1 23

European Journal of AppliedPhysiology ISSN 1439-6319Volume 111Number 9 Eur J Appl Physiol (2011)111:2213-2219DOI 10.1007/s00421-011-1857-8

Post-dive ultrasound detection of gas inthe liver of rats and scuba divers

Antonio L’Abbate, Claudio Marabotti,Claudia Kusmic, Antonino Pagliazzo,Alessandro Navari, Vincenzo Positano,Mario Palermo, Antonio Benassi, et al.

1 23

Your article is protected by copyright and

all rights are held exclusively by Springer-

Verlag. This e-offprint is for personal use only

and shall not be self-archived in electronic

repositories. If you wish to self-archive your

work, please use the accepted author’s

version for posting to your own website or

your institution’s repository. You may further

deposit the accepted author’s version on a

funder’s repository at a funder’s request,

provided it is not made publicly available until

12 months after publication.

Eur J Appl Physiol (2011) 111:2213–2219

Author's personal copy

DOI 10.1007/s00421-011-1857-8

ORIGINAL ARTICLE

Post-dive ultrasound detection of gas in the liver of rats and scuba divers

Antonio L’Abbate · Claudio Marabotti · Claudia Kusmic · Antonino Pagliazzo · Alessandro Navari · Vincenzo Positano · Mario Palermo · Antonio Benassi · Remo Bedini

Received: 3 August 2010 / Accepted: 29 January 2011 / Published online: 12 February 2011© Springer-Verlag 2011

Abstract In a previous study, we obtained histologic doc-umentation of liver gas embolism in the rat model of rapiddecompression. The aim of the study was to assess in thesame model occurrence and time course of liver embolismusing 2-D ultrasound imaging, and to explore by this meansputative liver gas embolism in recreational scuba divers.Following 42 min compression at 7 ATA breathing air and12 min decompression, eight surviving female rats wereanesthetized and the liver imaged by ultrasound at 20 minintervals up to 120 min. A signiWcant enhancement of echosignal was recorded from 60 to 120 min as compared toearlier post-decompression times. Enzymatic markers ofliver damage (AST, ALT, and GGT) increased signiWcantlyat 24 h upon decompression. Twelve healthy experienceddivers were studied basally and at 15-min intervals up to60 min following a 30-min scuba dive at 30 msw depth. At

30 min upon surfacing echo images showed signiWcant sig-nal enhancement that progressed and reached plateau at 45and 60 min. Total bilirubin at 24 h increased signiWcantly(p = 0.02) with respect to basal values although within thereference range. In conclusion, 2-D ultrasound liver imag-ing allowed detection of gas embolism in the rat and deW-ned the time course of gas accumulation. Its application toscuba divers revealed liver gas accumulation in all subjectsin the absence of clear-cut evidence of liver damage or ofany symptom. The clinical signiWcance of our Wndingsremains to be investigated.

Keywords Decompression gas embolism · Decompression rat model · Scuba diving · Liver gas embolism · Liver dysfunction · Liver ultrasound

Introduction

During diving, the body is exposed to increased hydrostaticpressure, and respiratory gases are forced to dissolve intobody Xuids according to their partial pressure, accumulat-ing with time. In divers breathing compressed gas, particu-larly if they ascend to the surface too quickly, the tissue gasoverload may give rise to a gas phase (bubbles) in both thetissues and blood. Bubbles entering the blood at the micro-circulation level Xow through the systemic venous circula-tion to the right heart and embolize the lungs. Flowing gasbubbles in the right heart chambers and/or in the pulmonaryartery trunk have been previously detected by transthoracicultrasound imaging or Doppler in scuba divers. The pres-ence of bubbles was reported regardless of safe dive proWleand even in the absence of any sign of decompression sick-ness (DCS). In fact, Xowing gas bubbles may be detected inthe pulmonary artery trunk and in the right ventricle of the

Communicated by Dag Linnarsson.

A. L’AbbateScuola Superiore Sant’Anna, Pisa, Italy

A. L’Abbate · C. Marabotti · M. Palermo · A. Benassi · R. BediniExtreme Centre of Scuola Superiore Sant’Anna, Pisa, Italy

C. Marabotti · C. Kusmic (&) · A. Pagliazzo · R. BediniCNR Institute of Clinical Physiology, Via G Moruzzi 1, 56124 Pisa, Italye-mail: kusmic@ifc.cnr.it

C. Marabotti · A. Navari · V. PositanoFondazione Toscana Gabriele Monasterio, Pisa, Italy

M. PalermoEndocrinology Service AOU, Sassari, Italy

A. BenassiCNR Institute of Information Science and Technology, Pisa, Italy

123

2214 Eur J Appl Physiol (2011) 111:2213–2219

Author's personal copy

majority of scuba divers (Bennett et al. 2007; Brubakk andEftedal 2001; Carturan et al. 2002; Marabotti et al. 1999;Ljubkovic et al. 2010).

Previously, we documented extensive gas bubble embo-lism in the liver by gross anatomy and histology in the rapiddecompression rat model adopted by WisloV and Brubakk(2001) to reproduce systemic gas embolization. These Wnd-ings challenged the existing paradigm that the liver isimmune from gas embolism in DCS. In addition, our datashowed evidence of zonal or diVuse hepatocellular damagefrom 3 to 24 h after decompression, associated with a sig-niWcant increase in enzymatic markers of liver damage(L’Abbate et al. 2010).

In the clinical setting, hepatic portal venous gas has beenreported as an unusual and unfavorable clinical entity andrecognized in association with ischemic and non-ischemicbowel diseases (Peloponissios et al. 2003; Schindera et al.2006). In contrast, only scanty and contradictory Wndingson portal, mesenteric, liver, and spleen gas embolism aswell as on blood markers of liver dysfunction have beenreported in DCS (Barak and Katz 2005; Butler et al. 1995;Jauchem et al. 1986; Philp et al. 1975; Bird 2007; Koeniget al. 2007; Sanders and Smerz 2010) favoring the beliefthat the liver is not a target organ of DCS.

With these considerations, the Wrst aim of our study wasto assess ultrasound imaging as a means to non-invasivelyevaluate liver gas embolism and to deWne its time course inan animal model characterized by consistent liver embo-lism upon decompression. The second aim of the study wasto verify this diagnostic approach in the examination of atest group of experienced scuba divers.

Materials and methods

Animals and experimental protocol

Experimental protocols were approved by the Animal CareCommittee of the Italian Ministry of Health and conformedto the “Guiding Principles for Research Involving Animalsand Human Beings” approved by the Council of the Ameri-can Physiological Society.

Twenty-one adult female Wistar rats 6–8 months of ageand weighing 274 § 47 g were studied. Animals werehoused under controlled 12/12 h light/dark cycle, tempera-ture (21 § 0.5°C) and relative humidity (55 § 2%) in animalrooms and fed with 4 RF 18 standard rodent diet for long-term maintenance (Mucedola, Italia) and water ad libitum.

Seventeen of the 21 rats were compressed (simulateddive) in a 50-l home-made hyperbaric chamber in pairs, buthoused in separate cages. Chamber characteristics have pre-viously been described (L’Abbate et al. 2010). Animalswere compressed at a rate of 200 kPa min¡1 (2 ATA) to a

Wnal pressure of 700 kPa (7 ATA), and maintained at thatpressure for 42 min, breathing air. At the end of the com-pression period, the rats were almost linearly decompressedto the ‘surface’ (100 kPa, 1 ATA) at a rate of 50 kPa min¡1

(0.5 ATA). Using the above proWle of a simulated dive, wepreviously documented consistent gas bubble presence inthe liver sinusoids by histologic examination, as well asincreased biomarkers of liver dysfunction at 24 h upondecompression (L’Abbate et al. 2010). The remaining fourrats were not compressed in the hyperbaric chamber andrepresented the control group for echo imaging.

Ultrasound acquisition and image analysis

After ‘surfacing’, surviving animals were anesthetized byintraperitoneal injection of Zoletil 100 + xylazine (50 and3 mg/kg, respectively) and the inferior vena cava and theliver were insonated using a portable ultrasound system(MyLab 25, Esaote SpA, Genoa, Italy) equipped with amulti-frequency linear transducer (LA 523, 12.5 MHz).Both transversal and longitudinal echo images of the liverwere obtained and recorded at 0.7 mechanical index (Pa/Hz) every 20 min for 2 h. Setting of echograph (ultrasoundpower, gain, persistency, dynamic range and gray scale)was kept constant along the series of acquisitions for allanimals. Images in DICOM (digital imaging and communi-cation in medicine) format were saved for oV-line analysis.The onset of bubble formation in the systemic circulationwas alerted by the appearance of Xashing bright spots,directed cranially, in the inferior vena cava. Throughout thestudy, data were stored and reviewed in cine-mode to allowa better deWnition of the liver for provisional blind analysis.For comparison, we followed the same imaging protocol inthe anesthetized control animals.

Images were analyzed using MIPAV 4.4 (CIT, NIH,Bethesda, MD, USA) software (McAuliVe et al. 2001). Foreach animal, circular regions of interest (ROI) of Wxeddimension were placed in correspondence of the same liverareas at diVerent frames. The mean and SD of the signalwithin the ROI was evaluated among the rats.

Measurement of plasma markers of liver function

Plasma aspartate aminotransferase (AST), alanine amino-transferase (ALT), gamma-glutamyl transferase (GGT),lactate dehydrogenase (LDH), creatine kinase (CK), andtotal bilirubin (TBil) were assayed by routine laboratorymethods using a CX9PRO autoanalyzer (Beckman CoulterInc., Brea, CA, USA). Heparinized blood samples wereobtained from the femoral vein in both control group anddecompressed rats. In the latter group, blood samples werecollected before the simulated dive in the hyperbaric cham-ber (baseline) and at 24 h post-decompression.

123

Eur J Appl Physiol (2011) 111:2213–2219 2215

Author's personal copy

Scuba divers and experimental design

A group of 12 experienced scuba (self-contained underwa-ter breathing apparatus) divers was enrolled in the study (9males; aged 37 § 10 years, age range 22–56 years; bodyweight 74 § 16 kg). The study protocol was approved bythe Local Ethics Committee. All participants were informedabout the aims and procedures of the study and gave theirwritten consent. Divers were convened at Asinara Island(Sardinia, Italy) on the occasion of the student summerstage of the Master in Underwater and Hyperbaric Medi-cine.

No subject had a history of clinical or instrumental (rest-ing ECG, Doppler echocardiography) evidence of arterialhypertension, cardiac or pulmonary diseases. All subjectswere non-smokers and had been fasting for at least 2 hbefore the study.

Ultrasound acquisition and image analysis

Subjects were studied using hepatic echography basally andat 15, 30, 45 and 60 min after surfacing from a “square”scuba dive (30 min from start to surfacing; 30 m depth).Dive proWle was according to recommendations of USNavy Diving Manual Revision 6 (US Department of theNavy 2008).

Hepatic echography was performed by a commerciallyavailable ultrasound system (Mylab25, Esaote SpA)equipped with phased-array linear transducer (CA 430,3.5 MHz). Setting of echograph (ultrasound power, gain,persistency, dynamic range and gray scale) was kept con-stant for all subjects along the series of acquisitions. A righthepatic lobe scan displaying hepatic veins merging intoinferior vena cava and a sagittal left hepatic lobe scan wereattained and recorded at 0.7 mechanical index (Pa/Hz) in allsubjects every 15 min upon decompression.

Imaging time-sequence was chosen according to thetime course observed in the rat (delayed S-shaped curve,see results below) and to the hands-on constraints ofon-Weld studies.

DICOM images were analyzed using the MIPAV 4.4(CIT, NIH, Bethesda, MD, USA) software (McAuliVe et al.2001). A circular ROI of Wxed dimension was placed insimilar liver regions of diVerent temporal frames. The meanand SD of the signal within the ROI was evaluated amongthe subjects.

Measurement of plasma markers of liver function in divers

TBil, AST, ALT, GGT, LDH, CK, creatine kinase isoen-zyme-MB (CK-MB), and serum cholinesterase (ChE) wereassayed by routine laboratory methods using an AbbottARCHITECT c8000 autoanalyzer.

Statistical analysis

All imaging data were analyzed using Medcalc v. 7.2.1(MedCalc Software, Mariakerke, Belgium) statistical pack-age. All continuous variables were expressed as the meanand SD. One-way repeated measures analysis of variance(ANOVA) was used to evaluate whether there was a sig-niWcant diVerence between diVerent measurements. Bonfer-roni correction for multiple comparison was applied.Correlation analysis was performed using Spearman’s test.Comparisons between the groups were made by paired orunpaired t test, as appropriate. In all tests, a two-tailedp < 0.05 was considered statistically signiWcant.

Results

Animal study

Eight of the 17 decompressed animals survived and wereanalyzed in the present study. The observed 53% mortalitywas in agreement with previous reports using the same diveproWle (L’Abbate et al. 2010; WisloV and Brubakk 2001).

Liver ultrasound imaging

The onset of bubble formation in the lower systemiccirculation became apparent, as from ultrasound imag-ing of inferior cava vein, between 20 and 120 min follow-ing decompression.

Post-processing analysis of echo liver images showedthat signal intensity increased progressively toward a pla-teau over a 2-h period as shown in Fig. 1. Tissue echo sig-nals were signiWcantly enhanced starting from 60 min ascompared to the Wrst time block of acquisition (p < 0.05).The control group showed no diVerence in the intensity ofecho signal over similar time-sequence of acquisition.

Plasma markers of liver function

Plasma enzymatic markers of liver function measuredbasally and 24 h upon decompression are reported inTable 1. The simulated dive induced functional hepaticdamage as assessed by the signiWcant increase in AST,ALT, and GGT levels at 24 h after decompression.

Study on scuba divers

Liver ultrasound imaging

Because of technical reasons imaging protocol wascompleted in all subjects but one. Analysis of variance(ANOVA) in the remaining 11 scuba divers showed a

123

2216 Eur J Appl Physiol (2011) 111:2213–2219

Author's personal copy

signiWcant dependence of the signal from the acquisitiontime (F ratio = 14.1, p < 0.001) A signiWcant hyper-enhancement of tissue echo signal with respect to the basalcondition at 30, 45, and 60 min following the dive wasfound (p < 0.001) (Fig. 2a). No signiWcant hyper-enhance-

ment was detected at 15 min (p = 0.49). Signal values at 15,30, 45, and 60 min were not signiWcantly diVerent. Accord-ing to analysis of individual subjects, the onset of signalhyper-enhancement was present in 6/11 subjects at 15 min,in 9 at 30 min and in all subjects at 45 and 60 min (Fig. 2b).

Fig. 1 Group intensity values in echo signal acquired in control rats(black bars, n = 4) and rats that survived a simulated dive (gray bars,n = 8) within regions of interest in comparable liver areas at 20-minintervals over a 2-h period following rapid decompression. Signalintensity indicates the mean gray scale/ROI and it is expressed in arbi-

trary units (a.u.). Signal values from diVerent animals were collectedand averaged in acquisition time blocks of 20 min. Data are presentedas mean § SD. * p < 0.05 versus the Wrst time block of acquisitionwithin group and versus parallel time block between groups

Table 1 Hematochemical markers of liver function in the rat model

Values are mean § SD in Unit/liter

n number of animals tested, AST aspartate aminotransferase, ALT alanine aminotransferase, GGT �-glutamyltransferase, LDH lactate dehydroge-nase, CK creatine kinase, TBil total bilirubin

* p < 0.05 versus basal value

n AST (U/liter) ALT (U/liter) GGT (U/liter) LDH (U/liter) CK (U/liter) TBil (mg/dl)

Basal 8 51 § 2 25 § 1 3 § 0.7 867 § 133 138 § 15 0.6 § 0.1

24 h upon decompression 8 130 § 12* 55 § 20* 6 § 1* 779 § 150 170 § 33 0.6 § 0.1

Fig. 2 a Group intensity values in echo signal, obtained in 11 scubadivers within regions of interest in comparable liver areas, in basal con-dition (before diving) and at 15-min intervals up to 60 min followingsurfacing. Signal intensity indicates the mean gray scale/ROI and it is

expressed in arbitrary units (a.u.). Data are presented as mean § SD.* p < 0.001 versus basal signal intensity. b Individual signal intensityvalues at basal and over a 1-h period upon decompression

123

Eur J Appl Physiol (2011) 111:2213–2219 2217

Author's personal copy

Plasma markers of liver function in divers

Basally, seven divers showed CK values above referencelaboratory range. This could likely be due to frequent exer-cise training of the subjects studied. However, as shown inTable 2, values of the majority of humoral markers of liverfunction did not show signiWcant changes at 24 h relative tobaseline, with the exception of bilirubin which increased sig-niWcantly (p = 0.02), although within the reference labora-tory range, and in association with a trend toward increasedAST and LDH (p = 0.05 and p = 0.048, respectively).

Discussion

In a previous study in the rat model of rapid decompres-sion, we documented by gross anatomy and histology thepresence of bubbles in the liver at 3 h following decom-pression (L’Abbate et al. 2010). The rat model we used wasthat previously employed by WisloV and Brubakk (2001) toinvestigate post-decompression gas bubble formation in thesystemic circulation by means of ultrasound imaging of theright heart cavities.

In the present study, we aimed at documenting in thesame animal model: (1) the ability of liver echography todetect gas embolism and (2) the time course of liver embo-lism by multiple ultrasound acquisitions. We noted a pro-gressive increase in echo signal which reached itsmaximum within 1 h.

Thus, on the basis of the experimental evidence thatultrasound imaging of the liver was a suitable tool forassessing gas accumulation in the organ, we used thisapproach to investigate putative liver gas embolism inscuba divers. The presence of gas bubbles in the systemiccirculation of scuba divers, even in the absence of any clin-ical eVect, has been repeatedly reported (Bennett et al.2007; Butler et al. 1995; Marabotti et al. 1999). In light ofour experimental results obtained in the rat model, although

aware of the striking diVerence between the two dive pro-Wles and their harmfulness, we reasoned that it would beworthwhile to investigate gas bubble formation in thesplanchnic circulation of scuba divers as well. We thoughtthat a 1-h period of multiple ultrasound acquisitions was areasonable compromise between the expected scientiWcanswers and the logistic requirements of on-Weld protocolin a relatively large group of divers.

Actually, the Wndings that we obtained in scuba diverswere beyond our expectation in terms of consistency andintensity of the signal. Although, there was neither subjec-tive nor objective evidence of liver damage, our results indi-cate that the systemic circulation is not the only source ofgas bubble formation in scuba divers. Instead our data sug-gest that the phenomenon occurs in the splanchnic circula-tion as well, thus leading to possible sinusoidal obstruction.

To the best of our knowledge, only a few case reports havebeen published regarding liver gas embolism in divers in asso-ciation or not with DCS and/or abdominal pain (Butler et al.1995; Bird 2007; Koenig et al. 2007; Sanders and Smerz2010). However, whether liver gas embolism was an inciden-tal phenomenon or rather a common event in scuba divers wasstill an unresolved question until the present study.

Study limitations

Interaction between bubbles and ultrasound

Gas microbubbles in a liquid medium exhibit uniquebehavior when exposed to an ultrasound Weld, dependingon the acoustic transmit power (Eatock and Nishi 1985). Atlow-emission power (mechanical index less than 1.0), bub-bles are exposed to a low-incident pressure and their vol-ume is not altered. The echo signal enhancement istherefore proportional, in theory and within a narrow rangeof volume, to the amount of microbubbles into the inson-ated tissue. This response is termed as linear backscatter(Burns et al. 1992; De Jong 1997). At higher output power,

Table 2 Hematochemical markers of liver function in scuba divers

Values are mean § SD in Unit/liter

n number of subjects tested, AST aspartate aminotransferase, ALT alanine aminotransferase, GGT �-glutamyltransferase, LDH lactate dehydroge-nase, CK creatine kinase, CK-MB muscle body mass creatine kinase, ChE cholinesterase, TBil total bilirubin

# p = 0.05§ p = 0.048

* p = 0.02 versus basal value

n AST (U/liter)

ALT (U/liter)

GGT (U/liter)

LDH (U/liter)

CK (U/liter)

CK-MB (U/liter)

ChE (U/liter) £1,000

TBil (mg/dl)

Reference range 5–34 10–55 12–64 M 125–243 30–200 M 0–3.6 5.1–11.7 M 0.2–1.3

10–55 F 29–168 F 4–12 F

Basal 12 25 § 2 27 § 5 25 § 4 190 § 37 255 § 6 1.4 § 0.2 11 § 1 0.4 § 0.1

24 h after dive 12 26 § 3# 28 § 6 25 § 4 212 § 31§ 287 § 13 1.6 § 0.1 11 § 1 0.5 § 0.1*

123

2218 Eur J Appl Physiol (2011) 111:2213–2219

Author's personal copy

the microbubbles produce nonlinear backscatter, i.e. whenexposed to ultrasound the bubbles vibrate in a rapid succes-sion of expansion–compression cycles and importantly,when subjected to critical ultrasound power, are destroyed(Apfel and Holland 1991; De Jong 1997; Tiemann et al.1997). Schematically, linear backscatter occurs at low-mechanical index (<0.1), while bubble disruption is the rulewith default settings (mechanical index > 1.0) (Becher andBurns 2000). In the present study, we used low-mechanicalindex in both the animal and human studies in order topreserve gas bubble survival. However, only semi-quantita-tive measurement of gas accumulation can be obtained inthis way.

Time course of liver echo signal enhancement

In the rat study, in each animal pre-diving data were notrecorded in order to prevent animals from undergoing gen-eral anesthesia twice. However, the compressed animalgroup was compared with a group of non-compressed ani-mals for the same time sequence and signal enhancementwith time in compressed animals was quite evident at bothintra- and inter-group analyses.

In scuba divers, due to logistic reasons related to the on-Weld study, echo imaging acquisition was limited to 60 min.Thus, in the present study, we cannot state how long liverenhancement of echo signals persists or even argue the timecourse of its regression. These data will supply importantinformation to explore in detail the kinetics of liver bubblesand will be included in further studies.

Furthermore, since in the present study, we did notassess the occurrence of bubbles in the systemic circulation,and we did not obtain any information on the coincidenceor not in the time course of portal versus systemic embo-lism. This topic deserves a dedicated study.

Plasma bilirubin increase upon decompression

Unlike in the rat, we observed a signiWcant increase of totalbilirubin in scuba divers associated with a trend towardincreased AST and LDH (p = 0.05). In a recent study Gla-vas et al. (2008) showed a signiWcant increase in LDH, CK,and CK-MB in eight scuba subjects diving to 54 msw for20 min bottom time; however, they did not measure totalbilirubin. At present, the interpretation of the total bilirubinincrease in our study is merely speculative and requiresmore speciWc investigation.

Perspectives and signiWcance

This is the Wrst report to consistently explore potential gasaccumulation in the liver as a consequence of venous

splanchnic embolism in scuba divers. The study was per-formed following the preliminary documentation that liverembolism actually occurs in the rapidly decompressed ratand that the phenomenon can be non-invasively detected byultrasound imaging of the liver.

Compared to the lung, embolizing bubbles in the livermay persist for a much longer time and potentially damagethe organ. However, the clinical relevance of our Wndingsrequires further investigation. Although, we did not observeany signiWcant increase in enzymatic markers of liver dam-age, but only a signiWcant increase in total bilirubin, thepathological consequences on liver function in human DCSor even outside DCS should be carefully sought and clini-cally evaluated.

A prudent policy for safe diving would require short-and long-term studies on putative liver dysfunction in sub-jects exposed to decompression, even when using com-monly accepted diving proWles.

Acknowledgments Authors wish to express their gratitude to Asin-ara National Park and the Conservatoria delle Coste della RegioneSardegna for their valuable logistic support, and to the students of theMaster in Underwater and Hyperbaric Medicine for their active contri-bution to the study on scuba divers. The work has been supportedby the Master in Underwater and Hyperbaric Medicine and the EX-TREME Centre of the Scuola Superiore Sant’Anna, Pisa, and by theCNR Institute of Clinical Physiology, Pisa, Italy. Preliminary data ofthe present work were presented in part at the European Underwaterand Baromedical Society, September 2010.

ConXict of interest There was no personal or Wnancial support orauthor involvement with organizations having Wnancial interest in thesubject matter, or any actual or potential conXict of interest for theauthors of the study.

References

Apfel RE, Holland CK (1991) Gauging the likelihood of cavitationfrom short-pulse, low-duty cycle diagnostic ultrasound. Ultra-sound Med Biol 17:175–185

Barak M, Katz Y (2005) Microbubbles pathophysiology and clinicalimplications. Chest 128:2918–2932

Becher H, Burns PN (2000) Contrast agents for echocardiography:principles and instrumentation. In: Becher H, Burns PN (eds)Handbook of contrast echocardiography. Left ventricular func-tion and myocardial perfusion. Springer-Verlag, Berlin, pp 2–42

Bennett PB, Marroni A, Cronje FJ, Cali-Corleo R, Germonpre P, PieriM, Bonuccelli C, Leonardi MG, Balestra C (2007) EVect of vary-ing deep stop times and shallow stop times on precordial bubblesafter dives to 25 msw (82 fsw). Undersea Hyperb Med 34:399–406

Bird N (2007) CT Wnding of VGE in the portal veins and IVC in a diverwith abdominal pain: a case report. Undersea Hyperb Med34:393–397

Brubakk AO, Eftedal O (2001) Comparison of three diVerent ultra-sonic methods for quantiWcation of intravascular gas bubbles.Undersea Hyperb Med 28:131–136

Burns PN, Powers JE, Fritzsch T (1992) Harmonic imaging: a newimaging and Doppler method for contrast enhanced ultrasound.Radiology 185(P):142

123

Eur J Appl Physiol (2011) 111:2213–2219 2219

Author's personal copy

Butler BD, Fife C, Sutton T, Pogodsky M, Chen P (1995) Hepatic por-tal venous gas with hyperbaric decompression: ultrasonographicidentiWcation. J Ultrasound Med 14:967–970

Carturan D, Boussuges A, Vanuxem P, Bar-Hen A, Burnet H, GardetteB (2002) Ascent rate, age, maximal oxygen uptake, adiposity, andcirculating venous bubbles after diving. J Appl Physiol 93:1349–1356

De Jong N (1997) Physics of bubble scattering. In: Nanda NC, SchliefR, Goldberg BB (eds) Advances in echo imaging using contrastenhancement. Kluwer Academic Publishers, Dubai, pp 39–64

Eatock VC, Nishi RY (1985) Numerical studies of the spectrum oflow-intensity ultrasound scattered by bubbles. J Acoust Soc Am77:1692–1701

Glavas D, Markotic A, Valic Z, Kovacic N, Palada I, Martinic R, Bre-skovic T, Bakovic D, Brubakk AO, Dujic Z (2008) Expression ofendothelial selectin ligands on human leukocytes following dive.Exp Biol Med 233:1181–1188

Jauchem JR, Waligora JM, Conkin J, Horrigan DJ Jr, Johnson PC Jr(1986) Blood biochemical factors in humans resistant and suscep-tible to formation of venous gas emboli during decompression.Eur J Appl Physiol 55:68–73

Koenig MD, Medak AJ, van Hoesen KB (2007) Portal venous gasassociated with decompression sickness in sport divers: a case se-ries. Undersea Hyperb Med Soc meeting: Abstract F3

L’Abbate A, Kusmic C, Matteucci M, Pelosi G, Navari A, PagliazzoA, Longobardi A, Bedini R (2010) Gas embolization of the liverin a rat model of rapid decompression. Am J Physiol Regul IntegrComp Physiol 299:R673–R682

Ljubkovic M, Marinovic J, Obad A, Breskovic T, Gaustad SE, Dujic Z(2010) High evidence of venous and arterial gas emboli at restafter diving without protocol relation. J Appl Physiol.doi:10.1152/japplphysiol.01369.2009

Marabotti C, Chiesa F, Scalzini A, Antonelli F, Lari R, Franchini C,Data PG (1999) Cardiac and humoral changes induced by recrea-tional scuba diving. Undersea Hyperb Med 26:151–158

McAuliVe MJ, Lalonde FM, McGarry D, Gandler W, Csaky K, TrusBL (2001) Medical image processing, analysis & visualization inclinical research. Proceedings of the 14th IEEE symposium oncomputer-based medical System. CMBS pp 381–386

Peloponissios N, Halkic N, Pugnale M, Jornod P, Nordback P, MeyerA, Gillet M (2003) Hepatic portal gas in adults: review of the lit-erature and presentation of a consecutive series of 11 cases. ArchSurg 138:1367–1370

Philp RB, Freeman DJ, Francey I (1975) Hematology and blood chem-istry in saturation diving: II. Open sea versus hyperbaric chamber.Undersea Biomed Res 2:251–265

Sanders WR, Smerz RW (2010) Severe DCS presents with hypoten-sion, abdominal pain and ST elevation. Undersea Hyperb MedSoc meeting: Abstract F7

Schindera ST, Triller J, Vock P, Hoppe H (2006) Detection of hepaticportal venous gas: its clinical impact and outcome. Emerg Radiol12:164–170

Tiemann K, Becher H, Bimmel D, Schlief R, Nanda NC (1997)Stimulated acoustic emission. Non backscatter contrast eVect ofmicrobubbles seen with harmonic power Doppler imaging echo-cardiography. J Cardiovasc Ultrasound Allied Tech 14:65–69

US Department of the Navy (2008) US Navy Diving Manual. Rev 6.SS521-Ag-PRO-010; NAVSEA 0910-LP-106-0957. http://www.supsalv.org/pdf/DiveMan_rev6.pdf

WisloV U, Brubakk AO (2001) Aerobic endurance training reducesbubble formation and increases survival in rats exposed to hyper-baric pressure. J Physiol 537:607–611

123