CoChat
Sign Up
Cited by 335Optical Imaging and Spectroscopy TechniquesAdvanced Fluorescence Microscopy TechniquesNeural dynamics and brain function

The DOI matches its Crossref record; 84 of the 85 references it cites were found in a registry or an index.

Wide-field optical mapping of neural activity and brain haemodynamics: considerations and novel approaches

Open access
Ma, Ying, Shaik, Mohammed A., Kim, Sharon H., Kozberg, Mariel G., Thibodeaux, David N., Zhao, Hanzhi T., Yu, Hang, Hillman, Elizabeth M. C.·2016·Philosophical Transactions of the Royal Society B: Biological Sciences
DOI 10.1098/rstb.2015.0360PMC PMC5003860
Open this paper in CoChat — free account

Abstract

Although modern techniques such as two-photon microscopy can now provide cellular-level three-dimensional imaging of the intact living brain, the speed and fields of view of these techniques remain limited. Conversely, two-dimensional wide-field optical mapping (WFOM), a simpler technique that uses a camera to observe large areas of the exposed cortex under visible light, can detect changes in both neural activity and haemodynamics at very high speeds. Although WFOM may not provide single-neuron or capillary-level resolution, it is an attractive and accessible approach to imaging large areas of the brain in awake, behaving mammals at speeds fast enough to observe widespread neural firing events, as well as their dynamic coupling to haemodynamics. Although such wide-field optical imaging techniques have a long history, the advent of genetically encoded fluorophores that can report neural activity with high sensitivity, as well as modern technologies such as light emitting diodes and sensitive and high-speed digital cameras have driven renewed interest in WFOM. To facilitate the wider adoption and standardization of WFOM approaches for neuroscience and neurovascular coupling research, we provide here an overview of the basic principles of WFOM, considerations for implementation of wide-field fluorescence imaging of neural activity, spectroscopic analysis and interpretation of results. This article is part of the themed issue ‘Interpreting BOLD: a dialogue between cognitive and cellular neuroscience’.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1098/rstb.2015.0360

Cites

84 of 84 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1. Demonstration of WFOM imaging of neural activity and haemodynamics in awake, behaving mouse brain. (a) A custom-made acrylic head plate is surgically implanted onto a thinned skull cranial window. After recovery, the mouse's head is held by attaching the head plate to an aluminium head plate holder to restrict head motion during imaging. The animal is free to run or rest on a saucer wheel positioned below. (b) Functional maps of Thy1-GCaMP6f (selectively expressed in excitatory neurons of layers 2/3 and 5 [17]) ΔF/F after haemodynamic correction, Δ[HbT], Δ[HbO] and Δ[HbR] 1.7 s after stimulus onset. (c) Grey scale raw fluorescence image of the cortical surface with four selected regions of interest (ROIs) in the contralateral (1) whisker-barrel and (2) motor areas, as well as in the ipsilateral (3) whisker-barrel and (4) motor areas. (d) Time course of neural GCaMP neural activity and haemodynamic changes in HbT, HbO and HbR corresponding to these four regions. Grey shading shows stimulus time. Black dashed line shows time of frames shown above. Data were acquired using Andor Ixon camera with 512 × 512 frames and 30 ms exposure time per frame. Data shown are an average of 38 repeated stimulation trials in one mouse, with 5 s duration, 25 Hz tactile whisker stimulation.
Figure 1. Demonstration of WFOM imaging of neural activity and haemodynamics in awake, behaving mouse brain. (a) A custom-made acrylic head plate is surgically implanted onto a thinned skull cranial window. After recovery, the mouse's head is held by attaching the head plate to an aluminium head plate holder to restrict head motion during imaging. The animal is free to run or rest on a saucer wheel positioned below. (b) Functional maps of Thy1-GCaMP6f (selectively expressed in excitatory neurons of layers 2/3 and 5 [17]) ΔF/F after haemodynamic correction, Δ[HbT], Δ[HbO] and Δ[HbR] 1.7 s after stimulus onset. (c) Grey scale raw fluorescence image of the cortical surface with four selected regions of interest (ROIs) in the contralateral (1) whisker-barrel and (2) motor areas, as well as in the ipsilateral (3) whisker-barrel and (4) motor areas. (d) Time course of neural GCaMP neural activity and haemodynamic changes in HbT, HbO and HbR corresponding to these four regions. Grey shading shows stimulus time. Black dashed line shows time of frames shown above. Data were acquired using Andor Ixon camera with 512 × 512 frames and 30 ms exposure time per frame. Data shown are an average of 38 repeated stimulation trials in one mouse, with 5 s duration, 25 Hz tactile whisker stimulation.CC BY · PubMed Central
Figure 2. Imaging geometry for WFOM (a) and absorption spectra of oxy- and deoxyhaemoglobin (HbO and HbR) (b). Vertical lines show common wavelengths for reflectance WFOM measurements. Raw data, reproduced from [26] is tabulated in the electronic supplementary material, appendix A.
Figure 2. Imaging geometry for WFOM (a) and absorption spectra of oxy- and deoxyhaemoglobin (HbO and HbR) (b). Vertical lines show common wavelengths for reflectance WFOM measurements. Raw data, reproduced from [26] is tabulated in the electronic supplementary material, appendix A.CC BY · PubMed Central
Figure 3. Example of raw reflectance WFOM data in awake mouse. (a) Imaging system including blue, green and red LEDs illuminating the surface of the cortex—the resulting reflectance and fluorescence images (500 nm long-pass filter, LPF) are recorded by the camera. (b,c) Raw, baseline green and red reflectance images of the surface of the cortex. (e,f) Green and red reflectance changes (%ΔR/R) at t = 3.05 s (1.44 s post-stimulus onset) corresponding to the data shown in figure 1 (response to 5 s, 25 Hz tactile whisker stimulus in awake, behaving mouse, average of 38 repeated trials). (d) Temporal profile of ΔR/R (% reflectance change) in green and red channels during a stimulus from regions indicated in maps to the right. Grey shading shows stimulus time, while black dashed line is at t = 3.05 s. Note red changes (right axis) are shown on 10× scale compared to green (left axis).
Figure 3. Example of raw reflectance WFOM data in awake mouse. (a) Imaging system including blue, green and red LEDs illuminating the surface of the cortex—the resulting reflectance and fluorescence images (500 nm long-pass filter, LPF) are recorded by the camera. (b,c) Raw, baseline green and red reflectance images of the surface of the cortex. (e,f) Green and red reflectance changes (%ΔR/R) at t = 3.05 s (1.44 s post-stimulus onset) corresponding to the data shown in figure 1 (response to 5 s, 25 Hz tactile whisker stimulus in awake, behaving mouse, average of 38 repeated trials). (d) Temporal profile of ΔR/R (% reflectance change) in green and red channels during a stimulus from regions indicated in maps to the right. Grey shading shows stimulus time, while black dashed line is at t = 3.05 s. Note red changes (right axis) are shown on 10× scale compared to green (left axis).CC BY · PubMed Central
Figure 4. The effects of scattering on spatial sensitivity and pathlength. (a) Depiction of light-scattering paths in diffuse reflectance. Photon 1 scatters shallowly and emerges close to the point of illumination. Photon 2 scatters more deeply, becomes less probable but is still detected. Photon 3 scatters and is not emitted within the NA of the detector. (b) The vascular architecture of the mouse cortex, where large arteries and veins cover the cortical surface, with diving arterioles feeding deeper capillary beds, which in turn are drained by ascending venules. WFOM illuminates the cortical surface, visualizing surface vessels with sharp contrast. Light penetrating more deeply is attenuated, but can report more blurred signals from capillary contributions as depicted in (c). (d) Shows Monte Carlo-based simulations of 488, 530 and 630 nm light, showing (y-averaged) spatial sensitivity of a camera pixel to changes in absorption at different depths and lateral positions (log10 scale on colour bar and contours). Properties: 488 nm: μa = 0.33 mm−1, μs = 21 mm−1, g = 0.8; 530 nm: μa = 0.55 mm−1, μs = 21 mm−1, g = 0.82; 630 nm: μa = 0.024 mm−1, μs = 24 mm−1, g = 0.87, values from [29], and assuming 3% blood in tissue and 75% average oxygen saturation. (e) Plot extracted from (d) showing the overall depth sensitivity of measurements at a single pixel (over all lateral positions) for the three wavelengths (linear scale) and (f) the lateral sensitivity measurements at a single pixel (over all depths). These plots show that red light can detect changes in much deeper tissues than green light (although this difference depends most strongly on absorption differences than scatter). (g) Simulated wavelength-dependent pathlength based on Monte Carlo simulations (methods and tabulated values included in the electronic supplementary material, appendix B). Measured spectra of blue, green and red LEDs with band-pass filters are overlaid. Pathlength values (and haemoglobin molar extinction values) used are integrated under these spectral curves for haemodynamic conversion. LEDs used: Thorlabs M490L3 (GCaMP excitation blue), M530L3 (green) and M625L3 (red). Filters used: Semrock FF01-530/43-25 (530/43 in front of the green) and FF01-475/28-25 (475/28 in front of the blue).
Figure 4. The effects of scattering on spatial sensitivity and pathlength. (a) Depiction of light-scattering paths in diffuse reflectance. Photon 1 scatters shallowly and emerges close to the point of illumination. Photon 2 scatters more deeply, becomes less probable but is still detected. Photon 3 scatters and is not emitted within the NA of the detector. (b) The vascular architecture of the mouse cortex, where large arteries and veins cover the cortical surface, with diving arterioles feeding deeper capillary beds, which in turn are drained by ascending venules. WFOM illuminates the cortical surface, visualizing surface vessels with sharp contrast. Light penetrating more deeply is attenuated, but can report more blurred signals from capillary contributions as depicted in (c). (d) Shows Monte Carlo-based simulations of 488, 530 and 630 nm light, showing (y-averaged) spatial sensitivity of a camera pixel to changes in absorption at different depths and lateral positions (log10 scale on colour bar and contours). Properties: 488 nm: μa = 0.33 mm−1, μs = 21 mm−1, g = 0.8; 530 nm: μa = 0.55 mm−1, μs = 21 mm−1, g = 0.82; 630 nm: μa = 0.024 mm−1, μs = 24 mm−1, g = 0.87, values from [29], and assuming 3% blood in tissue and 75% average oxygen saturation. (e) Plot extracted from (d) showing the overall depth sensitivity of measurements at a single pixel (over all lateral positions) for the three wavelengths (linear scale) and (f) the lateral sensitivity measurements at a single pixel (over all depths). These plots show that red light can detect changes in much deeper tissues than green light (although this difference depends most strongly on absorption differences than scatter). (g) Simulated wavelength-dependent pathlength based on Monte Carlo simulations (methods and tabulated values included in the electronic supplementary material, appendix B). Measured spectra of blue, green and red LEDs with band-pass filters are overlaid. Pathlength values (and haemoglobin molar extinction values) used are integrated under these spectral curves for haemodynamic conversion. LEDs used: Thorlabs M490L3 (GCaMP excitation blue), M530L3 (green) and M625L3 (red). Filters used: Semrock FF01-530/43-25 (530/43 in front of the green) and FF01-475/28-25 (475/28 in front of the blue).CC BY · PubMed Central
Figure 5. Demonstration of spatio-temporal unmixing of WFOM haemodynamic signals. Data were acquired on an alpha-chloralose anaesthetized rat (from data shown in [11]) during 4 s electrical forepaw stimulation imaged using 470 and 530 nm reflectance measurements, converted to haemoglobin concentrations using equation (2.7). Unmixing method (as described in [31] and in the text) uses basis functions shown in (c), sampled from regions in the field of view (indicated by coloured dots in (a). Non-negative unmixing yields maps showing the ‘concentration’ of each basis set in the haemoglobin response dataset for each compartment (shown grey scale in (d), and as three-component and five-component colour merges in (b)). This analysis shows how WFOM can delineate clearly distinct time courses of each vascular compartment despite being a superficially weighted two-dimensional representation of the haemodynamic response.
Figure 5. Demonstration of spatio-temporal unmixing of WFOM haemodynamic signals. Data were acquired on an alpha-chloralose anaesthetized rat (from data shown in [11]) during 4 s electrical forepaw stimulation imaged using 470 and 530 nm reflectance measurements, converted to haemoglobin concentrations using equation (2.7). Unmixing method (as described in [31] and in the text) uses basis functions shown in (c), sampled from regions in the field of view (indicated by coloured dots in (a). Non-negative unmixing yields maps showing the ‘concentration’ of each basis set in the haemoglobin response dataset for each compartment (shown grey scale in (d), and as three-component and five-component colour merges in (b)). This analysis shows how WFOM can delineate clearly distinct time courses of each vascular compartment despite being a superficially weighted two-dimensional representation of the haemodynamic response.CC BY · PubMed Central
Figure 6. Spatial dependence of fluorescence WFOM sensitivity and haemodynamic cross-talk. (a–c) Depict different patterns of fluorescence expression that could be generated in the brain (indicated by green shading). Brighter regions are intended to denote a region of increased intracellular calcium. Panels below show estimated images of fluorescence that could be detected at the cortical surface in each case. (d) The absorption spectra of HbO and HbR overlaid with the excitation and emission bands of GCaMP, as well as bands indicating LED wavelengths used for WFOM results shown herein. (e) Schematic of light paths in WFOM travelling to and from a GCaMP fluorescence interaction.
Figure 6. Spatial dependence of fluorescence WFOM sensitivity and haemodynamic cross-talk. (a–c) Depict different patterns of fluorescence expression that could be generated in the brain (indicated by green shading). Brighter regions are intended to denote a region of increased intracellular calcium. Panels below show estimated images of fluorescence that could be detected at the cortical surface in each case. (d) The absorption spectra of HbO and HbR overlaid with the excitation and emission bands of GCaMP, as well as bands indicating LED wavelengths used for WFOM results shown herein. (e) Schematic of light paths in WFOM travelling to and from a GCaMP fluorescence interaction.CC BY · PubMed Central
Figure 7. Haemodynamic correction of WFOM GCaMP data and validation with electrophysiology. (a) Data shown in figure 1 before and after haemodynamic correction using the Ex–Em method. Maps are for 1.7 s after onset of a 5 s tactile whisker stimulus in an awake, behaving mouse (average of 38 trials). Corrected image shows only minor vessel artefacts, whereas original response is dominated by negative contrast corresponding to the haemodynamic response. (b) Time courses show the GCaMP signal before and after correction, in comparison with the D[HbT] haemodynamic response. (c) Raw uncorrected and single-wavelength ratiometrically corrected and GCaMP signals acquired in a urethane-anaesthetized Thy1-GCaMP3 mouse undergoing an 8 s electrical hindpaw stimulation. (d) An electrode inserted into the responding region recorded multi-unit activity (MUA). The GCaMP fluorescence signal after haemodynamic correction can be closely replicated by convolving MUA with a gamma function (e) to model the combined dynamics of intracellular calcium and GCaMP fluorescence changes.
Figure 7. Haemodynamic correction of WFOM GCaMP data and validation with electrophysiology. (a) Data shown in figure 1 before and after haemodynamic correction using the Ex–Em method. Maps are for 1.7 s after onset of a 5 s tactile whisker stimulus in an awake, behaving mouse (average of 38 trials). Corrected image shows only minor vessel artefacts, whereas original response is dominated by negative contrast corresponding to the haemodynamic response. (b) Time courses show the GCaMP signal before and after correction, in comparison with the D[HbT] haemodynamic response. (c) Raw uncorrected and single-wavelength ratiometrically corrected and GCaMP signals acquired in a urethane-anaesthetized Thy1-GCaMP3 mouse undergoing an 8 s electrical hindpaw stimulation. (d) An electrode inserted into the responding region recorded multi-unit activity (MUA). The GCaMP fluorescence signal after haemodynamic correction can be closely replicated by convolving MUA with a gamma function (e) to model the combined dynamics of intracellular calcium and GCaMP fluorescence changes.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 85 references

Cited by

Sorted by citations

See all 335

Related

Metadata

Authors
Ma, Y., Shaik, M. A., Kim, S. H., Kozberg, M. G., Thibodeaux, D. N., Zhao, H. T., Yu, H., & Hillman, E. M. C.
Affiliations
Mortimer B. Zuckerman Mind Brain Behavior Institute (US)
Corresponding author
Hillman, Elizabeth M. C.
ORCID
Venue
Philosophical Transactions of the Royal Society B: Biological Sciences
Publisher
The Royal Society
Published
2016-10-05
Volume
371
Issue
1705
Pages
20150360
ISSN
  • 0962-8436 (ISSN-L)
  • 1471-2970
DOI
10.1098/rstb.2015.0360
PMC
PMC5003860
OpenAlex
W2513047215
Cited by
335
References
85
Citations by year
2026: 24 · 2025: 54 · 2024: 52 · 2023: 47 · 2022: 40 · 2021: 31 · 2020: 24 · 2019: 21 · 2018: 24 · 2017: 15 · 2016: 3
FWCI
3.25
Type
journal article
Language
en
Update notices
correction (2017-02-19)
Open access
Licence
Locations
  • Philosophical Transactions of the Royal Society B Biological Sciences (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • National Science Foundation (0954796)
  • U.S. Department of Defense (MURI W911NF-12-1-0594)
  • National Institute of Neurological Disorders and Stroke (1R01NS063226, 5U01NS094296, 1R21NS053684, 1R01NS076628)
  • National Eye Institute (1R01EY019500)
Topics
  • Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Optical Imaging and Spectroscopy Techniques
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Biophysics › Advanced Fluorescence Microscopy Techniques
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
Keywords
neuronal activity imaging, neurovascular coupling, genetically encoded fluorescent indicators, wide-field fluorescence imaging, spectroscopic analysis, cortical imaging, BOLD signal
MeSH terms
Animals, Brain, Brain Mapping, Hemodynamics, Humans, Neurons, Mice, Rats, Optical Imaging

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review