul National University Hospital,
101 Daehak-ro, Jongno-gu, Seoul 03080, Korea.
E-mail: seil@snu.ac.kr
References
[1] Boyden PA. Purkinje physiology and patho-
physiology. J Interv Card Electrophysiol 2018;
52: 255-262.
[2] Duan D, Yu S, Cui Y and Li C. Morphological
study of the atrioventricular conduction sys-
tem and Purkinje bers in yak. J Morphol 2017;
278: 975-986.
[3] De Almeida MC, Lopes F, Fontes P, Barra F,
Guimaraes R and Vilhena V. Ungulates heart
model: a study of the Purkinje network using
India ink injection, transparent specimens and
computer tomography. Anat Sci Int 2015; 90:
240-250.
[4] Xu Y, Peng Y, Qu R, Zheng G, Feng F, Feng Y, He
L, He S, Li Z, Liu C, Xiao Z, Ouyang J and Dai J.
Purkinje network visualization
3020 Int J Clin Exp Pathol 2020;13(12):3013-3020
Locating the human cardiac conduction sys-
tem using a 3D model of its nutritious arteries.
Sci Rep 2017; 7: 344.
[5] Atkinson A, Inada S, Li J, Tellez JO, Yanni J,
Sleiman R, Allah EA, Anderson RH, Zhang H,
Boyett MR and Dobrzynski H. Anatomical and
molecular mapping of the left and right ven-
tricular His-Purkinje conduction networks. J
Mol Cell Cardiol 2011; 51: 689-701.
[6] Davies F. The conducting system of the verte-
brate heart. Br Heart J 1942; 4: 66-76.
[7] Stephenson RS, Boyett MR, Hart G, Nikolaidou
T, Cai X, Corno AF, Alphonso N, Jeffery N and
Jarvis JC. Contrast enhanced micro-computed
tomography resolves the 3-dimensional mor-
phology of the cardiac conduction system in
mammalian hearts. PLoS One 2012; 7:
e35299.
[8] Tracy EP, Gettler BC, Zakhari JS, Schwartz RJ,
Williams SK and Birla RK. 3D bioprinting the
cardiac purkinje system using human adipo-
genic mesenchymal stem cell derived Purkinje
cells. Cardiovasc Eng Technol 2020; 11: 587-
604.
[9] Tawara S. Das reizleitungssystem de saugeti-
erherzens: eine anatomich-hisologische studie
uber das atrioventricularbundel und die pur-
kinjeschen faden. Jena, Germany: Verlag von
Gustav Fischer; 1906.
[10] Haissaguerre M, Vigmond E, Stuyvers B, Hocini
M and Bernus O. Ventricular arrhythmias and
the His-Purkinje system. Nat Rev Cardiol 2016;
13: 155-166.
[11] Wilde AAM, Garan H and Boyden PA. Role of
the Purkinje system in heritable arrhythmias.
Heart Rhythm 2019; 16: 1121-1126.
[12] Rivera S, Vecchio N and Ricapito P. Anatomical
connections between the papillary muscles
and the ventricular myocardium. Circ Arrhythm
Electrophysiol 2019; 12: e007004.
[13] Nogami A. Purkinje-related arrhythmias part I:
monomorphic ventricular tachycardias. Pacing
Clin Electrophysiol 2011; 34: 624-650.
[14] Huang D, Swanson EA, Lin CP, Schuman JS,
Stinson WG, Chang W, Hee MR, Flotte T, Greg-
ory K, Puliato CA, et al. Optical coherence to-
mography. Science 1991; 254: 1178-1181.
[15] Loukas M, Louis RG Jr, Black B, Pham D, Fu-
dalej M and Sharkees M. False tendons: an
endoscopic cadaveric approach. Clin Anat
2007; 20: 163-169.
[16] Komatsu Y, Nogami A, Kurosaki K, Morishima I,
Masuda K, Ozawa T, Kaneshiro T, Hanaki Y,
Shinoda Y, Talib AK, Kowase S, Sekiguchi Y
and Aonuma K. Fascicular ventricular tachy-
cardia originating from papillary muscles: Pur-
kinje network involvement in the reentrant cir-
cuit. Circ Arrhythm Electrophysiol 2017; 10:
e004549.
[17] Yamada T, McElderry HT, Okada T, Murakami
Y, Doppalapudi H, Yoshida N, Allred JD, Muro-
hara T and Kay GN. Idiopathic focal ventricular
arrhythmias originating from the anterior papil-
lary muscle in the left ventricle. J Cardiovasc
Electrophysiol 2009; 20: 866-872.
[18] Nogami A. Purkinje-related arrhythmias part ii:
polymorphic ventricular tachycardia and ven-
tricular brillation. Pacing Clin Electrophysiol
2011; 34: 1034-1049.
[19] Kim YH, Xie F, Yashima M, Wu TJ, Valderrabano
M, Lee MH, Ohara T, Voroshilovsky O, Doshi
RN, Fishbein MC, Qu Z, Garnkel A, Weiss JN,
Karagueuzian HS and Chen PS. Role of papil-
lary muscle in the generation and mainte-
nance of reentry during ventricular tachycardia
and brillation in isolated swine right ventricle.
Circulation 1999; 100: 1450-1459.
[20] Good E, Desjardins B, Jongnarangsin K, Oral H,
Chugh A, Ebinger M, Pelosi F, Morady F and
Bogun F. Ventricular arrhythmias originating
from a papillary muscle in patients without
prior infarction: a comparison with fascicular
arrhythmias. Heart Rhythm 2008; 5: 1530-
1537.
[21] Takase S, Mukai Y, Honda N, Chishaki A, Sad-
amatsu K and Tsutsui H. Catheter ablation for
verapamil-sensitive fascicular ventricular ta-
chycardia guided by precise mapping using a
multi-spline duodecapolar catheter with small
electrodes and close spacings. HeartRhythm
Case Rep 2019; 5: 48-52.
[22] Ambrosi CM, Moazami N, Rollins AM and E-
mov IR. Virtual histology of the human heart
using optical coherence tomography. J Biomed
Opt 2009; 14: 054002.
[23] Ono N, Yamaguchi T, Ishikawa H, Arakawa M,
Takahashi N, Saikawa T and Shimada T. Mor-
phological varieties of the Purkinje ber net-
work in mammalian hearts, as revealed by
light and electron microscopy. Arch Histol Cytol
2009; 72: 139-149.
[24] Yoshimura A, Yamaguchi T, Kawazato H, Taka-
hashi N and Shimada T. Immuno-histochemis-
try and three-dimensional architecture of the
intermediate laments in Purkinje cells in
mammalian hearts. Med Mol Morphol 2014;
47: 233-239.
Purkinje network visualization
1
Supplementary Figure 1. Sheep heart tissue preparation for microscopic observation. The lateral free wall of the
left ventricle between two papillary muscles was incised to expose the anterior mitral leaet. The whole part of
left ventricular wall was cut out to a 3 mm depth and observed with both new optical coherence microscopy and
conventional light microscope with hematoxylin and eosin staining. APM, Anterolateral papillary muscle. PPM, Pos-
terolateral papillary muscle. LCC, left coronary cusp.
Purkinje network visualization
2
Supplementary Figure 2. Myocardial segmentation for microscopic observation. The ovine left ventricular endocar-
dial area was divided into 14 regions based on traditional 16 segments. Two incised walls (basal and mid anterolat-
eral walls) were excluded from the analysis.
Supplementary Figure 3. Frontal view and modules of the OCM system.
Purkinje network visualization
3
Supplementary Figure 4. Multi-angular optical coherence microscopic images. With optical coherence microscopy,
the region of interest can be observed at different angles.
Supplementary Figure 5. The direction of Purkinje bers and working myocardium bundles. With optical coherence
microscopy, the direction can be followed relatively easily. The direction of two different bundles running through
different angles on ventricular surface.
Purkinje network visualization
4
Supplementary Figure 6. Left bundle branch. In the left bundle, the Purkinje bers are already separated into thin
bundles according to their nal destination.
Purkinje network visualization
5
Supplementary Figure 7. Purkinje ber distribution at apical area. The Purkinje bers are fragmented and bifurcated
and turned retrogradely at distal area.