Deep Brain Fluorescence Imaging in the Second Near-Infrared Window

The body of terrestrial animals, including us, human beings, is usually nontransparent in the spectral region visible to the eye (400 – 750 nm, hence, the namesake of “visible spectrum”). This fact, as a result of million-year-scale evolution of terrestrial habituation has a significant consequence on today’s imaging techniques that scientists and clinicians employ to understand the anatomy, functions and pathology inside the body of animals and patients: fluorescence imaging employing photons in the visible spectrum is limited to a very superficial layer of the tissue in live animals despite its superior resolution in visualizing small features. Fortunately, in another spectral window not far from “visible” spectrum with doubled and even tripled wavelengths (1000 – 1700 nm, coined “second near-infrared window” or “NIR-II window” to differentiate from the conventional near-infrared window in the 750 – 1000 nm), most biological tissues become a lot more “transparent” (sadly, our eye, especially the photoreceptors in the retina, also becomes transparent in this window, making us unable to see anything in this window) owing to much reduced absorption, scattering and autofluorescence, affording deep-tissue fluorescence imaging with the right choice of fluorophores (Nat. Biomed. Eng. 2017).


Since the first demonstration of in vivo fluorescence imaging in the NIR-II window using solubilized and bio-functionalized carbon nanotubes in 2009 by the Dai lab at Stanford (Nat. Nanotechnol. 2009), in vivo NIR-II fluorescence imaging has been applied to visualize tumors (JACS 2012), peripheral vasculature (Nat. Med. 2012) and cerebral vasculature (Nat. Photonics 2014) deep inside the body in a noninvasive manner. Besides carbon nanotubes (Chem. Rev. 2015), a wide variety of other inorganic and organic materials, including quantum dots (Angew. Chem. 2012), conjugated polymers (Nat. Commun. 2014) and small molecules (Nat. Mater. 2016) have been added to the palette of biocompatible NIR-II contrast agents and reporters. The fluorescence emission wavelengths have also been tuned widely to 1700 nm and beyond (Angew. Chem. 2015), the longest photon wavelengths that have ever been used in fluorescence imaging.


The deep-tissue penetration of NIR-II light provides a noninvasive interface between the brain and an external neural recording / stimulation instrumentation, which employs free-space NIR-II photons alone as the medium for noncontact neural signal transduction to interrogate and modulate the activities at single-neuron level deep inside the brain. Knowledge gained and tools developed based thereon will pave the way for enabling paradigms of mind reading and mind control, boosting next-generation brain-machine interfaces and even direct brain-to-brain interfaces that bridge the gap between the minds.

Seeing through the brain in the NIR-II window

Deep tissue imaging of neovasculature in a limb ischemia model with NIR-II fluorescence

Long-Term Chronic Interface to the Brain and the Retina With Mesh Electronics (under advice of Prof. Charles M. Lieber, Harvard University)

The beautiful drawings of Ramon y Cajal more than a century ago reveal the intricate interconnection of neurons and glia cells into a 3D network structure. The retina bears a similar network structure as revealed in its name, meaning “net” in Latin. However, existing electrophysiology probes for understanding the brain and retina are usually designed with structures and mechanics unlike those of the brain and the retina, leading to an interface that only allows electrical recording of neuron signals shortly after acute implantation and often becomes unstable over long time periods. The Lieber Lab at Harvard pioneers in the mesh electronics paradigm to blur the distinction between the living biological system – the brain – and the nonliving electronic system – the probe (Curr. Opin. Neurobiol. 2018).


Mesh electronics is designed with similar structural and mechanical properties as the neural tissue, and can be delivered into virtually any part of the nervous system via syringe injection like pharmaceuticals owing to its ultraflexibility (Nat. Nanotechnol. 2015). With controlled injection to target specific brain regions (Nano Lett. 2015) and high-throughput input / output electrical connection method (Nano Lett. 2017), mesh electronics is found to integrate seamlessly with the endogenous neural network, affording chronically stable recording of brain activity at the single-neuron level (Nat. Methods 2016) and long-term intimate neural interface free of chronic immune response (PNAS 2017). Moreover, mesh electronics offers the first chronically stable in vivo electrophysiology of retina, which is considered as an approachable part of the brain, revealing functionally diverse retinal ganglion cells in awake mice.


The unique capability of mesh electronics for chronic interrogation and manipulation of single neuron activity opens up a myriad of exciting new directions, from fundamental neuroscience study of age-related cognitive decline to neurological application of syringe-deliverable “electroceuticals” for neurodegenerative diseases. The highly fenestrated mesh electronics offers a unique window to peek into the brain by eavesdropping on individual neurons and their encompassing neural circuit, thus allowing one to understand the neurobiological underpinnings of the mind and how the mind drives different behaviors.


“Mesh electronics interfaces with individual neurons chronically”

“Mesh electronics floating in saline like macromolecules and colloids”