Room: Karl Dean Ballroom A1
Purpose: To optimize the iThera Medical MSOT inVision 256-TF photoacoustic (PA) imaging system to quantitatively assess oxygen saturation (SOâ‚‚) in murine femoral bone marrow for the purpose of monitoring leukemia progression and treatment response.
Methods: C57 albino mice were imaged on the MSOT system; matched x-ray CT data were then co-registered with 800-nm PA data to confirm co-localization of the PA signature with the femoral bone marrow cavity. To select optimal parameters for linear unmixing of hemoglobin and calculation of SOâ‚‚, mice were imaged with different combinations of wavelengths (715, 730, 745, 760, 800, 850, 865, 880 nm) and frame averaging (5, 10, 20 averages per frame). An oxygen challenge (100% vs. 21% Oâ‚‚) was performed to measure the frame rate needed to monitor SOâ‚‚ changes during changing inhalation conditions. Additionally, in a pilot study, three control (wild-type) and four leukemic mice were imaged; PA-based SOâ‚‚ estimates over the length of the femur were compared across five time-points.
Results: Between two- and five-wavelength unmixing, the mean SOâ‚‚ estimate differed by 4%; increasing the number of frame averages decreased variation in SOâ‚‚ estimates, but it did not statistically significantly (Î±=0.05) change the mean SOâ‚‚ estimate. In the oxygen challenge, the transient was observed to last longer than one minute; therefore, we determined that the five-wavelength/five-averages combination is acceptable (frame rate=2.5 seconds). In the pilot study, changes in mean SOâ‚‚ over the length of the femur increased from 5% (between baseline time-points) to 20% (between the terminal time-point and baseline) in leukemic mice.
Conclusion: We have designed an imaging protocol to obtain repeatable PA-based SOâ‚‚ estimates in murine femoral bone marrow. Results indicate that we can achieve repeated measure variability below 5% SOâ‚‚ in wild-type mice, which is far below pathologic-based variation (20%) observed in a pilot cohort of leukemic mice.