Development of a Multi-Spectral Photoacoustic Imaging System for the Detection of Intracranial Hematoma

Abstract

Head trauma patients are susceptible to secondary injuries where afflicted tissues can propagate towards dysfunction after the initial injury is treated. Monitoring blood oxygenation (SpO2) below the skull is crucial for the early detection of secondary head injury such as hematoma. To obtain such information in a point-of-care setting, photoacoustic (PA) imaging can be used to differentiate optical contrast between hemoglobin (Hb) species by measuring resultant ultrasonic waves emitted by optically irradiated tissue. Given the ratio of Red:NIR light absorption, information regarding SpO2 can be determined in vivo. In this project, computer simulations involving PA imaging of tissue models have been performed and techniques tested using an optic fiber/transducer PA imaging system. In simulated and physical PA scans, Red:NIR ratio values are computed for various tissue models to evaluate optical contrast in target absorbers. Images reconstructed from simulations showed the ability to visualize differences in SpO2 across a layer of skull tissue using multi-spectral optical irradiation. Red:NIR ratios were calculated using PA signals produced by 750 nm light and 850 nm light. Physical image reconstructions were conducted using a 5 ns pulsed laser and near-infrared (NIR) optical parametric oscillator (OPO). The pulse energy used during physical PA raster scans reached up to 5.3 mJ/pulse. Tissue phantoms scanned consisted of optical absorbers surrounded by various tissue-mimicking materials. Images from physical acquisitions were acquired using a 2D CNC-controlled moving stage, but free-hand tracking using inertial and optical sensors have been investigated. At 5.3 mJ/pulse, images reconstructed from physical scans could not resolve optical absorbers positioned beneath a layer of skull-mimicking tissue. A limiting factor contributing to low signal-to-noise ratio (SNR) from inferior absorbers was the percentage of power lost during beam focusing. However, simulation results encourage future work to improve pulse energy output before simulation results can be validated.