Researchers image visual cortex with diffuse optical tomography
Technology could help assess brain function
Scientists at Washington University School of Medicine in St. Louis and at the University of Exeter in the UK have developed an optical method that is portable and that has the potential to assess brain function in a wide range of patients, from newborns to the critically ill in intensive care. The system relies on diffuse optical tomography and measures brain function by imaging changes in oxygenated hemoglobin, in deoxygenated hemoglobin and in total hemoglobin.
Using diffuse optical tomography, researchers mapped the response in the visual cortex as the polar angle changed on a video screen. Images courtesy of Joseph P. Culver.
According to principal investigator Joseph P. Culver, diffuse optical tomography offers important advantages over MRI or PET scanning. “Diffuse optical tomography provides a cap-based approach to brain imaging that is portable and wearable,” he explained. “These features are important for imaging subjects not amenable to the fixed enclosures of MRI and PET, [such as] children sitting on their parents’ laps or patients in intensive care units.”
Most attempts to use optical methods to perform functional brain imaging rely on topographic systems that synthesize an image from measurements at a single source-detector pair. They do not collect signals from other source-detector pairs that allow for calculations to determine scattering and absorption through various depths of tissue. As a result, these systems have limited lateral resolution and are unable to section the image according to depth. Researchers elsewhere have successfully used diffuse optical tomography to image breast cancer. However, using the technology for brain imaging has proved challenging because light attenuates very quickly, Culver explained.
Researchers developed a molded imaging cap containing a high-density fiber optic setup of 24 source fibers (red) and 28 detector fibers (blue) arranged in a grid measuring 5.2 × 2.6 in. The grid collects signals from the first-, second-, third- and fourth-nearest neighbor source-detector pair as indicated by green lines.
The new system uses a high-density fiber optic setup of 24 source fibers and 28 detector fibers arranged in a grid measuring 5.2 × 2.6 in. The grid is fixed inside an imaging cap that is molded to fit the back of the subject’s head, above the visual cortex. Culver and his group set up the grid to collect first-, second-, third- and fourth-nearest neighbor source-detector pairs with separations of 13, 30, 40 and 48 mm, respectively. Each detector collects light from each source, allowing a total of 348 measurements with acceptable signal-to-noise ratios.
As published in the July 17 issue of PNAS, the researchers improved the dynamic gain by digitizing the signal using isolated detector channels, each with a dedicated 24-bit analog-to-digital converter, as well as dedicated control lines for each source. This setup creates high-dynamic-range and extremely low channel crosstalk. It also has a much faster frame rate (12 Hz) and higher dynamic range (120 dB) than 16-bit time-shared multiplexed systems.
As a light source, the group used two types of Roithner Laser LEDs operating at 750 and 850 nm modulated through high-bandwidth digital input/output lines from National Instruments. The researchers used optical fiber bundles from CeramOptec to relay light from the LEDs to the optical cap. Each source fiber delivered 0.2 mW of power to the subjects at each wavelength. The group also gave each detection channel its own dedicated Hamamatsu avalanche photodiode and digitized its signal with a dedicated 24-bit analog-to-digital converter chip.
The researchers tested the system’s ability to create eccentricity maps — polar angle maps of the human visual cortex. To map the eccentricity, they imaged a subject as he viewed changing patterns — three types of stimuli presented sequentially on a 50 percent gray background — on a video screen. The stimuli, a black-and-white grid with a flashing region that moved from the center of the screen to the periphery, lasted 10 s each and were separated by 30 s of plain 50 percent gray screen.
The polar angle study also used a black-and-white grid. “In the polar angle study, we used pie slice-shaped patterns and rotated the position of the slice around the center of the screen,” Culver explained. The group averaged 30 full-sweep cycles of 36 s for each of five subjects. The technique clearly showed that it could map the polar angle response within the visual cortex.
The researchers found that their method produced functional images that are in agreement with previous MRI or PET studies. They concluded that differences in the mapped locations were expected: “For example, it is known from anatomical and functional studies that cortical folding differs between subjects, and thus the localization and borders of functional responses also vary significantly between individuals.”
The researchers noted that they were able to reproduce activations in a single subject over multiple sessions. This reproducibility could prove useful in localizing the focus of seizures before brain surgery, for example. They also noted that, to prepare the system for widespread use, they must have diffuse optical tomography images coregistered with anatomical images, and that they would like to apply an MRI-derived head model instead of a simpler hemispheric one to create more accurate images for clinical applications.
The researchers are working on developing a larger optical array that can image the entire head instead of a small part. “In another project, we are developing more wearable imaging caps for use in imaging infants,” Culver added.
The basic instrument is suitable for both human research and clinical use. They are refining a number of features, including ease of use, imaging and data processing speeds, and the flexibility and wearability of the cap. In addition, they are beginning to work with clinicians to evaluate potential uses. Lastly, Culver added that they are interested in using the system to study brain development in children during the preschool and early elementary years.
MORE FROM PHOTONICS MEDIA