Network Imaging

The FEMTO3D Atlas microscope with its unique technical solutions enables the precise spatial and real-time complexity of neuronal coding to be resolved by scanning large number of cells distributed in a 3D volume. 

• Imaging of thousands of cells simultaneously 
• Maximal scanning speed: 100 kHz 
• High singal-to-noise ratio 
• Online motion correction even in the axial dimension in behaving animal models 
• Quasi-simultaneous 3D photostimulation with 3D imaging 
• All available in the FEMTO3D Atlas two-photon imaging system 

The imaging of cell populations in 3D is the fastest with FEMTO3D Atlas

---

3D random-access point scanning

Several thousand cells can be measured near-simultaneously (up to 100 kHz per point) by 3D random-access point scanning by restricting the imaging to sub-regions of the 3D volume, which results ultra-high scanning speed and signal-to-noise ratio.The figure (A-D) shows the changes in Ca2+ concentration of 2000 cells from the visual cortex of a GCaMP-expressing mouse as a function of time. See more:
Wertz et al., Science, 2015; Katona et al., Nat Meth, 2012. FEMTO3D Atlas

Imaging of multiple somata distributed in 3D in behaving animals

---

3D chessboard scanning

Chessboard scanning is a planar extension of random-access point scanning, where random-access points are extended to small squares by drifting the laser beam. These squares can be located anywhere in the imaging volume and include the somata with the surrounding area. Up to 300 somata can be measured simultaneously. The chessboard arrangement of the squares helps to visualize somata, analyze their activity, and correct for motion artifacts, and completely eliminating them with the help of the 3D Real-Time Motion Correction. The video shows in vivo neuronal activity, Ca2+ transients from 100 neuronal somata from the mouse V1 region, labeled with GCaMP6. See also Szalay et al., Neuron, 2016.

Imaging of sparsely distributed interneurons in a large neuronal network during behavior in real-time in 3D

---

3D chessboard scanning in a large neuronal network

Sparsely distributed brain cells (such as interneurons) can be followed in real-time in a 3D volume of using chessboard scanning. In the study depicted in the figure below, the calcium activity from nearly 250 interneurons was measured simultaneously, across every CA1 hippocampal layer, in an 800 × 800 × 500 µm volume with the chessboard scanning mode, while the mouse was performing behavioral tasks on a treadmill. Using these recordings, the biochemical composition of the imaged cells was also determined from data acquired from post hoc immunohistochemistry. Such a novelty – identifying distinct interneuron types and their activity dynamics simultaneously in a large neural network – could not have been achieved with other traditional recording techniques. See more: Geiller et al., Neuron, 2020.  Watch the interview with one of the authors.

Imaging of multiple somata during large motions

---

3D multi-cube scanning 

Multi-cube scanning is a spatially extended mode of chessboard scanning, where a Z dimension is added to the chessboard squares to cover the whole extent of the somata, therefore preserving all fluorescent information from the somata even during large amplitude movements Learn more: Bharioke et al. Neuron (2022)

Real-time motion-correction with the Femtonics 3D-RTMC

---

Our solution takes advantage of acousto-optics technology and elevates it to a whole new level. The flexible drift scanning method creates the opportunity to perform correction calculations with a predefined and adjustable frequency. The algorithm is based on correction calculations derived from selecting stable objects within the field of view (FOV), which serve as reference points during scanning. With our method providing an effortless means to scan in any direction in 3D, reference measurements are equally efficient in the Z direction as they are in the XY direction, which is a unique feature of the Femtonics 3D Real-Time Motion Correction.

Fastest frame scanning with rotational freedom

---

High-speed arbitrary frame scanning

Cells assembled into a neural network can be acquired with a frame rate of 40 Hz in a 500×500 µm field-of-view (FOV) by the High-speed arbitrary frame scanning mode of the Atlas. Beyond the high speed, another important and unique advantage of this mode is that the selected plane can be rotated into any in space to match your layer of interest.

The video below demonstrates the rotation of the selected plane in High-speed arbitrary frame scan mode. The left panel shows the relative position of the objective’s focal plane with blue and the actual two-photon focal plane with pink. The movement of the pink plane exemplifies the possibilities Atlas offers you to orient the plane in a three dimensional tissue. The right panel shows the neurons of a network in the actual scanned plane during rotation in live mode. 

FEMTOSmart offers numerous imaging opportunities and with its vast modularity researchers can study neuronal networks from varied perspectives and save time in the experimental work. 

Imaging of somata as ROIs in 2D

---

FEMTOSmart Galvo with flexible ROI scanning possibilities

The FEMTOSmart Galvo, with its flexible scanning patterns such as 2D random-access point, 2D multiple-line, and folded-frame scanning, supports the manual selection of individual cells in a neural network in a 2D plane. By skipping the measurement of the entire field, it is possible to reach a fast scanning speed and maintain a high signal-to-noise ratio. Using the TravellingSalesman additional software module, it is possible to determine the shortest pathway visiting defined points arbitrarily dispersed in the field of view. The short round-trip time results in a higher measurement repetition speed, up to 100 Hz for about 30 cells.

---

Multiple-line scanning

Fast scanning of the entire FOV or an entire volume

---

FEMTOSmart Resonant equipped with Piezo objective positioner kit

Fast-frame scanning by FEMTOSmart Resonant combined with fast Z-focusing performed by a Piezo objective positioner kit is a well-known approach for studying any three-dimensional neural network. In this case, the entire field of view is imaged continuously by the fast scanner, while the objective positioner moves between planes. More than one (cortical) layer might be recorded simultaneously this way, or the frames may be assembled into volumes resulting in a four-dimensional set of data.