Microbubble detection following hyperbaric chamber dives using dual-frequency ultrasound.

Venous gas emboli (VGE) can be readily detected in the bloodstream using existing ultrasound methods. No method currently exists to detect decompression-induced microbubbles in tissue. We hypothesized that dual-frequency ultrasound (DFU) could detect these microbubbles. With DFU, microbubbles are driven with two frequencies: a lower "pump" (set to the resonant frequency of the desired bubble size) and a higher "image" frequency. A bubble of the resonant size emits the sum and difference of the two transmitted frequencies. For this study we used a pump frequency of 2.25 MHz and an image frequency of 5.0 MHz, which detects bubbles of roughly 1-10 µm in diameter in a water tank. Four anesthetized swine were pressurized at 4.5 ATA for 2 h and decompressed over 5 min, inducing moderate to very severe VGE scores. Four sites on the thigh of each swine were monitored with DFU before and after the dives. A single mock dive was also performed. The number of sites returning signals consistent with microbubbles increased dramatically after the chamber dive (P < 0.01), but did not change with the mock dive. The increase in DFU signal after the chamber dive was sustained and present at multiple sites in multiple swine. This research shows for the first time that decompression-induced tissue microbubbles can be detected using DFU and that DFU could be used to monitor decompression-induced microbubbles at multiple sites on the body. Additionally, DFU could be used to track the time course of microbubble formation and growth during decompression stress.

Signals consistent with microbubbles detected in legs of normal human subjects after exercise.

Exercise may produce micronuclei (presumably gas-filled bubbles) in tissue, which could serve as nucleation sites for bubbles during subsequent decompression stress. These micronuclei have never been directly detected in humans. Dual-frequency ultrasound (DFU) is a resonance-based, ultrasound technique capable of detecting and sizing small stationary bubbles. We surveyed for bubbles in the legs of six normal human subjects (ages 28-52 yr) after exercise using DFU. Eleven marked sites on the left thigh and calf were imaged using standard imaging ultrasound. Subjects then rested in a reclining chair for 2 h before exercise. For the hour before exercise, a series of baseline measurements was taken at each site using DFU. At least six baseline measurements were taken at each site. Subjects exercised at 80% of their age-adjusted maximal heart rate for 30 min on an upright bicycle ergometer. After exercise, the subjects returned to the chair, and multiple postexercise measurements were taken at the marked sites. Measurements continued until no further signals consistent with bubbles were returned or 1 h had elapsed. All subjects showed signals consistent with bubbles after exercise at at least one site. The percentage of sites in a given subject showing signals significantly greater than baseline (P < 0.01) at first measurement ranged from 9.1 to 100%. Overall, 58% of sites showed signals consistent with bubbles at the first postexercise measurement. Signals decreased over time after exercise. These data strongly suggest that exercise produces bubbles detectable using DFU.

Dual-frequency ultrasound detection of stationary microbubbles in tissue.

Indirect evidence suggests that microbubbles that exist normally in tissue may play a key role in decompression sickness (DCS). Their sizes and locations are unknown. Dual-frequency ultrasound (DFU) exploits bubble resonance to detect bubbles over a wide size range and could potentially detect stationary tissue microbubbles. To test this capability, DFU was used to detect stationary microbubbles of known size (2-3 microm mean diameter) over a range of ultrasound pressures and microbubble concentrations. In gelatin phantoms doped with microbubbles and in ex vivo porcine tissue, signals indicative of bubbles were detected for microbubble concentrations of 5x10(5) per mL and greater. Signals were not returned from solid particle microspheres of similar size to the microbubbles or from saline controls. In the thigh of an anesthetized swine, signals were detected for concentrations of 5x10(7) per mL and greater. Because of its ability to detect bubbles over a wide range of sizes, this technique could potentially detect naturally-existing microbubbles in tissue and lead to (a) an improved understanding of the mechanics of bubble formation during decompression and (b) a new metric for evaluating DCS.