Content uploaded by William Holmes
Author content
All content in this area was uploaded by William Holmes on Jun 27, 2018
Content may be subject to copyright.
Measures of Cardiac Function in Theraphosidae Spiders using in vivo
1
Magnetic Resonance Imaging
2
3
Authors
4
Gavin D. Merrifield* (1)
5
Nichola M. Brydges (2)
6
Lynsey Hall (3)
7
James Mullin (1)
8
Lindsay Gallagher (1)
9
Romain Pizzi (4, 5)
10
William M. Holmes (1)
11
Addresses:
12
1. Glasgow Experimental Magnetic Resonance Imaging Centre, College of
13
Medical, Veterinary and Life Sciences, University of Glasgow, Glasgow, UK
14
2. Institute of Psychological Medicine and Clinical Neurosciences, Cardiff
15
University School of Medicine, Cardiff, UK
16
3. University of Newcastle, Newcastle, UK
17
4. Royal Zoological Society of Scotland, UK
18
5. School of Veterinary Medicine and Science, University of Nottingham, UK
19
20
*Corresponding Author
21
g.d.merrifield@gmail.com
22
23
Keywords
24
MRI, cardiac function, Theraphosidae, tarantula, ejection fraction, in vivo, magnetic
25
resonance, heart rate
26
27
28
29
30
Abstract
31
32
We present the first in vivo cardiac Magnetic Resonance Imaging (MRI)
33
measurements of Theraphosidae spiders. MRI scanning was performed on six
34
spiders under isoflurane-induced anaesthesia. Retrospective Self-Gating Cine-
35
Cardiac MRI (RG-CINE-MRI) was used to overcome the difficulties of prospective
36
cardiac gating in this species. The resulting RG-CINE-MRI images were successfully
37
analysed to obtain functional cardiac parameters from live spiders at rest. Cardiac
38
ejection fraction was found to increase with animal mass (Pearson correlation 0.849,
39
p= 0.03) at a faster rate than myocardial tissue volume, while heart rate stayed
40
constant across animals. Suggesting the spider heart undergoes additional
41
biomechanical loading with age. The acquisition of these results demonstrates the
42
potential for retrospective gating to evaluate aspects of cardiac function in a wide
43
range of previously inaccessible species.
44
45
46
Introduction
47
48
To date, the cardiac physiology of invertebrates in general, and spiders in particular,
49
has been comparatively little studied next to the great volume of cardiac literature
50
amassed for rodents and humans, particularly in the medical sciences. Additionally
51
for spiders the techniques used have been restricted to measurements of heart
52
action potential via electrocardiogram (ECG) (Dunlop et al.,1992) or monitoring
53
exterior cuticle movement as a proxy for cardiac motion (Bromhall, 1987; Coelho and
54
Amaya 2000) . These latter observations have included both visual observation and
55
the use of attached magnets and sensors to gauge movement. Many of these
56
methods involve either attaching experimental apparatus to the animal, penetrating
57
the outer cuticle (which could potentially lead to lowering of the internal
58
hydrodynamic pressure) or indirect observations of heart function. Therefore, there
59
is little quantitative information on spider cardiac function and outputs in the
60
literature.
61
62
This is in contrast to many vertebrates where cardiovascular Magnetic Resonance
63
Imaging (MRI) routinely provides a non-invasive method of assessing both function
64
and structure in live subjects. MRI is widely used for clinical assessment of
65
cardiovascular disease in humans, providing measurements such as myocardial
66
mass, ventricular volumes, stroke volume, ejection fraction (Epstein, 2007) and
67
even quantitative myocardial perfusion (Jerosch-Herold, 2010) and blood flow maps
68
(Markl M, 2007) . Applications of cardiovascular MRI to other vertebrates, such as
69
rodents, are predominately focussed on biomedical research using disease models.
70
This research has driven the development of specialised MRI systems optimised for
71
mice and rats. The data needed for an MR image generally needs to be acquired
72
over several separate acquisitions. For cardiac MRI where the heart is in continual
73
motion, the MRI scanner typically needs to be triggered/gated by an ECG signal
74
identifying the phase of the cardiac cycle. This is called prospective gating.
75
76
Although MRI has been used previously to acquire basic MRI images of spider
77
anatomy (Pohlmann et al., 2007) it has never been used to study the cardiac
78
function of spiders. Indeed, conventional prospective gating with its need for ECG
79
electrodes and respiration probes is not practical for spiders due to the difficulty of
80
attaching electrode pads or needles. However, the recent development of
81
retrospectively gated cardiac cine-MRI (RG-CINE-MRI) (Heijman et al., 2006) has
82
simplified cardiac MRI experiments by removing the need for ECG and respiratory
83
contact probes (Holmes 2008, 2009). Here we seek to demonstrate this technology
84
to acquire the first in vivo cardiac images of a spider heart and obtain quantitative
85
measures of cardiac function. The Theraphosidae family and Grammostola genus
86
were chosen as subjects, being a good fit in terms of physical size for existing
87
commercial rat cardiac MRI coils. In addition, possessing hearts of comparable size
88
to rodent hearts helps produce good quality MRI images for quantitative analysis.
89
However it should be noted that the cardiac method described here would be equally
90
applicable to smaller spiders, if an appropriately sized MRI coil was used (Merrifield
91
et al 2017) . With the advent of coil-on-a-chip technology, the size of MRI coils now
92
ranges from several tens of centimetres down to just 50microns diameter (Webb
93
2013).
94
95
RG-CINE-MRI
96
Materials and Method
97
98
Animal Ethics
99
100
Experiments were conducted according to UK legislation (UK Animals (Scientific
101
Procedures) Act (1986)). Subjects were anaesthetised throughout scanning for
102
immobilisation and to reduce potential subject stress. Efforts were made to avoid
103
direct handling of subjects.
104
105
Animal Details and Housing
106
107
Six captive bred adult spiders (gender undetermined, exact age unknown but of adult
108
size, n = 4 Grammostola rosea (Walcknenaer), n = 2 Grammostola porteri (Mello-
109
Leitao), Mean Mass = 15.7 g ± 1.75) were obtained from a UK-based supplier
110
(Virginia Cheeseman, High Wycombe, UK). Spiders were individually housed in
111
plastic vivariums (Length = 29 cm, Width = 19 cm, Height = 23 cm). Sterilised
112
vermiculite substrate was provided (~4 cm deep) along with a retreat. A one week
113
acclimatisation period followed delivery of spiders before subsequent scanning. Free
114
access to water was provided. Food was withheld until after full recovery from
115
scanning and anaesthesia.
116
117
Animal Anaesthesia, Handling and Positioning
118
119
Following standard procedures in animal MRI research, subjects were anaesthetised
120
using 5% isoflurane delivered in a 30%/70% mixture of O2/N2O gas (1000 ml min-1)
121
to minimise subject motion derived imaging artefacts (Fig 1B). Under anaesthesia all
122
measurements would also be taken at a physiological baseline, eliminating variation
123
due to involuntary movement or behaviour.
124
125
Subjects were positioned in an MRI compatible animal cradle, lying supine with the
126
heart close to the MRI coil (Fig. 1C). Restraints were cushioned by folded medical
127
gauze swabs that were placed along the length and width of the spider. The
128
assembly was then enclosed in a sealed clear plastic chamber to allow maintenance
129
of anaesthesia and for visual observation of the subject (Fig. 1D). This was then
130
placed into the MRI scanner.
131
132
After scanning was complete (>1 hour) subjects were in an unresponsive state
133
suggesting deep anaesthesia had been achieved. Locator scans performed before
134
and after cardiac scanning confirmed animals had remained in position during the
135
scanning procedure. Activity returned to normal over a subsequent 24-48 hour
136
period. After recovery no adverse health effects or anomalous behaviours were
137
noted.
138
139
MRI Scanning
140
141
Animals were scanned in a 7T Bruker Biospec MRI scanner (Figure 1A) equipped
142
with a 400 mTm-1 gradient insert and 4-channel phased array cardiac coil (Rapid
143
Biomedical GmbH, Germany). Ambient room temperature during scanning was 18-
144
21ºC. The animal was not directly heated by any additional equipment in this time.
145
146
Anatomical MRI scans were obtained to set up imaging slice prescriptions for Fast
147
Low Angle Shot (FLASH) based retrospective RG-CINE-MRI scans (repetition time
148
TR = 8.00ms, echo time TE = 3.30ms, field of view (FOV) = 30.0mm x 30.0mm,
149
matrix = 256 x 256, in-plane resolution 117μm x117μm, slice tyhickness = 1.50mm,
150
14-18 slices depending on size of individual spider, 300 continuous k-space
151
acquisitions, 5mins 7secs imaging time per slice). These were obtained using an in-
152
slice navigator echo as part of the Intragate software package on the scanner
153
(Paravision v.4, Bruker).This navigator is used retrospectively to determine the
154
phase of the cardiac cycle associated with each k space acquisition, allowing images
155
to be created for 10 different phases of the cardiac cycle [Heijman et al., 2006;
156
Bovens SM et al., 2011].Axial image slices along the length of the heart were
157
obtained sequentially until the whole heart was scanned marked by the distal and
158
proximal aortas.
159
160
Image Reconstruction and Analysis
161
162
Cardiac images were reconstructed using the software tools available in Bruker's
163
Intragate software. Residual navigator pulse trace discontinuities were excluded.
164
Heart rates for each acquired slice were individually outputted as part of this
165
process. Images were analysed using Image J (Schneider et al., 2012). Three
166
independent researchers were trained in cardiac image analysis and then each
167
conducted a separate analysis of all images. Researchers were blinded to which
168
subjects the images came from.
169
170
For each slice, images were acquired for 10 different phases of the cardiac cycle.
171
From the 10 images of the cardiac cycle for the central chamber, the images
172
corresponding to the diastolic and systolic phases were identified. For each slice at
173
diastole and systole, a region of interest was manually drawn around the heart
174
perimeter giving the area of the ventricle at diastole and systole. This area is then
175
converted to a volume for that slice by multiplying by the image slice thickness of
176
1.5mm. The total ventricular volume of the heart, the end diastolic volume (EDV) and
177
end systolic volume (ESV) are then given by summing the volumes from each slice.
178
The cardiac ejection fraction (EJ) was then determined by,
179
180
𝐸𝐽 = (𝐸𝐷𝑉 − 𝐸𝑆𝑉)
𝐸𝐷𝑉
. Calculations of global heart parameters were then made from these measurements.
181
182
183
184
Statistical Analysis
185
186
Where group averages are given they are presented with plus/minus the standard
187
deviation. In Figure 3a the error bars represent the range of values of the
188
measurement made by the three independent researchers. Researchers were
189
blinded to which spider the images had come from and slice ordering was
190
randomised for each researcher. From the RG-CINE-MRI navigator signal a heart
191
rate is measured for each of the (14 to 18) slices acquired. The heart rate was then
192
presented as the mean and standard deviation of these values (figure 3b).
193
Correlations were performed using Pearson correlation coefficient and a 2-tailed test
194
of significance (OriginPro 8, OriginLab Corporation).
195
196
197
198
Results
199
200
Cardiac Anatomy
201
202
Scans revealed anatomy matching that broadly outlined for spiders in existing
203
literature (Paul et al. 1994, Foelix 1996, Huckstorf K 2013). Figures 2a and 2b show
204
typical image slices acquired from the RG-CINE-MRI scans. Figure 2c shows a set
205
of ostia in the open position as the heart chamber fills with blood. Figure 2d shows
206
the same slice but with the heart now filled and the ostia closed. Blood pooling
207
between the myocardium and pericardium prior to injection to the interior of the
208
myocardium was also visible.
209
210
Heart Rate
211
212
Resting heart rates were measured via the RG-CINE-MRI technique as outputted by
213
navigator signal data. Multiple samplings on each spider were yielded by measuring
214
the heart rate in each image slice. The mean heart rate for the group of spiders was
215
20 bpm ± 2 and showed good consistency between subjects using this method (Fig.
216
3B). No significant correlation was found between body mass and heart rate
217
(Pearson correlation -0.256, p=0.62).
218
219
Cardiac function
220
221
Table 1 shows quantitative measurements of cardiac function derived from the RG-
222
CINE-MRI datasets for each spider. As may be expected there was a significant
223
correlation between body mass and total heart volume (Pearson correlation 0.882,
224
p=0.02) and between body mass and end diastolic ventricular volume (EDV)
225
(Pearson correlation 0.956, p= 0.006). The fraction of blood ejected from the heart
226
with each heart beat (the cardiac ejection fraction (EJ)) was successfully measured
227
in each spider as a measure of cardiac function (Fig. 3A). These are the first
228
measurements of in vivo ejection fraction in spiders that we were able to find in the
229
literature. Interestingly, we find a significant correlation between body mass and EJ
230
(Pearson correlation 0.849, p= 0.03). The difference in measurement of EJ between
231
the three researchers gave a mean observer difference of EJ = 5.9%. This mirrors
232
accepted levels of observer variability in corresponding rodent cardiac MRI (Heijman
233
et al., 2008). The motion of the heart over the full cardiac cycle at different positions
234
along the axis of the heart can clearly be seen in Supplementary Information 1-3.
235
236
Discussion
237
238
The spider heart differs in many structural and biochemical aspects to those of
239
vertebrates previously studied with MRI. However, the spider’s contiguous nested
240
two chamber system - the heart myocardium surrounded by a pericardium (Paul et
241
al., 1994; Huckstorf et al., 2013) - provides a resultant MRI image similar to that of a
242
standard human/rodent short-axis view (Figure 2). Hence we used similar image
243
analysis methods to quantify cardiac function.
244
245
246
As might be expected we found a significant positive correlation between body mass
247
and both heart volume and ventricular volume. However, we also found a significant
248
positive correlation between body mass and ejection fraction. This is interesting as
249
in humans there is no significant correlation between body mass and ejection
250
fraction (Dorbala S. et al., 2006; Seo J.et al., 2017), but there is some correlation
251
between age and ejection fraction (Gebhard C. et al., 2012). A possible explanation
252
for this is that as spiders grow by iterative stages of moulting, a spider’s mass can
253
serve as an approximate proxy for age (Foelix 1996). Therefore the correlation
254
observed between EJ and mass can also be broadly considered to be one of EJ and
255
age.
256
257
While the cardiac ventricular volume and to a lesser extent the total heart volume
258
changes between diastolic and systolic phases of the cardiac cycle as expected, the
259
total volume of the myocardial tissue remains largely unchanged in each spider and
260
between systolic and diastolic phases of the cardiac cycle. This suggests that the
261
material is being expanded and compressed across the cardiac cycle as would be
262
expected. No obvious trend for increase in myocardial material over animal mass
263
was observed in contrast to the trend observed for ventricular ejection fraction.
264
265
It can be speculated that if the growth of the heart volume and/or myocardial tissue
266
from one moulting instar to the next is less than that of the overall animal's volume
267
then a corresponding increase in EJ would be required. This means that the heart
268
would have to pump a larger volume of blood as it aged in order to maintain
269
adequate cardiac function, just as we observed. Further assessment of spiders of
270
different masses (and so at different ages or developmental instars at least) would
271
provide a more complete understanding of this relationship as well as potentially
272
revealing any additional physiological costs of this increased output as subjects age.
273
These costs might mirror mechanical heart degradation and disease in vertebrate
274
animals but could be of novel interest to cardiac researchers given the evolutionary
275
independent nature of the spider cardiac system as an invertebrate. Additionally, if
276
measurements of the mechanical properties of the spider heart could be conducted
277
ex vivo, then it would be possible to combine these with MRI to calculate the
278
biomechanical forces at work in the spider heart in vivo. Spider growth gradually
279
tapers off both physically and with frequency dependent on species, age and food
280
availability (Foelix 1996). It is conceivable that the end point to this process might be
281
marked or triggered by the condition of the heart, precluding further growth when
282
specific mechanical limits have been reached. The MRI technique we have
283
demonstrated here would be suitable for investigating this further.
284
285
The consistency of mean heart rate across subjects supports the use of non-contact
286
methods to assess heart function in spiders. Although some reports suggest that
287
heart rate across all spider species is correlated with animal mass (Carrel and
288
Heathcote 1976) this was not found to be the case for the subjects involved in our
289
study. However, the Carrel and Heathcote study treats multiple spider types in a
290
collective fashion, combining measurements from them all into a single trend for
291
heart rate and mass. This is despite known differences between these spiders in
292
terms of behaviour, environmental conditioning and physiology. This variation
293
between types is visible in the Carrel and Heathcote study itself. Our study suggests
294
that within groupings of spider type heart rate does not vary - certainly within the
295
Theraphosidae species. Comparative study of different types of spiders using MRI
296
would clarify this situation further.
297
298
A wider range of heart rate values was observed in two subjects in this study. These
299
two animals were scanned first. This greater range of obtained heart rates could be
300
explained by a poorer quality Intragate navigator signal resulting from increased
301
animal motion. In turn this could be due to less restrictive restraints used on these
302
initial two subjects as we finalised the experimental set up. However, mean values
303
are still acceptably consistent with those found in the subsequent scanned subjects.
304
305
We were unable to absolutely determine the gender of the subjects scanned, but
306
their large size (indicating greater age) and continued life-span post-scanning (years)
307
suggests they were all female. This would also tally with the assessment of the
308
supplier. It is possible that our group of subjects included a mix of both male and
309
female subjects which may have resulted in some of the variation in results.
310
However, sexing of spiders is notoriously difficult until males reach sexual maturity in
311
the final instar of growth and so may continue to be difficult to determine.
312
313
Despite reports of the effectiveness of isoflurane-based inhalation anaesthesia
314
induction in spiders (Zachariah et al., 2009; Dombrowski et al., 2013) we found it to
315
have variable performance (Pizzi 2012). Some spiders were rendered lethargic after
316
5-10 minutes. In others it appeared to have minimal effect even for induction times
317
greater than 30 minutes. The brief time needed to move the animal from the
318
induction chamber to the scanning chamber (<15 seconds) was often sufficient for
319
recovery from the initial lethargic state.
320
321
Although alternative injectable anaesthetics have been studied recently in spiders
322
(Gjeltema 2014) more effective inhalation anaesthesia agents should be investigated
323
for future use. Additionally, the design and use of a combined induction/scanning
324
chamber is recommended to avoid the need to remove spiders from the anaesthesia
325
environment. Placement of the spiders in an oxygen enriched environment post-
326
scanning is recommended to potentially accelerate recovery from the deep level of
327
anaesthesia induced in subjects over the scan duration.
328
329
330
The RG-CINE-MRI sequence used was designed for rodent hearts beating at much
331
higher rates compared to spiders (~20 beats per minute (bpm) compared to
332
~350bpm in rats and ~550 bpm in mice). Therefore, a potential concern was that the
333
spider heart rate would be insufficient and provide too few complete cardiac cycles
334
for the reconstruction algorithms of the RG-CINE-MRI software to work. This did not
335
prove as problematic as originally thought and the reconstruction appeared robust
336
and consistent with experience in house.. Our own particular scan settings may not
337
be optimal in terms of balancing total scan time (and so cost) against the obvious
338
benefits of in vivo MRI as a technique. Fewer image averages and Intragate cardiac
339
cycles should be possible to speed up the imaging process without compromising
340
the final measurement quality. Our high imaging resolution enabled us to see the
341
operation of the heart ostia, but for simple analysis of heart function the image
342
resolution could also be reduced, speeding up image acquisition further. It should be
343
noted that in these experiments the RG-CINE-MRI for each slice was acquired
344
sequentially. We found this gave stable pseudo cardiac and respiratory gating
345
signals that are needed for the retrospective reconstruction of the cardiac images.
346
However, a more efficient multi-slice approach is often used in rodent cardiac
347
studies, which could be potentially applied to spiders (Heijman et al., 2006).
348
349
These results suggest that RG-CINE-MRI can potentially be applied to spiders and
350
other invertebrates. The availability of MRI micro-coils, from just 50um diameter
351
(Webb 2013), would allow even small invertebrates to be studied. However, it is
352
worth noting that the centralised, cohesive heart structure of spiders is comparatively
353
rare. Amongst both arachnids and insects, a chain of small ‘pseudo-hearts’ that act
354
collectively is more the norm (Klowden 2007). Though, the application of RG-CINE-
355
MRI to such a chain of pseudo hearts should be technically possible, it would need
356
to be practically tested.
357
358
In summary, we acquired the first in vivo cinematic cardiac MRI images from spiders.
359
From these images we were able to directly measure common cardiac functional
360
parameters for the first time in an identical manner to existing human and rodent
361
cardiac MRI. Cardiac ejection fraction was found to increase with animal mass at a
362
faster rate than myocardial tissue volume while heart rate stayed constant across
363
animals, suggesting the spider heart undergoes additional biomechanical loading
364
with age. The RG-CINE-MRI technique provides much potential for in vivo cardiac
365
MRI research to be expanded into a wider range of novel species.
366
367
Acknowledgements
368
369
The authors thank Dr M. A. Jansen and Prof. W. N. McDicken for advice throughout
370
this project.
371
372
Supporting Information
373
Additional Supporting Information may be found in the online version of this article.
374
Movie M1 Example CINE cardiac movie
375
Movie M2 Example CINE cardiac Movie
376
Movie M3 Example CINE cardiac Movie
377
378
Contributions
379
380
GM and WMH conceived and designed the experiments. GM, JM, LG, RP and WMH
381
performed the experiments. GM, NB and LH analysed the resultant images and data.
382
GM, NB and WMH prepared the manuscript
383
384
385
References
386
387
Bromhall C. (1987). Spider heart-rates and locomotion. Journal of Comparative
388
Physiology B 157, 451-460.
389
390
Bovens S.M., Boekhorst B.C.M.T., den Ouden K., van de Kolk K.W.A., Nauerth
391
A., Nederhoff M.G.J., Pasterkamp G., ten Hove M., van Echteld C.J.A. (2011)
392
Evaluation of infarcted murine heart function: comparison of prospectively triggered
393
with self-gated MRI. NMR in Biomedicine. 24, 307-315.
394
395
Carrel J.E. and Heathcote R.D. (1976) Heart Rate in Spiders: Influence of Body
396
Size and Foraging Energetics. Science, New Series, 193, 148-150.
397
398
Coelho F.C. and Amaya C.C. (2000). Measuring the heart rate of the spider,
399
Aphonopelma hentzi: a non-invasive technique . Physiological Entomology, 25, 167–
400
171. (doi:10.1046/j.1365-3032.2000.00182.x)
401
402
Dombrowski, D.S., De Voe R.S. and Lewbart G.A. (2013) Comparison of
403
Isoflurane and Carbon Dioxide Anesthesia in Chilean Rose Tarantulas (Grammostola
404
rosea). Zoo Biol. 32, 101–103. (doi: 10.1002/zoo.21026)
405
406
Dorbala S., Crugnale S., Yang D., Di Carli M.F. (2006). Effect of body mass index
407
on left ventricular cavity size and ejection fraction. American Journal of Cardiology
408
97, 725-729
409
410
Dunlop J.A., Altringham J.D. and Mill P.J. (1992). Coupling between the heart and
411
sucking stomach during ingestion in a tarantula. J. Exp. Biol. 166, 83-93.
412
413
Epstein F.H. (2007). MRI of left ventricular function. Journal of Nuclear Cardioogy.
414
40, 729-744.
415
416
Foelix R.F. (1996) Biology of Spiders. Oxford, UK; Oxford University Press.
417
418
Gebhard C., Staehli, B. E, Gebhard, C. E., Jenni, R.,Tanner, F. C. (2012). Effect of
419
age on left ventricular ejection fraction assessed by echocardiography. European
420
Heart Journal, 33, 565-565
421
422
Gjeltema J., Posner L.P., and Stoskopf M. (2014). The use of injectable
423
alphaxalone as a single agent and in combination with ketamine, xylazine, and
424
morphine in the Chilean rose tarantula, Grammostola rosea. Journal of Zoo and
425
Wildlife Medicine 45(4):792-801. 2014 (doi: http://dx.doi.org/10.1638/2013-0223.1)
426
427
428
Heijman E., de Graaf W., Niessen P., Nauerth A., van Eys G., de Graaf L.,
429
Nicolay K. and Strijkers G.J. (2006). Comparison between prospective and
430
retrospective triggering for mouse cardiac MRI. NMR in Biomedicine 20, 439447
431
(doi:10.1002/nbm.1110)
432
433
Heijman E., Aben J.P., Penners C. et al. (2008). Evaluation of manual and
434
automatic segmentation of the mouse heart from CINE MR images. Journal of
435
Magnetic Resonance Imaging 27, 86-93. (doi:10.1002/jmri.21236)
436
437
Holmes W.M., McCabe C., Mullin J.M., Condon B., Bain M.M. (2008) Non-
438
Invasive Self-Gated MR Cardiac Imaging of Developing Chick Embryos In-Ovo.
439
Circulation. 117, E346-347.
440
441
Holmes W.M., McCabe C., Mullin J.M., Condon B., Bain M.M. (2009). In ovo non-
442
invasive quantification of the myocardial function and mass of chick embryos using
443
magnetic resonance imaging. NMR in Biomedicine. 22: 745-752.
444
445
Huckstorf K., Kosok G., Seyfarth E.A., Wirkner C.S. (2013). The hemolymph
446
vascular system in Cupiennius salei (Araneae: Ctenidae), Zoologischer Anzeiger
447
252, 76-87. (doi:10.1016/j.jcz.2012.03.004)
448
449
Jerosch-Herold M. (2010). Quantification of myocardial perfusion by cardiovascular
450
magnetic resonance. Journal of Cardiovascular Magnetic Resoance. 12, 57.
451
452
Klowden M.J. (2007) Physiological Systems in Insects. Burlington, USA; Academic
453
Press.
454
455
Markl M., Frydrychowicz A., Kozerke S. (2007). 4D Flow MRI. Journal of Magnetic
456
Resonance Imaging. 36, 1015-1036.
457
458
Merrifield G.D., Mullin J., Gallagher L., Tucker C., Jansen M.A., Denvir M.,
459
Holmes W.M. (2017). Rapid and recoverable in vivo MRI of adult zebrafish at 7T.
460
Magnetic Resonance Imaging. 37, 9-15.
461
462
463
Paul R.J., Bihlmayer S., Colmorgen M. and Zahler S. (1994). The Open
464
Circulatory System of Spiders (Eurypelma californicum, Pholcus phalangioides): A
465
Survey of Functional Morphology and Physiology. Physiological Zoology 67, 1360-
466
1382.
467
468
Pizzi, R. (2012). Spiders. In: Invertebrate Medicine, 2nd edition (ed Lewbert, G.A.)
469
pp. 187-222, Chichester, UK; Blackwell Publishing
470
471
Pohlmann A., Möller M., Decker H., and Schreiber W.G. (2007). MRI of tarantulas:
472
morphological and perfusion imaging. Magn Reson Imaging. 25, 129-35.
473
474
Seo J.S., Jin H.Y., Jang J.S., Yang T.H., Kim D.K., Kim D.S. (2017). The
475
Relationships between Body Mass Index and Left Ventricular Diastolic Function in a
476
Structurally Normal Heart with Normal Ejection Fraction. J Cardiovasc Ultrasound.
477
25, 5-11
478
479
Schneider C.A., Rasband W.S. and Eliceiri K.W. (2012). NIH Image to ImageJ: 25
480
years of image analysis. Nature Methods 9, 671-675.
481
482
Webb A.G. (2013). Radiofrequency microcoils for magnetic resonance imaging and
483
spectroscopy. Journal of Magnetic Resonance. 229, 55-66.
484
485
Zachariah T.T., Mitchell M.A., Guichard C.M. and Singh R.S. (2009). Isoflurane
486
anesthesia of wild-caught goliath birdeater spiders (Theraphosa blondi) and Chilean
487
rose spiders (Grammostola rosea). J Zoo Wildl Med 40, 347–349.
488
489
490
Spider
1
2
3
4
5
6
Mean ± std
Body Mass (g)
11.7
12.00
12.00
15.10
16.09
18.8
14 ± 3
Heart Rate (beats/min)
20.2
(±1)
20.6
(±1)
21.1
(±2)
22.9
(±4)
16.0
(±3)
20.6
(±6)
20 ± 2
End systolic ventricular
Volume (ESV) (mm3)
29.6
50.0
30.7
28.6
53.7
49.4
40 ± 12
End Diastolic ventricular
Volume (EDV) (mm3)
42.5
57.5
51.1
59.9
96.5
115.1
70 ± 29
Ejection Fraction (EJ)
0.25
(±0.04)
0.13
(±0.02)
0.37
(±0.04)
0.52
(±0.02)
0.48
(±0.04)
0.57
(±0.01)
0.38 ± 0.17
Heart volume (systole) (mm3)
114
134
91.8
135
176
184
139 ± 35
Heart volume (diastole)
(mm3)
122
140
114
170
213
232
165 ± 48
Volume of Myocardium
(systole) (mm3)
84.0
83.7
61.1
106
123
135
99 ± 28
Volume of Myocardium
(diastole) (mm3)
79.3
82.6
63.1
110
116
117
95 ± 23
Change in myocardium
Thickness (mm3)
0.06
0.01
-0.02
-0.03
0.08
0.15
0.04 ± 0.07
Table 1. Cardiovascular physiological measurements made on each of the six
491
spiders from the cine MRI datasets. The mean values from all six spiders is given in
492
the end column
493
494
495
496
497
Figure Legends
498
499
500
Figure 1. Experimental set up for scanning. A. 7T preclinical MRI scanner used in
501
the experiments, B. Spider in anaesthetic chamber undergoing anaesthesia
502
induction, C. Anaesthetised spider lying prone on back in place above MRI coil, D.
503
Spider ready for scanning now with restraints in place and with plastic sleeve in
504
place over the coil and cradle.
505
506
507
508
Figure 2. Example cardiac images from RG-CINE-MRI data. Axial image slices
509
showing A. Diastolic phase (heart arrowed), B. Systolic phase (heart arrowed), C.
510
Heart with ostia (arrowed) in the open position, D. Heart with ostia (arrowed) in the
511
closed position.
512
513
514
515
Figure 3. Results of Cardiac MRI Analysis showing A. Mean and standard
516
deviation of measurements of cardiac ejection fraction determined between three
517
independent researchers (for n=6 spiders). B. RG-CINE-MRI navigator sourced
518
mean heart rates in Beats Per Minute (BPM) with s.d. error bars. After RG-CINE-MRI
519
navigator signal processing the mean heart rate for each slice of cardiac data from
520
an individual subject is estimated and then a mean heart rate is generated for the
521
entire heart (n=6).
522
523
