Statistically unbiased prediction enables accurate denoising of voltage imaging data
Creators
- 1. School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea
- 2. Department of Materials Science and Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea
- 3. Department of Chemistry and Chemical Biology, Harvard University, Cambridge, United States
- 4. Department of Biology, Chungnam National University, Daejeon, South Korea
- 5. School of Biological Sciences, Seoul National University, Seoul, Republic of Korea.
- 6. Allen Institute for Neural Dynamics, Seattle, WA, USA.
Description
Here we report SUPPORT (Statistically Unbiased Prediction utilizing sPatiOtempoRal information in imaging daTa), a self-supervised learning method for removing Poisson-Gaussian noise in voltage imaging data. SUPPORT is based on the insight that a pixel value in voltage imaging data is highly dependent on its spatially neighboring pixels in the same time frame, even when its temporally adjacent frames do not provide useful information for statistical prediction. Such spatiotemporal dependency is captured and utilized to accurately denoise voltage imaging data in which the existence of the action potential in a time frame cannot be inferred by the information in other frames. Through simulation and experiments, we show that SUPPORT enables precise denoising of voltage imaging data while preserving the underlying dynamics in the scene.
We also show that SUPPORT can be used for denoising time-lapse fluorescence microscopy images of Caenorhabditis elegans (C. elegans), in which the imaging speed is not faster than the locomotion of the worm, as well as static volumetric images of Penicillium and mouse embryos. SUPPORT is exceptionally compelling for denoising voltage imaging and time-lapse imaging data, and is even effective for denoising calcium imaging data.
For more details, please see the accompanying research publication "Statistically unbiased prediction enables accurate denoising of voltage imaging data".
Datasets for volumetric structural imaging of penicillium and calcium imaging of zebrafish.
Volumetric structural imaging of penicillium
Penicillium/penicillium_low_snr.tif
--> Low SNR image
Penicillium/penicillium_high_snr.tif
--> High SNR image
Calcium imaging of zebrafish
Zebrafish/zebrafish_multiple_brain_regions.tif
--> Multiple brain regions
Zebrafish/zebrafish_Cerebellar_plate.tif
--> Cerebellar plate
Zebrafish/zebrafish_Dorsal_telencephalon.tif
--> Dorsal telencephalon
Zebrafish/zebrafish_Medulla_oblongata.tif
--> Medulla oblongata
Zebrafish/zebrafish_Olfactory_bulb.tif
--> Olfactory bulb
Zebrafish/zebrafish_Optic_tectum.tif
--> Optic tectum
Zebrafish/zebrafish_Habenula.tif
--> Habenula
Files
data_explanations.txt
Additional details
Related works
- Is required by
- Preprint: 10.1101/2022.11.17.516709 (DOI)