PROCEEDINGS OF SPIE
SPIEDigitalLibrary.org/conference-proceedings-of-spie
Using color intensity projections to
visualize air flow in operating theaters
with the goal of reducing infections
Keith S. Cover, Niek van Asperen, Joost de Jong, Rudolf
M. Verdaasdonk
Keith S. Cover, Niek van Asperen, Joost de Jong, Rudolf M. Verdaasdonk,
"Using color intensity projections to visualize air flow in operating theaters with
the goal of reducing infections ," Proc. SPIE 8572, Advanced Biomedical and
Clinical Diagnostic Systems XI, 857215 (22 March 2013); doi:
10.1117/12.2002777
Using color intensity projections to visualize air flow in operating
theaters with the goal of reducing infections
Keith S. Cover*, Niek van Asperen, Joost de Jong, Rudolf M. Verdaasdonk
Department of Physics and Medical Technology, VU University Medical Center,
Amsterdam, The Netherlands
ABSTRACT
Infection following neurosurgery is all too common. One possible source of infection is the transportation of dust and other contaminates into the open wound by airflow within the operating theatre. While many modern operating theatres have a filtered, uniform and gentle flow of air cascading down over the operating table from a large area fan in the ceiling, many obstacles might introduce turbulence into the laminar flow including lights, equipment and personal. Schlieren imaging - which is sensitive to small disturbances in the laminar flow such as breathing and turbulence caused by air warmed by a hand at body temperature – was used to image the air flow due to activities in an operating theatre. Color intensity projections (CIPs) were employed to reduce the workload of analyzing the large amount of video data. CIPs – which has been applied to images in angiography, 4D CT, nuclear medicine and astronomy – summarizes the changes over many gray scale images in a single color image in a way which most interpreters find intuitive. CIPs uses the hue, saturation and brightness of the color image to encode the summary. Imaging in an operating theatre showed substantial disruptions to the airflow due to equipment such as the lighting. When these disruptions are combined with such minor factors as heat from the hand, reversal of the preferred airflow patterns can occur. These reversals of preferred airflow patterns have the potential to transport contaminates into the open wound. Further study is required to understand both the frequency of the reversed airflow patterns and the impact they may have on infection rates.
Keywords: Surgical infections, color intensity projections, air flow, medical imaging, Schlieren imaging.
1. INTRODUCTION
A great deal of effort and costs in surgery is invested in reducing the chance of infection. However between 1% to 5% of surgeries still result in infection. As detailed in another study [1] major potential source of infection is air blowing dust and other contaminates into open wounds. All modern operating rooms (OR) have a large air conditioning systems in the ceiling over the OR table that blows a flow down over the table called the laminar flow. However, lights, surgeons, residents, nurses and equipment all can disrupt air flow potentially leading to infection.
It is possible to image air flow in an OR with Schlieren imaging [1-4]. However, manually reviewing the Schlieren air flow video over many hours of surgery, perhaps from multiple cameras, for events where contaminates may be blow into the open wounds is a tedious and demanding task.
Figure 1 shows a selection of Schlieren images from a 4 second Schlieren video of a hand in the air flow generated by the air conditioning system in the ceiling of an OR. A hand in the laminar flow of an OR offers a simple example of complicated flow patterns that may arise during an OR. The smooth laminar flow of the air from the ceiling to the top of the hand is clearly visible. Equally visible is the turbulence below the hand due to the laminar flow being broken up by the hand.
Figure 1. Selected images from a sequence of 111 grayscale images acquired over 4 seconds of a hand in the laminar flow in an operating room.
2. METHOD
As demonstrated by Cover et al. color intensity projects (CIPs) is a method for summarizing a sequence of grayscale images in a signal color image with the hue, saturation and brightness of the color encoding the changes over the sequence [5-8]. Since its initial publication, the usefulness of CIPs has been demonstrated in angiography studies [9-10] and is being integrated in some of the standard angiography packages. Its usefulness has also been demonstrated in 4D CT [5]. It has also been applied in other fields including astronomy [7].
Two different versions of CIPs have been used – percentage time and arrival time. In percentage time CIPs the hue encodes the percentage time a signal spends at a location. For example, due to respiration-induced movement, a tumor may be present at a certain location only for part of the respiratory cycle - for example at end-expiration - while at the more central regions the tumor will be present throughout the respiratory cycle.
In arrival time CIPs the hue of the color (red-yellow-green-light blue-blue-purple) encodes the time that a signal arrives at a location in a sequence of images, for example, when contrast arrives at a location in digital subtraction angiography. For implementation of percent time CIPs in the current paper, the standard encoding of the hue will be used. The hue used will be red-yellow-green-light blue-blue. Red will represent pixels that are bright about 10% of the time. Green will represent pixels that are bright about 50% of the time and blue will represent pixels that are bright about 90% of the time.
For implementing of arrival time CIPs in the current paper, hue will range between red and blue with no hue cycling. Red will represent the earliest arrival time corresponding to the first image and blue will represent the latest arrival time.
3. RESULTS
The results for percentage time CIPs applied to air flow of Figure 1 are shown in Figure 2. The percentage time CIPs
Figure 2. Percentage time CIPs of the sequences of images used in Fig 1.
When compared to the air flow well below the hand, most of the air flow above the hand is of fairly uniform intensity with little color saturation. While there is a wide variation of hue, the low saturation indicates little variation in intensity. As we know there is laminar flow about the hand generated by the ceiling air conditioning system. Thus, the CIPs pattern above the hand is indicative with that of laminar flow.
The air flow pattern well below the hand has green and blue hues with high saturation. The high saturation is indicative of the intensity of the pixels varying almost the full amplitude of the pixels, as is expected from turbulence. As green indicates pixels were bright about 50 % of the time and blue indicates 90% of the time, the hue indicates that most pixels where bright most of the time. Since in Schlieren imaging, pixels are dark unless there is some sort of turbulence this shows information about both the air flow and the motion of the hand over the 4 seconds of the original Schlieren grayscale video.
indicates a large amount of turbulence. Examination of the greyscale images confirms the turbulence in the region well below the hand.
In the few millimeters above the hand there are highly saturated yellow, green and blue. This is due to the brief upward motion of the hand at the end of the greyscale video. As yellow, green and blue indicate 25%, 50% and 90% of the time the pixels were bright, it is clear the top of the hand was just briefly in the blue region.
The color at the bottom of the hand looks much different than those at the top. There is a wide red band at the bottom of the hand indicates the pixel were bright about 10% of the time. The motion which explains both the red band at the bottom on the hand and the yellow-green-blue at the top of the hand is a brief top and upward motion of the hand. This motion can be confirmed by inspection of the greyscale moving.
Thus, the percent time CIPs provides a concise summary of the motion of air and the hand over the 4 second video.
Figure 3. Arrival time CIPs of the sequences of images used in Fig 1.
However, in the regions well above and below the hand the arrival time CIPs shown in Figure 3 has a much wider variety of hues than the percentage time CIPs. This is because hue in the arrival time CIPs is determined by the timing of the maximum pixel brightness. Both above and below the hand the timing is random thus yielding wide varieties of hues. For the arrival time CIPs in Figure 3, the saturation well above the hand is much less than well below the hand, which is the same as with the percentage time CIPs. It makes sense that both the brightness and saturation for both the percentage time and arrival time CIPs are the same since they use identical equations (Cover et al. 2013).
In Figure 3, the hue band just below the hand once again indicates a sudden upward and forward motion of the hand. However, in contrast to the percentage time CIPs, the arrival time CIPs allows the timing of the sudden motion to be determined. The blue band at the bottom of the hand indicates the maximum brightness was at the end of the 4 second clip. This is consistent with the percentage time CIPs which indicates the band just below the hand was bright about 10% of the time. Close inspection of the greyscale video also confirms this.
That there is no band on the top of the hand corresponding to the lower blue band in Figure 3 illustrates an important point on how arrival time CIPs works. As the hand moved up during the last 10% of the video, the band just above the hand was bright for the first 90% of the video. Thus the maximum value of each of the pixels in the upper band are randomly located during the first 90% of the video. And since the hue is determined by the timing of the brightest value of a pixel, the hue of the upper band is a mixture of hues rather than the single blue hue in the lower band.
4. DISCUSSION AND CONCLUSION
Examples of both arrival time and percentage time CIPs have been presented for visualizing air flow in an OR. Both percentage time and arrival time CIPs clearly show the difference between laminar and turbulent air flow.
While structures, such as the OR lights and equipment may generate constant areas of turbulence or even reverse flow, other events may be transient – only lasting a few seconds – but still introducing substantial contamination into the wound. Example of such transients events include opening of doors, moving equipment including OR lights, and movement of staff. A CIPs image can summarize the flow over the many hours of a surgery in a single image and provide the location in the room of a transient event. In addition, the time of the transient event can be obtained in a couple of ways.
If an arrival time CIPs is used then the hue of the image will encode the time of a single transient event. However, if there are multiple transient events at the same location then arrival time CIPs may only provide the arrival time of the maximum intensity for each pixel in the events.
A second, and more versatile way, to determine the time of a transient event is to break the Schlieren video of the air flow into segments and calculate a CIPs of each segment. Any segments without the transient can be ignored. But those segments recording the transient can be broken down further to sub segments. The sub segments can then be summarized as CIPs and inspected for transients. This process can be applied recursively until the timing of the transient is narrowed down to the desired accuracy.
Arrival time CIPs works well when contrast is introduced into the flow. For example, in angiography x-ray opaque contrast is injected into the subject to measure the flow of blood through the subject. Unfortunately the demands of OR prevent contrast agents such as smoke or soap bubbles being introduced into the air flow.
Verdaasdonk et al. [1] have suggested a novel method for introducing contrast into laminar air flow generated by the ceiling air conditioning system in the OR. It should be possible to introduce small bubbles of cold air into the laminar flow just below the air conditioning. The bubbles should remain intact in the laminar flow until they hit turbulence. Thus the cold air bubbles should provide more information on both the laminar flow and the turbulence.
While the data presented are only from initial tests of monitoring air flow in ORs, the results are encouraging. CIPs hold the potential of being a valuable tool to speed up the analysis of Schlieren air flow videos in ORs.
5. ACKNOWLEDGMENTS
This work was funded by the VU University Medical Center in Amsterdam.
Disclosure: The authors’ employer, the VU University Medical Center Amsterdam, is pursuing patents on color
intensity projections covering applications in astronomy, medical imaging, and other fields. The first author of this paper have a financial interest in the patents.
REFERENCE
[1] Verdaasdonk, R. M., Jong, J., Van der Veen, J. A., Cover, K. S., “Large field air flow visualization in the operating room to study potential sources for contaminations during surgery,” Proc. SPIE 8573-19, (2013).
[2] Merzkirch, W., [Flow visualization] Academic Press, New York (1987).
[3] Settles, G. S., [Schlieren and shadowgraph techniques: Visualizing phenomena in transparent media.] Berlin: Springer-Verlag, Berlin, (2001).
[5] Cover, K. S., Lagerwaard, F. J., and Senan, S., “Color intensity projects: a rapid approach for evaluating four-dimensional CT scans in treatment planning,” Int. J. Radiat. Oncol. Biol. Phys. 64, 954 (2007).
[6] Cover, K. S., Lagerwaard, F. J., van den Berg, R., Buis, D. R. and Slotman, B. J., “Color intensity projection of digital subtracted angiography for the visualization of brain arteriovenous malformations,“ Neurosurgery 60, 511-515 (2007).
[7] Cover, K. S., Lagerwaard, F. J. and Senan, S., “Color Intensity Projections: A Simple Way to Display Changes in Astronomical Images ,” Publications of the Astronomical Society of the Pacific 119, 523–526 (2007).
[8] Cover, K. S., Lagerwaard, F. J., Verdaasdonk, R. M., “Color intensity projections with hue cycling for intuitive and compressed presentation of motion in medical imaging modalities,” Proc. SPIE 8574-25, (2013).
[9] Strother, C. M., Bender, F., Deuerling-Zheng, Y., Royalty, K., Pulfer, K. A., Baumgart, J., Zellerhoff, M., Aagaard-Kienitz, B., Niemann, D. B. and Lindstrom, M. L,. “Parametric color coding of digital subtraction angiography.” AJNR Am J Neuroradiol 31, 919–924 (2010).
[10] Lin, C. J., Hung, S. C., Guo, W. Y., Chang, F. C., Luo, C. B., Beilner, J., Kowarschik M., Chu W. F., Chang C. Y., “Monitoring peri-therapeutic cerebral circulation time: a feasibility study using color-coded quantitative DSA in patients with steno-occlusive arterial disease,“ Am J Neuroradiol 33, 1685–1690 (2012).
