Application of Functional Ultrasound Imaging in the Study of the Human Spinal Cord

Background Introduction

The spinal cord is an important center for sensory and motor integration in the nervous system, responsible for monitoring the kinematics and posture of various body parts. Interruption of spinal cord information flow due to injury or disease can lead to a series of adverse outcomes, such as enhanced reflex activity, chronic pain, partial or complete loss of motor or sensory function, and disturbances in bowel/bladder function. Despite the importance of the spinal cord in sensory, motor, and autonomic functions, research on its functional architecture remains limited. Currently, functional brain imaging technologies (such as fMRI and stereotactic electroencephalography) are widely used in brain science research but have limited application in spinal cord studies. Since the late 1990s, when functional magnetic resonance imaging (fMRI) was introduced into the field of spinal cord research, only a few studies have revealed the functional organization of the spinal cord. However, challenges in imaging techniques (such as the small cross-sectional area of the spinal cord, magnetic field inhomogeneity, and motion artifacts) have limited its progress.

Functional Ultrasound Imaging (FUSI) is an emerging imaging technology with advantages such as wide coverage, high temporal and spatial resolution, and high sensitivity. It was first validated in cerebral blood volume imaging in mice and has gradually been extended to freely moving rodents, awake and active non-human primates, as well as brain imaging in adult and pediatric patients. Recent studies have shown that functional ultrasound imaging can detect hemodynamic responses of the spinal cord induced by electrical stimulation, but its effectiveness has yet to be verified in humans.

Paper Source

The paper, titled “Functional Ultrasound Imaging of the Human Spinal Cord,” was authored by K.A. Agyeman and others from institutions including the University of California, Riverside, and the University of Southern California Medical Center. It was published in the journal “Neuron” on May 15, 2024, with the DOI: 10.1016/j.neuron.2024.02.012.

Research Process and Methods

Research Process

The study was conducted on six patients who underwent standard spinal cord stimulator implant surgery due to chronic back pain, to collect functional ultrasound imaging data. The research process included the following steps:

1. Patient Selection and Surgical Preparation: Patients with chronic back pain were selected and underwent partial laminectomy at the T10 level for the implantation of spinal cord stimulation electrodes.
2. Installation of Functional Ultrasound Imaging Probe: A 15 MHz linear array probe was inserted into the surgical area to obtain functional ultrasound imaging images in a transverse view.
3. Electrical Stimulation Experiment: Hemodynamic data of the spinal cord were collected under stimulated and non-stimulated conditions, with electrical stimulation parameters including a frequency of 40 Hz, pulse width of 250 ms, and current intensity of 3.0 mA and 4.5 mA.
4. Data Collection and Processing: Continuous ultrasound images were collected and underwent motion artifact correction, low-frequency filtering, and other preprocessing.

Data Analysis and Methods

1. Hemodynamic Change Analysis: Statistical Parametric Mapping (SPM) and Event-Related Averaging (ERA) methods were used to calculate local spinal cord blood volume changes induced by stimulation and generate hemodynamic change maps.
2. Single Trial Decoding: Machine learning algorithms (such as Principal Component Analysis and Linear Discriminant Analysis) were used to reduce dimensionality and classify single-trial data to assess the effectiveness of stimulation protocols.
3. Spatial Resolution and Blood Flow Analysis: Image resolution was reduced to 200 microns and 400 microns to evaluate decoding accuracy at different resolutions. Blood flow signals were sorted by size to analyze the influence of different vessel sizes on decoding accuracy.

Research Results

Hemodynamic Response

The study found that electrical stimulation caused significant spatiotemporal hemodynamic changes in the spinal cord, mainly manifested as localized increases or decreases in blood flow. Statistical analysis showed significant changes in local spinal cord blood volume during stimulation, with these changes gradually returning to baseline levels after the stimulation ended.

Decoding Analysis

Through machine learning algorithms, the study successfully decoded spinal cord states in single trials with an accuracy exceeding 90%. Specifically, blood flow changes in different regions showed different response peaks and recovery times, revealing complex hemodynamic responses within the spinal cord.

Spatial Resolution Analysis

The study showed that high spatial resolution (100 microns) was more effective in detecting and decoding spinal cord blood flow changes than lower spatial resolutions (200 microns and 400 microns). The most informative blood flow signals mostly came from small (medium-small) vessels.

Conclusion and Significance

Research Conclusion

The study successfully quantified hemodynamic responses of the spinal cord to electrical stimulation using functional ultrasound imaging technology for the first time in humans. High-precision single-trial decoding demonstrated the potential of functional ultrasound imaging technology for applications in real-time closed-loop neuromodulation systems.

Scientific and Application Value

This study provides important experimental evidence for the application of functional ultrasound imaging technology in spinal cord research. By revealing hemodynamic changes in the spinal cord induced by electrical stimulation, the study offers new perspectives for understanding spinal cord function and its clinical neuromodulation effects. Additionally, functional ultrasound imaging technology can be used to optimize spinal cord stimulation parameters to enhance the effectiveness of neuromodulation therapy, indicating broad application prospects.

Research Highlights

• Applied functional ultrasound imaging technology to the human spinal cord for the first time, successfully quantifying spinal cord hemodynamic responses.
• Achieved high-precision single-trial decoding, providing technical support for the development of real-time closed-loop neuromodulation systems.
• Revealed the impact of different vessel sizes on blood flow signal decoding, providing new insights for optimizing spinal cord stimulation parameters.

Prospects and Challenges

Although functional ultrasound imaging technology demonstrated great potential in this study, there are still some challenges to its clinical application, such as the invasiveness of the imaging process and image quality issues. In the future, non-invasive functional ultrasound imaging technologies (such as low-frequency imaging with intravenous microbubble contrast agents and phased array transducers) are expected to further broaden its application scope. Moreover, developing thinner ultrasound probes to fit the limited space in clinical surgeries will also improve imaging quality and operational convenience. The application of functional ultrasound imaging technology in the human spinal cord provides a new tool for studying spinal cord function and its clinical modulation. With continuous optimization of technology and data processing methods, this technology is expected to play an important role in neuroscience research and clinical neuromodulation therapy.