This article has Open Peer Review reports available.
A novel software program for detection of potential air emboli during cardiac surgery
© Secretain et al.; licensee BioMed Central. 2015
Received: 31 October 2014
Accepted: 30 December 2014
Published: 12 January 2015
Risks associated with air emboli introduced during cardiac surgery have been highlighted by reports of postoperative neuropsychological dysfunction, myocardial dysfunction, and mortality. Presently, there are no standard effective methods for quantifying potential emboli in the bloodstream during cardiac surgery. Our objective was to develop software that can automatically detect and quantify air bubbles within the ascending aorta and/or cardiac chambers during cardiac surgery in real time.
We created a software algorithm (“Detection of Emboli using Transesophageal Echocardiography for Counting, Total volume, and Size estimation”, or DETECTS™) to identify and measure potential emboli present during cardiac surgery using two-dimensional ultrasound. An in vitro experiment was used to validate the accuracy of DETECTS™ at identifying and measuring air emboli. An experimental rig was built to correlate the ultrasound images to high definition camera images of air bubbles created in water by an automatic bubbler system. There was a correlation between true bubble size and the size reported by DETECTS™ in our in vitro experiment (r = 0.76). We also tested DETECTS™ using TEE images obtained during cardiac surgery, and provide visualization of the software interface.
While monitoring the heart during cardiac surgery using existing ultrasound technology and DETECTS™, the operative team can obtain real-time data on the number and volume of potential air emboli. This system will potentially allow de-airing techniques to be evaluated and improved upon. This could lead to reduced air in the cardiac chambers after cardiopulmonary bypass, possibly reducing the risk of neurological dysfunction following cardiac surgery.
KeywordsCardiac surgery Cardiopulmonary bypass Air emboli Cerebral emboli Neurological dysfunction Myocardial dysfunction Ultrasound Transcranial doppler Transesophageal echocardiography
During cardiac surgery, particularly during cardiopulmonary bypass (CPB), air may be introduced into circulation [1, 2]. These air bubbles can cause neurological dysfunction, myocardial dysfunction, and even death [2–6]. To avoid these potentially fatal effects, all surgeons employ a variety of de-airing techniques including venting and altering the patient’s position, and carbon dioxide field flooding in an attempt to reduce the impact of air emboli . However, these de-airing techniques are largely non-standardized. Additionally, the current methods for identifying potential air emboli do not allow for rapid, automatic, real-time detection during cardiac surgery. Having this capability would better enable surgeons to assess their de-airing techniques and adjust where necessary.
Transcranial Doppler (TCD) focused on the middle cerebral artery may assist in the identification of cerebral microemboli [8, 9]; however, this method does not detect the emboli until they reach the brain, where they pose a serious neurological threat. Thus, TCD cannot be used as a preventative solution. Transesophageal echocardiography (TEE) can be used to visualize potential emboli in the heart and circulation once the patient has been separated from CPB. TEE requires trained personnel to identify potential emboli, and does not provide a size estimate for each embolus . The ability to automatically detect the presence and estimate the size of potential air emboli in the heart in real time would allow the operating team to evaluate their de-airing techniques, and might contribute to the improvement and development of new and more effective de-airing practices. In turn, this may lead to a reduction in air emboli, thereby decreasing the rates of neurological dysfunction after cardiac surgery using CPB.
Here, we present a new software algorithm named Detection of Emboli using Transesophageal Echocardiography for Counting, Total volume and Size estimation (DETECTS™) as a solution to identify potential emboli in a rapid, real-time, and automatic manner. This program was designed to operate as an online system to benefit the operative team during cardiac surgery. We also present the results of an in vitro experiment validating the accuracy with which DETECTS™ quantifies the number and size of potential air emboli. Use of DETECTS™ in cardiac surgery patients was approved by the Queen’s University Health Sciences & Affiliated Teaching Hospitals Research Ethics Board.
DETECTS™ software algorithm and interface
In vitrovalidation of DETECTS™
Our apparatus was comprised of a “bubbler” at the base of a clear Plexiglas™ rise chamber, through which air bubbles could rise in tap water. Air was delivered into the rise chamber through an air input using a pneumatic cylinder connected to a stepper motor. The air bubbles were produced at the ends of glass tubes connected to one of the air inputs. A portion of one wall of the rise chamber was replaced with a laminated plastic sheet. An optical measurement plane was created using two perpendicular 250 Watt halogen lamps that shone into the rise chamber through a narrow slit to produce a light sheet inside the middle of the rise chamber. A VF13-5SP footprint transducer was oriented in line with the optical slits on the side of the rise chamber, coupled to the laminated plastic sheet with ultrasonic gel to eliminate any air gaps. The rest of the validation apparatus was covered with black photographer’s background cloth.
DETECTS™ software installed on a Siemens Acuson/Antares ultrasound machine was used to quantify the number of potential emboli in consecutive images from the VF13-5SP footprint transducer. A High Definition Sony Handycam™ video camera was oriented to view facing down into the rise chamber (ie. normal to the plane of the light sheet). This video camera recorded light reflected from the air bubbles at 250 frames per second as the bubbles passed through the light sheet created inside the rise chamber.
Small amounts of air were introduced into the bubbler by incrementally moving the stepper motor, which forced air through the pneumatic cylinder. Air bubbles with a radius ranging from 0.1 to 1.5 mm produced at the ends of the glass tubes inside the bubbler rose through the rise chamber, and travelled through the measurement plane of the optical and DETECTS™ systems for simultaneous detection by the cameras and the ultrasound transducer. The Matlab (R2009) imaging toolbox was used to analyze the optical frames. A computer program was created to automatically process each of the optical frames. This included thresholding the images using a basic greyscale threshold and then measuring the number and size of the bubbles in the frame; the result produced centre of geometrics and size estimations of the bubble slices in the optical images. A 3D reconstruction similar to that used for DETECTS™ was used to link the individual bubble slices together and estimate the volume of the bubble. The bubbles captured from the ultrasound were analyzed with the DETECTS™ algorithm directly without any modification. The data from the two imaging (optical and ultrasound) techniques were then compared in terms of the number of bubbles, the location of the bubbles, and the bubble radii. Correlation of the optical and ultrasound data used a least squares technique which minimized the total error of three specific parameters (the position, time, and radius differences).
The thresholds to extract each bubble from the ultrasound and optical data were determined. The mechanical index and gain on the ultrasound machine were adjusted to test the sensitivity of the bubble threshold value calculated using DETECTS™ versus the mechanical index and ultrasound image gain. The bubble threshold value was more influenced by the mechanical index than by the image gain, but both results indicated that DETECTS™ is able to extract the bubble threshold value for an extreme range of ultrasound inputs.
Cardiac surgery is commonly performed in the elderly population, which is particularly susceptible to post-operative neurologic dysfunction [11, 12]. Since neurological dysfunction is thought to be in part caused by cerebral air microemboli , efforts are taken by operative teams to reduce microembolism associated with heart surgery . Despite these efforts, air bubbles are frequently observed with TEE after a cardiac surgery patient is separated from CPB . Although TEE alone is a useful tool for determining the presence of potential emboli in the blood postoperatively, and it is commonly used to qualitatively estimate air trapped in the heart, it is unable to accurately quantify the number of bubbles in the heart.
Our new system incorporating the DETECTS™ software algorithm conveniently integrates the use of TEE technology to provide an accurate estimate of the number and size of air bubbles in the heart. Qualitative and quantitative analysis provided promising results from the DETECTS™ algorithm compared to optical measurements. The shape and intensity of the reflected acoustic signal was determined to uniquely characterize the true bubble radius; individual air bubbles could be reliably distinguished from each other. Although there was some noise in the bubble 3D reconstruction due to phantom bubbles, DETECTS™ still showed a strong correlation between optical and ultrasound measurements, with minimal centre of geometry and time differences between the measurement techniques.
By enabling the operative team to visualize the number and size of air emboli in the cardiac chambers and/or aorta in real time, it is our hope that they will adjust their de-airing techniques to further reduce the amount of air trapped in the heart. Ultimately, we believe this will lead to improved de-airing techniques and, in turn, reduced rates of neurological dysfunction in cardiac surgery patients. In the future, DETECTS™ may also prove useful for the evaluation of other cardiac conditions associated with microembolism, and for the detection of bubbles and potentially solid particles in other flow systems.
Future work in the development of DETECTS™ will ideally involve assessment of the software performance using blood products. It was necessary in the current study to use water instead of blood to allow for the use of optical measures to validate DETECTS™ by comparing optical and ultrasound output. Future studies should provide data on the capacity of DETECTS™ to estimate the volume of air in blood. Furthermore, once permitted for in vivo use, the clusters of gaseous emboli visualized by DETECTS™ during cardiac surgery in real time could be correlated to the surgeon’s maneuvers to gauge the most effective technique to remove air from the heart.
Availability and requirements
Project Name: Detection of Emboli using Transesophageal Echocardiography for Counting, Total volume, and Size estimation
Project Homepage: N/A
Operating System(s): Windows XP Eng mode (on ACUSON Antares Ultrasound Machine)
Programming Language: C++
Other Requirements: N/A
License: There is no license for this product; however, it is currently under review for a patent
Any restrictions to use by non-academics: Patent pending
FS is a PhD candidate in the Department of Mechanical and Materials Engineering at Queen’s University, Canada. AP is a Professor and Queen’s Research Chair in Fluid Dynamics and Multi-scale Phenomena in the Department of Mechanical and Materials Engineering at Queen’s University, Canada. CGB has a MSc in Mechanical Engineering from Queen’s University, Canada, and is currently a resident in Anatomical Pathology at the University of Ottawa in the Department of Pathology and Laboratory Medicine, Canada. MU is an Associate Professor in the Department of Mechanical Engineering and Engineering Science at University of North Carolina Charlotte, USA, and is the Director of NC Motorsports and Automotive Research Center, USA. AH is an Associate Professor, Chair of the Cardiac Division, and Deputy Head of the Department of Surgery at Queen’s University, Canada, and is an attending cardiac surgeon at the Kingston General Hospital, Canada. RCT is an Assistant Professor in the Department of Anesthesiology & Perioperative Medicine at Queen’s University, Canada, and is an attending anesthesiologist at the Kingston General Hospital, Canada. JBT holds a PhD in Psychology, Neuroscience & Behaviour, and is a Clinical Research Associate in the Department of Anesthesiology & Perioperative Medicine at Queen’s University, Canada. BM is an Adjunct Professor in the Department of Anesthesiology & Perioperative Medicine at Queen’s University, Canada, Professor Emeritus at Queen’s University, Canada, and a former attending anesthesiologist at the Kingston General Hospital, Canada.
The authors would like to thank Dr. Mohamed J. Ali for his advice and expertise regarding the conception of this idea. This work was supported by grants from NSERC awarded to AP, the Garfield Kelly Cardiovascular Research and Development Fund awarded in part to BM, AP and MU, and a Parteq Proof of Principle Fund/Ontario Ministry of Research and Innovation awarded in part to AP, BM, and AH. These funding agencies had no role in the design, or in the collection, analysis, and interpretation of the data, in the writing of the manuscript, or in the decision to submit the manuscript for publication.
- Gallagher EG, Pearson DT: Ultrasonic identification of sources of gaseous microemboli during open heart surgery.Thorax 1973, 28:295–305. 10.1136/thx.28.3.295View ArticlePubMedPubMed CentralGoogle Scholar
- Abu-Omar Y, Balacumaraswami L, Pigott DW, Matthews PM, Taggart DP: Solid and gaseous cerebral microembolization during off-pump, on-pump, and open cardiac surgery procedures.J Thorac Cardiovasc Surg 2004, 127:1759–65. 10.1016/j.jtcvs.2003.09.048View ArticlePubMedGoogle Scholar
- Barak M, Katz Y: Microbubbles: pathophysiology and clinical implications.Chest 2005, 128:2918–32. 10.1378/chest.128.4.2918View ArticlePubMedGoogle Scholar
- Gerriets T, Schwarz N, Sammer G, Baehr J, Stolz E, Kaps M, et al.: Protecting the brain from gaseous and solid micro-emboli during coronary artery bypass grafting: a randomized controlled trial.Eur Heart J 2010, 31:360–8. 10.1093/eurheartj/ehp178View ArticlePubMedGoogle Scholar
- Pugsley W, Klinger L, Paschalis C, Treasure T, Harrison M, Newman S: The impact of microemboli during cardiopulmonary bypass on neuropsychological functioning.Stroke 1994, 25:1393–9. 10.1161/01.STR.25.7.1393View ArticlePubMedGoogle Scholar
- Borger MA, Peniston CM, Weisel RD, Vasiliou M, Green REA, Feindel CM: Neuropsychologic impairment after coronary artery bypass surgery: effect of gaseous microemboli during perfusionist interventions.J Thorac Cardiovasc Surg 2001, 121:743–9. 10.1067/mtc.2001.112526View ArticlePubMedGoogle Scholar
- Kalpokas MV, Nixon IK, Kluger R, Beilby DS, Silbert BS: Carbon dioxide field flooding versus mechanical de-airing during open-heart surgery: a prospective randomized controlled trial.Perfusion 2003, 18:291–4. 10.1191/0267659103pf671oaView ArticlePubMedGoogle Scholar
- Grosset DG, Georgiadis D, Kelman AW, Cowburn P, Stirling S, Lees KR, et al.: Detection of microemboli by transcranial Doppler ultrasound.Tex Heart Inst J 1996, 23:289–92.PubMedPubMed CentralGoogle Scholar
- Ringelstein EB, Droste DW, Babikian VL, Evans DH, Grosset DG, Kaps M, et al.: Consensus on microembolus detection by TCD.Stroke 1998, 29:725–9. 10.1161/01.STR.29.3.725View ArticlePubMedGoogle Scholar
- Oka Y, Moriwaki KM, Hong Y, Chuculate C, Strom J, Andrews IC, et al.: Detection of air emboli in the left heart by M-mode transesophageal echocardiography following cardiopulmonary bypass.Anesthesiology 1985, 63:109–13. 10.1097/00000542-198507000-00020View ArticlePubMedGoogle Scholar
- Etzioni DA, Starnes VA: The Epidemiology and Economics of Cardiothoracic Surgery in the Elderly. In Cardiothoracic Surgery in the Elderly. Edited by: Katlic MR. New York: Springer New York; 2011:5–24.View ArticleGoogle Scholar
- Ahonen J, Salmenperä M: Brain injury after adult cardiac surgery.Acta Anaesthesiol Scand 2004, 48:4–19. 10.1111/j.1399-6576.2004.00275.xView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.