Comparison of three different ultrasonic methods for quantification of intravascular gas bubbles

For evaluating different decompression schedules, the use of ultrasound is common. Systems based on the Doppler principle have mostly been used. However, ultrasonic scanners producing images where the bubbles are easily detected, may be an alternative, because analysis of the signals is simpler than when using Doppler methods. In this study, three methods of bubble detection were used following a series of air dives. The divers were investigated using a "blind" Doppler system where only auditory signals were used for positioning the probe. They were also studied using ultrasonic images and finally an "image-assisted" Doppler method was used, where the sample volume of the Doppler system was positioned using the images. Both Doppler systems were pulsed Doppler systems. The agreement between the methods was determined using weighted kappa statistics. The results show that, at rest, the agreement between the images and the blind Doppler method was very good, and between the two Doppler methods and the images and the image-assisted method the agreement was good. Generally, the agreement is better at higher bubble grades. After movement, the agreement was not good. We conclude that grades from the different methods can be directly compared at rest.