Abstract
The primate cerebral cortex is organized into specialized areas representing different modalities and functions along a continuous surface. The functional maps across the cortex, however, are often investigated a single modality at a time (e.g., audition or vision). To advance our understanding of the complex landscape of primate cortical functions, here we develop a polarization-gated wide-field optical imaging method for measuring cortical functions through the un-thinned intact skull in awake marmoset monkeys (Callithrix jacchus), a primate species featuring a smooth cortex. Using this method, adjacent auditory, visual, and somatosensory cortices are noninvasively parcellated in individual subjects with detailed tonotopy, retinotopy, and somatotopy. An additional pure-tone-responsive tonotopic gradient is discovered in auditory cortex and a face-patch sensitive to motion in the lower-center visual field is localized near an auditory region representing frequencies of conspecific vocalizations. This through-skull landscape-mapping approach provides new opportunities for understanding how the primate cortex is organized and coordinated to enable real-world behaviors.
The Challenge of Primate Brain Mapping
Understanding how the primate brain organizes different senses—like sight, sound, and touch—into a continuous functional landscape is a major challenge in neuroscience. Traditionally, researchers face a difficult trade-off: high-resolution but invasive methods like intrinsic signal optical imaging (ISOI) usually require a craniotomy, which limits the field of view and carries surgical risks. Conversely, non-invasive methods like fMRI lack the fine-grained mesoscopic resolution needed to see detailed functional maps, such as the specific layout of frequencies in the auditory cortex.
While through-skull imaging is common in mice, primate skulls are thick enough to scatter light and create a "glare" that obscures underlying brain activity. This paper introduces XINTRINSIC (cross-linear polarization enhanced intrinsic imaging), a way to bypass that glare without thinning the bone.
XINTRINSIC: Seeing Through the Skull
The XINTRINSIC method utilizes two primary technical innovations to achieve high-resolution imaging through intact bone:
- Polarization Gating: The setup illuminates the skull with linearly polarized light but records only the light in the orthogonal (cross) polarization. Light reflecting off the bone surface maintains its original polarization and is filtered out, while light that travels deep into the cortex and undergoes multiple scattering events becomes depolarized, allowing the camera to "see" past the bone.
- Wavelength Optimization: Researchers found that green light (530 nm) provided the best signal-to-noise ratio. This wavelength measures changes in total blood volume, making the signal robust and comparable to contrast-enhanced fMRI.
Mapping the Marmoset Cortical Landscape
The study utilized 11 common marmosets (Callithrix jacchus), a primate species with a smooth (lissencephalic) cortex ideal for wide-field optical imaging. By registering functional data to established structural brain templates, the researchers were able to map the complex geography of the primate brain with high fidelity.
Key Insights and Results
- The "Social" Network: The researchers localized a "face patch" specialized for processing faces and motion in the lower-center visual field. This patch sits adjacent to auditory regions representing frequencies used in marmoset vocalizations, suggesting a specialized cortical hub for social communication.
- New Auditory Geography: A previously unknown tonotopic gradient was discovered in the auditory cortex extending into the "parabelt" regions, suggesting a more complex organization than current models indicate.
- Simultaneous Parcellation: Somatosensory, auditory, and visual cortices were successfully mapped in single 6.6-minute imaging sessions. The through-skull maps showed a high correlation ($r = 0.88$) with traditional invasive through-window maps.
Limitations and Future Work
While a major advancement, XINTRINSIC is currently "minimally invasive" rather than completely non-invasive, as the scalp must be removed to allow light access. Additionally, like all optical methods, it is limited to the cortical surface and cannot image deep structures like the hippocampus. Future iterations may use longer wavelengths (NIR) to penetrate even thicker skulls or be miniaturized for use in freely moving animals. This approach—costing less than $10,000 to build—could significantly accelerate primate neurophysiology research.
Summary updated