Opportunity
In hybrid photoacoustic and ultrasound imaging systems, the speed of sound within the imaged sample is a critical reconstruction parameter that directly impacts image quality, including resolution and contrast. Inaccurate speed of sound values lead to artifacts such as feature splitting, fringing, distortion, and blurring, which degrade diagnostic utility. The speed of sound is not a fixed constant; it varies with factors like sample size, temperature, and physiological status, making an accurate, sample-specific value often unknown prior to imaging. Existing techniques for determining this value are problematic. Manual adjustment by operators is slow and subjective, prone to inaccuracies. Automated computational methods often rely on complex, iterative algorithms that are resource-intensive and inefficient, hindering real-time or video-rate imaging applications. This creates a significant barrier to achieving consistently high-quality, co-registered dual-modal images in both preclinical and clinical settings, where rapid, accurate visualization is essential.
Technology
This patent introduces an adaptive system and method for processing photoacoustic and ultrasound imaging data that automatically optimizes the speed of sound value(s) used in image reconstruction. The core innovation lies in leveraging the acquired ultrasound imaging data itself to determine the optimal reconstruction parameter. The technology encompasses two primary optimization strategies. The first involves a single, global speed of sound optimization. The system processes the same set of ultrasound channel data multiple times, each trial using a different candidate speed of sound value. For each trial, it calculates a coherence factor (CF) map, which measures signal phase coherence. The sum of coherence values across the map (total coherence factor) is computed. The candidate speed of sound that yields the maximum total coherence factor is identified as the optimal value, often refined via interpolation of the coherence factor curve. This optimized value is then applied in delay-and-sum beamforming algorithms to reconstruct both the ultrasound and the co-registered photoacoustic images. The second, more advanced strategy performs a two-compartment optimization. It first automatically segments the ultrasound data to identify a boundary, typically between the sample object (e.g., tissue) and the coupling medium (e.g., water). This segmentation uses algorithms like an improved Akaike information criterion picker on transmission channel data to detect first-arrival signal positions. Once segmented, it determines two optimal speed of sound values: one for the sample and one for the coupling medium. The value for the coupling medium can be derived from a temperature-based lookup table. The value for the sample is determined similarly to the first strategy but applied only to the data subset within the sample region, potentially using a multi-stencil fast marching method for time-of-flight calculation and a golden-section search for efficiency. Finally, image reconstruction proceeds using these two distinct speed of sound values in their respective regions, significantly reducing boundary artifacts and improving focus throughout the image.
Advantages
- Automated and Objective: Eliminates slow, subjective manual speed-of-sound tuning, providing consistent, data-driven optimization.
- Enhanced Image Quality: Minimizes reconstruction artifacts (splitting, fringing, blurring), leading to sharper, more accurate images with better-defined features and boundaries.
- Computational Efficiency: The coherence-factor-based search, especially when combined with strategies like golden-section search, is more efficient than complex iterative methods, enabling faster processing.
- Real-Time Potential: The single-value optimization method is sufficiently fast to support video-rate (e.g., 10Hz) dual-modal image reconstruction and display.
- Robustness: The optimization uses ultrasound data, making it insensitive to variables that affect photoacoustic signals, such as optical fluence attenuation.
- Dual-Strategy Flexibility: Offers both a faster single-value method for real-time imaging and a higher-quality dual-value method for detailed analysis, which can be selected based on application needs.
- Automatic Segmentation: The two-compartment method includes an automatic, accurate segmentation of the sample from the coupling medium, removing another manual step.
Applications
- Preclinical Research: High-resolution, dynamic imaging of small animals (e.g., mice, rats) for studying anatomy, physiology, cardiopulmonary dynamics, tumor vascularization, and pharmacokinetics.
- Clinical Diagnostics: Potential for imaging human extremities (e.g., finger joints, breast) to visualize vasculature, inflammation, and other pathologies with complementary contrast.
- Intraoperative Monitoring: Real-time, co-registered imaging during surgical procedures to guide interventions and monitor tissue perfusion.
- Phantom and Sample Imaging: Quality control and characterization of tissue-mimicking phantoms and biological samples in laboratory settings.
- Functional Imaging: Mapping hemodynamic activity by analyzing dynamic photoacoustic signals, such as creating arterial maps based on heartbeat frequency.
