ll

Perspective
Integrated Neurophotonics:
Toward Dense Volumetric Interrogation of
Brain Circuit Activity—at Depth and in Real Time
Laurent C. Moreaux,1,* Dimitri Yatsenko,2,15 Wesley D. Sacher,3,4,5 Jaebin Choi,6 Changhyuk Lee,6,7 Nicole J. Kubat,1
R. James Cotton,8,15 Edward S. Boyden,9,10,11,12 Michael Z. Lin,13 Lin Tian,14 Andreas S. Tolias,2,15,18 Joyce K.S. Poon,5,16
Kenneth L. Shepard,6 and Michael L. Roukes1,3,4,17,*
1Division of Physics, Mathematics and Astronomy, California Institute of Technology, Pasadena, CA 91125, USA
2Vathes LLC, Houston, TX 77030, USA
3Kavli Nanoscience Institute, California Institute of Technology, Pasadena, CA 91125, USA
4Division of Engineering and Applied Science, California Institute of Technology, Pasadena, CA 91125, USA
5Max Planck Institute for Microstructure Physics, Halle, Germany
6Departments of Electrical Engineering and Biomedical Engineering, Columbia University, New York, NY 10027, USA
7Center for BioMicrosystems, Brain Science Institute, Korea Institute of Science and Technology, Korea
8Shirley Ryan AbilityLab, Northwestern University, Chicago, IL 60611, USA
9Howard Hughes Medical Institute, Cambridge, MA, USA
10McGovern Institute, MIT, Cambridge, USA
11Koch Institute, MIT, Cambridge, USA
12Departments of Brain and Cognitive Sciences, Media Arts and Sciences, and Biological Engineering, MIT, Cambridge, USA
13Departments of Neurobiology and Bioengineering, Stanford University, Stanford, CA 94305, USA
14Department of Biochemistry and Molecular Medicine, School of Medicine, University of California, Davis, CA 95616, USA
15Center for Neuroscience and Artificial Intelligence and Department of Neuroscience, Baylor College of Medicine, Houston, TX 77030, USA
16Department of Electrical and Computer Engineering, University of Toronto, 10 King’s College Rd., Toronto, ON M5S 3G4, Canada
17Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, CA 91125, USA
18Department of Electrical and Computer Engineering, Rice University, Houston, TX 77005, USA

*Correspondence: moreauxl@caltech.edu (L.C.M.), roukes@caltech.edu (M.L.R.)

SUMMARY

We propose a new paradigm for dense functional imaging of brain activity to surmount the limitations of pre-
sent methodologies. We term this approach ‘‘integrated neurophotonics’’; it combines recent advances in
microchip-based integrated photonic and electronic circuitry with those from optogenetics. This approach
has the potential to enable lens-less functional imaging from within the brain itself to achieve dense, large-
scale stimulation and recording of brain activity with cellular resolution at arbitrary depths. We perform a
computational study of several prototype 3D architectures for implantable probe-array modules that are de-
signed to provide fast and dense single-cell resolution (e.g., within a 1-mm3 volume of mouse cortex
comprising 100,000 neurons). We describe progress toward realizing integrated neurophotonic imaging
modules, which can be produced en masse with current semiconductor foundry protocols for chip
manufacturing. Implantation of multiple modules can cover extended brain regions.

Massively Parallel Interrogation of Brain Activity Realizing instrumentation to monitor population activity within
the brain with single-neuron resolution is a profoundly difficult chal-
Within the central nervous system, the events in each unit
lenge; Figure 1 provides a sense of the scale involved. The slow
are not so important. We are more concerned with the in-
rate of technological development in neuroscience is elucidated
teractions of large numbers, and our problem is to find the
in Figure 2; it charts the evolution of our ability to simultaneously
way in which such interactions can take place.—Edward
resolve and track the activity of a multiplicity of neurons in vivo,
D. Adrian (1926)
over the six decades since the invention of whole-cell recording
These final lines from Lord Adrian’s Nobel lecture (Adrian, (Stevenson and Kording, 2011). Today’s state-of-the-art technol-
1926) illustrate the extraordinary prescience of this researcher ogy permits simultaneous, full bandwidth recording in vivo in
who first discovered neuronal spiking. He anticipated that under- awake rodents from multi-shank neural probe modules, each
standing brain computation is not likely to be achieved solely by with up to 1,024 channels (Rios et al., 2016; Shobe et al., 2015).
studies of individual neurons but instead by observing coordi- With implantation of multiple probes of these types, many thou-
nated interactions of neurons and their collective activity sands of neurons are now being simultaneously recorded (Stein-
patterns. metz et al., 2019). Although it is unequivocal that these advances

66 Neuron 108, October 14, 2020 ª 2020 Elsevier Inc.

Figure 1. Brain Complexity, ‘‘Brain Fields,’’ and Structural Length Scales Vis-à-Vis Cell-Body Location, Density, and Heterogeneity in the
Rodent Brain
Strong light scattering and absorption in brain tissue make it extremely difficult to achieve dense, volumetric functional imaging with cellular resolution.
(A) Biophysical scales for electrical, neurochemical, and optical domain recordings and relative sizes of brain structures.
(B) A 2-mm-thick optical section of an adult rat brain slice, stained with a fluorescent nuclear stain, wet mounted, and imaged by large-scale serial two-photon
microscopy. Beneath this image, we enumerate three ‘‘brain fields’’—that is, domains of neural activity: the electrical, neurochemical, and mechanical.
(C–E) Cellular nuclear density at multiple scales (C, 500 mm; D, 200 mm; E, 20 mm), from the macroscopic down to the level of individual cells. Image credits for (B)–
(E): L. Moreaux.

open exciting research frontiers, the number of observable neu- microscale photonic emitter and detector pixels (hereafter, E-
rons has continued to remain comparable to the total electrode and D-pixels) positioned on a 3D spatial lattice (Roukes, 2011;
count. This is consistent with the empirical observation that Roukes et al., 2016). These pixel arrays are integrated onto nar-
multi-site extracellular electrodes yield, on average, just one or row silicon shanks (needles), which leverage recent advances
two units per site, even with optimal spike-sorting algorithms (Mar- in silicon-nanoprobe-based fabrication (Rios et al., 2016; Shobe
blestone et al., 2013). At this rate of development, another 90 years et al., 2015; Steinmetz et al., 2018). Used with functional molec-
must elapse before the activity of an entire mouse brain, containing ular reporters (Andreoni et al., 2019; Chen et al., 2013; Lin and
roughly 75 million neurons, will become observable (Figure 2). Schnitzer, 2016) and optogenetic actuators (Boyden, 2011; Mie-
Clearly, we must significantly accelerate this rate of development. senböck, 2011), this novel instrumentation offers the prospect of
This was a central aim of our proposal (Alivisatos et al., 2012) approaching the interrogation of all neuronal activity from within
that eventually culminated in the launching of the US BRAIN a 1-mm3 volume (100,000 neurons in mouse cortex). The
Initiative (Bargmann and Newsome, 2014). Our initial vision, approach leverages recent breakthroughs in molecular reporters
which still remains true, is that advances in nanotechnology, mo- that can enable multimodal and multi-physical sensing
lecular reporters, and large-scale integration of semiconductor (Figure 1B), advances in optogenetic actuators that enable opti-
devices now make it feasible to precipitously upscale the rate cal control of neural activity, and the genetically encoded deliv-
of progress toward massively multiplexed interrogation of brain ery of molecular reporters and actuators that provide specificity
circuits (Alivisatos et al., 2012). of cell type. Further, the methodology is potentially scalable—
Here, we focus in more depth on these prospects. Our aim in multiple modules can be tiled to densely cover extended regions
this Perspective is not solely to identify ways to increase the total deep within the brain. We anticipate that this will ultimately
number of neurons that can be recorded from simultaneously. permit interrogation —that is, simultaneous recording and
Instead, we explore the possibility of achieving dense recording patterned stimulation of millions of neurons, at arbitrary positions
from within a targeted tissue volume to ultimately achieve com- and depths in the brain—to unveil dynamics of neural networks
plete interrogation of local brain circuit activity. We use the word with single-cell resolution and specificity of cell type. Like their
interrogation to denote recording and direct causal manipulation contemporary counterparts for highly multiplexed electrophysi-
of a brain circuit’s individual neurons by the application of ology (Ephys), ultranarrow photonic neural probes perturb brain
patterned, deterministic stimulation with single-neuron resolu- tissue minimally given their small cross sections and passivated
tion. To achieve this, we are pursuing a new approach, which surfaces. They impose negligible tissue displacement upon im-
we term integrated neurophotonics, that offers significant poten- plantation while dissipating low power during operation—com-
tial for accelerating progress toward Lord Adrian’s vision. This parable to today’s active, multi-site Ephys probes that also
technological path offers the prospect of dense functional imag- employ complementary metal-oxide-semiconductor (CMOS)
ing of neuronal activity in highly scattering neural tissue, technology. And, importantly, they offer near-term prospects
providing cellular-scale resolution at arbitrary depths in the brain. for wide deployment to the neuroscience research community,
Our approach is based on implanting an entire lens-less imaging as they are mass producible by well-validated semiconductor
system within the brain itself by distributing dense arrays of foundry (microchip-production factory) methods.

Neuron 108, October 14, 2020 67