Thread by @adam_k_glaser

How do you image the large and the small at the same time? We developed new 🔬 technology to image centimeter-scale specimens - including whole mouse brains 🧠 - with diffraction-limited resolution and without sectioning. #mesoscale #imaging

biorxiv.org/content/10.110…

🧵 (1/n)

ExA-SPIM is a technology based on innovations in (1) tissue processing and (2) large-scale microscopy 🔬. Our team developed new methods for expansion for microscopy (ExM) of large tissue volumes. #teamscience #expansionmicroscopy

protocols.io/view/whole-mou…

🧵 (2/n)

By combining ExM with a new SPIM system based on technologies from electronics metrology, ExA-SPIM pushes past the volumetric imaging barrier 🚧 that constrains previous imaging approaches. (inspired by recent plot from @Daetwyler_St, @RetoPaul) #lightsheet #microscopy

🧵 (3/n)

The lens 🔎 at the heart of ExA-SPIM provides NA = 0.305 (1 µm resolution) over an incredible 16.8 mm FOV. This is paired with a 151 megapixel CMOS camera 📷, far exceeding the capabilities of traditional technologies. They are big! #bigscience

vision.vieworks.com/en/lens/veo_jm

🧵 (4/n)

We modified the optical system to achieve an aberration free working distance of 35 mm (!) into liquid 💧 mounting media. #deepimaging #highresolution

🧵 (5/n)

ExA-SPIM can image an incredible 200 x 52 x 35 mm3 volume, with 1 µm near-isotropic resolution over a 10.6 x 8.0 mm field of view. - at up to a 1 gigavoxels/sec 💨!
#largevolume

🧵 (6/n)

ExA-SPIM can therefore image an entire 3X expanded 🐁 mouse brain - with 300 nm effective resolution - in only 15 tiles - and in less than one day. This produces ~100 TB of raw data!
#highspeed

🧵 (7/n)

With no sectioning and few tiles, these volumes are easily stitched and merged for downstream analysis, including tracing ✏️ and reconstruction of individual neurons across the mouse brain. #neuralcircuits #hortacloud

hortacloud.janelia.org

🧵 (8/n)

ExA-SPIM is also a charm for larger brains 🧠. We imaged a 1x1x1.5 cm block of macaque brain. ExA-SPIM reveals brightly ⚡ labeled cortico-spinal tract neurons, their dendritic arbors, and incredibly dense dendritic spines as well as descending axons. #connectomics

🧵 (9/n)

ExA-SPIM required new 🐍 acquisition software for controlling the microscope. Data is streamed using ACQUIRE, a new multi-camera 📸 streaming software, led by @ChanZuckerberg in collaboration with our team.
#acquisitionsoftware #napari

github.com/acquire-project

🧵 (10/n)

Using a combination of high-speed networking, fast on-premises storage, and real-time compression, our current imaging pipeline enables an imaging throughput to cloud ☁️ storage of >100 TB per day, using commodity hardware.
#cloudcomputing

🧵 (11/n)

This collaboration was only possible at @AllenInstitute for Neural Dynamics, and represents the hard work of many team members! Including @vcj029, @svoboda314, @dyfbrain, @rclayreid, @sejdevries, @Poofjunior, @nclack, @kjcao, @KalmbachsCortex, and many others.
#teamscience