Functional ultrasound imaging (fUSI) is a developing brain imaging modality currently being tested in human studies. Unlike structural ultrasound imaging, which creates static images, fUSI captures rapid sequences of brain tissue images, tracking changes in blood flow associated with neural activity. This technique has the potential to offer ten times higher linear resolution compared to functional MRI, is smaller and less expensive since it does not require magnets, and has a higher signal-to-noise ratio.
fUSI can be paired with low-intensity focused ultrasound (LIFU) to modulate brain activity, presenting new treatment options for neuropsychiatric and neurological disorders. Future developments may include the use of contrast agents or gene therapies to measure vasculature at super-resolution or directly assess neural activity. Although ultrasound has difficulty penetrating the skull, fUSI has been successfully demonstrated in intraoperative settings and in patients with acoustically transparent cranial implants following skull surgery.
The author conducted extensive research on fUSI, noting that much of the technical information is often buried in application papers or lengthy textbooks. The article aims to clarify how fUSI operates and its analytical methods. fUSI utilizes transducers to transmit and receive ultrasound waves, which backscatter when encountering changes in tissue density, such as red blood cells. By analyzing these backscattered waves, researchers can reconstruct images that reflect local changes in impedance.
The resolution of these images is determined by the wavelength of the ultrasound, allowing for high-resolution structural imaging. For functional imaging, fUSI capitalizes on the hemodynamic response, where neuronal activity leads to increased blood flow and changes in blood properties. fUSI tracks these changes using the interference patterns of moving red blood cells, acquiring images rapidly—over a kHz—and measuring changes in speckle patterns, which correlate with blood flow and volume.
The article provides an analogy comparing fUSI to high-speed video analysis, illustrating how movement can be quantified through pixel brightness variations. However, it also acknowledges potential flaws in this approach, such as decreased signal-to-noise ratio and motion artifacts, which are challenges faced by both video analysis and fUSI.
fUSI employs a multiplexed strategy, transmitting unfocused plane waves to capture entire imaging planes at once, allowing for rapid image acquisition. The reconstruction of images involves complex algorithms that infer scatterer density from the time series data collected by each transducer. The article discusses the Power Doppler (PD) estimate, which measures changes in image texture as a proxy for blood volume, while also addressing the limitations of this method, particularly its sensitivity to global movement.
To improve accuracy, the article suggests using motion estimation techniques and exploring alternative methods for analyzing blood flow, such as autocorrelation functions. It also mentions the potential for deep learning methods to enhance image sequence analysis, although the lack of available data poses a challenge for training such models.
The author concludes by discussing the application of fUSI in analyzing motor cortex activity, suggesting that time-locked external signals can be effectively studied using statistical models similar to those used in fMRI analysis.