L.B. Christie, W.V. Giegerich, P. Schneider, K.W. Oh
University at Buffalo,
United States
Keywords: Photoacoustics, Test Phantom
Summary:
This research focuses on the design, manufacturing, and testing of test phantoms mimicking human finger vasculature for use in photoacoustic (PA) applications. Phantoms were designed for modularity, allowing the researcher to adjust channel diameter, length, depth, quantity, distention, pulse waveform, and material in order to meet any testing parameters. A series of different phantoms were created, tested with PA, and the results were compared to those from live human finger signals. Prior Work: PA imaging modalities present unique qualities to the medical imaging space by taking advantage of optical and acoustic imaging methods. This allows for medical scans of physiological features to be taken at high resolution. A current downfall of PA is a lack of quantification of system performance due to the absence of devices such as test phantoms. Test phantoms are tools with known parameters used by an operator to quantify the system. Previous research outlines the use of test targets created with microfluidics for quantification of PA resolution as a function of depth [1]. This research improves upon that method by introducing anatomically/physiologically correct phantoms capable of showing system resolution with respect to the features found in the human body. Methods: Test phantoms were designed to mimic the physiological characteristics of the human finger. Several material properties considered for the phantom design were extinction coefficient (combination of optical absorption and optical scattering coefficient), optical index of refraction, acoustic impedance, density, speed of sound, acoustic attenuation, Young’s Modulus, and thermal expansion. These parameters were tuned to closely match those of the human finger. This impacted the choice of materials for multiple different channels, phantom silicone rubbers, 3D printing PLA, and PA blood simulant (a custom manufactured mixture of black ink, water, and glycerin). An example of a test phantom containing three channels at staggered depths and a custom 3D printed shell can be seen in Figure 1. A synthetic heart rate waveform was modeled after a human heart waveform captured from a photoplethysmography sensor placed on the finger. This waveform was pulsed throughout the phantom channel using a micro actuating pump capable of pulsing uL/min with high accuracy and a PA signal was captured (Figure 2). Experimentation and Results: Test phantoms were tested using a PA testbench consisting of laser light source, pump, and acoustic receiver. Several targets were created with a channel diameter of 1.2mm and channel depths of 2mm, 5mm, and 8mm with respect to the acoustic sensor. These were filled with a custom designed PA blood simulant and a signal was captured for each static phantom (Figure 3). The peaks of these signals agreed with the mathematical calculations for the depth of each channel. Heart rate waveforms were also pulsed through a single artery phantom for 20s. A heart rate PA signal that was outputted by the phantom was compared to that of a human finger (Figure 4).