Opto-acoustic imaging involves using light to produce sound waves for visualizing blood in biological tissue. By using two different optical wavelengths, diagnostic images of blood oxygen saturation can be generated using endogenous optical contrast, without injection of any external contrast agent and without using any ionizing radiation. The technology has been used in recent clinical studies for diagnosis of breast cancer to help distinguish benign from malignant lesions, potentially reducing the need for biopsy through improved diagnostic imaging accuracy. To enable this application, techniques that can accurately map and effectively display small differences in oxygen saturation are necessary.
We analyze the ability of an opto-acoustic imaging system to display a colorized parametric map for oxygen saturation of blood using biologically-relevant opto-acoustic phantoms. The relationship between colorized image output and known oxygen saturation values is examined. To mimic breast tissue, a material with closely matching properties for optical absorption, optical scattering, acoustic attenuation and speed of sound is used. The phantoms include two vessels filled with whole blood at oxygen saturation levels determined using a gold-standard sensor-based approach. A flow system with gas-mixer and membrane oxygenator adjusts the oxygen saturation of each vessel independently. Data is collected with an investigational Imagio breast imaging system. An opto- acoustic relative color map is generated using a novel statistical mapping approach. In addition, we propose a technique to characterize the ability to distinguish small differences in oxygen saturation as the oxygenation level is varied. When applied to the phantom, with reference vessel at 99% saturation, hematocrit of 42% and depth of 1.5cm, a contrast distinction threshold was reached when an adjusted vessel achieved a difference of approximately 4.6% saturation compared to the reference.