Tissue perfusion rate estimation with compression-based photoacoustic-ultrasound imaging

Characterizing a photoacoustic and fluorescence imaging platform for preclinical murine longitudinal studies

Significance

To effectively study preclinical animal models, medical imaging technology must be developed with a high enough resolution and sensitivity to perform anatomical, functional, and molecular assessments. Photoacoustic (PA) tomography provides high resolution and specificity, and fluorescence (FL) molecular tomography provides high sensitivity; the combination of these imaging modes will enable a wide range of research applications to be studied in small animals.

Aim

We introduce and characterize a dual-modality PA and FL imaging platform using in vivo and phantom experiments.

Approach

The imaging platform's detection limits were characterized through phantom studies that determined the PA spatial resolution, PA sensitivity, optical spatial resolution, and FL sensitivity.

Results

The system characterization yielded a PA spatial resolution of 173 ± 17    μ m in the transverse plane and 640 ± 120    μ m in the longitudinal axis, a PA sensitivity detection limit not less than that of a sample with absorption coefficient μ a = 0.258    cm - 1 , an optical spatial resolution of 70    μ m in the vertical axis and 112    μ m in the horizontal axis, and a FL sensitivity detection limit not < 0.9    μ M concentration of IR-800. The scanned animals displayed in three-dimensional renders showed high-resolution anatomical detail of organs.

Conclusions

The combined PA and FL imaging system has been characterized and has demonstrated its ability to image mice in vivo, proving its suitability for biomedical imaging research applications.

Monitoring peripheral hemodynamic response to changes in blood pressure via photoacoustic imaging

Chronic wounds and amputations are common in chronic kidney disease patients needing hemodialysis (HD). HD is often complicated by drops in blood pressure (BP) called intra-dialytic hypotension. Whether intra-dialytic hypotension is associated with detectable changes in foot perfusion, a risk factor for wound formation and impaired healing remains unknown. Photoacoustic (PA) imaging is ideally suited to study perfusion changes. We scanned the feet of 20 HD and 11 healthy subjects. HD patients were scanned before and after a dialysis session whereas healthy subjects were scanned twice at rest and once after a 10 min exercise period while BP was elevated. Healthy (r = 0.70, p < 0.0001) and HD subjects (r = 0.43, p < 0.01) showed a significant correlation between PA intensity and systolic BP. Furthermore, HD cohort showed a significantly reduced PA response to changes in BP compared to the healthy controls (p < 0.0001), showing that PA can monitor hemodynamic changes due to changes in BP.

Towards microvascular pressure estimation using ultrasound and photoacoustic imaging

Microvascular pressure drives perfusion in tissues but is difficult to measure. A method is proposed here to estimate relative pressures in microvessels using photoacoustic and ultrasound tracking of small vessels during calibrated tissue compression. A photoacoustic-ultrasound dual imaging transducer is used to directly compress on tissue in vivo. Photoacoustic signals from blood vessels diminish as an external load is applied and eventually reaches a minimum or vanishes when external pressure is sufficiently greater than the internal pressure. Two methods were proposed to estimate relative pressures. In the first approach, vessels were tracked during compression and when the vessel photoacoustic signals vanished below a set threshold, the internal pressures were assigned as the external loading pressure at the respective collapse point. In this approach pressures required to collapse vessel signatures completely were found to be much greater than physiological blood pressures. An alternative approach was to track the cross-sectional area of small vessels with changing external load and fitting the data to a known Shapiro model for thin-walled vessel compression. This approach produced estimates of internal pressures which were much more realistic. Both approaches produced the same rank-ordering of relative pressures of various vessels in vivo. Approaches thus far require future work to become fully quantitative but the present contributions represent steps towards this goal.