[
International Worm Meeting,
2021]
Unwanted background fluorescence in microscopy can occur when light emitted by fluorescent structures is scattered by nearby tissues. In our in vivo imaging of green fluorescent protein (GFP)-tagged neurons in the roundworm C. elegans, scattered light produces a haze surrounding the cell body that can obscure the imaging of target structures, such as an axon or dendrite. The thin fibers appear dimmer than the much larger cell body and cannot be clearly visualized due to the low contrast between it and the bright background. Here, we describe a method to model and remove the cell body haze utilizing an inverse square intensity distribution. Such distributions are common in nature and can describe the intensity of light that emanates away from a point or sphere, such as a fluorescent cell body. Assuming that scattering is proportional to intensity, it follows that an inverse square distribution also describes the intensity of the scattered light. Utilizing this model, we have developed a post processing procedure to subtract background from an image. Removal of the haze surrounding the bright cell body enhances contrast of the dim axon, particularly in the region close to the cell body. Preliminary algorithms demonstrate a signal to background ratio improvement of >5x. This improvement in image quality allows us to more clearly visualize the axon. We intend to broadly disseminate this technique via an ImageJ extension and in MATLAB. We may further release our fitting technique on other imaging platforms. We are also investigating methods to accelerate the computations for real-time image improvement.
[
International Worm Meeting,
2021]
Widefield microscopes are commonly used in many biological laboratories. However, their inability to reject scattered or out-of-focus light often produces images with obscured thin or dim structures when there are some bright structures nearby. Scanning microscopes like confocal microscopes and multiphoton microscopes can solve this problem, but they are comparatively slower and much more expensive. We developed an inexpensive way to empower confocal imaging capacity on a widefield microscope, by inserting a spatial light modulator (SLM) into the field stop of the widefield microscope and customizing the illumination pattern and acquisition methods. We assessed the performance of this SLM-inserted setup by comparing images taken at our widefield microscope, our widefield microscope with the SLM-inserted setup, and a commercial confocal microscope. While a widefield microscope showed no sectioning capability, our SLM-inserted setup showed 0.85 ± 0.04 mum and the commercial confocal showed 0.68 ± 0.04 mum optical sectioning capability. Additionally comparing images of the FLP neuron and the tightly bundled amphid neurons in C. elegans taken by the widefield, SLM-inserted setup, and confocal microscopes, we confirmed that the SLM-inserted setup greatly reduces haze from the bright cell body, allowing visualization of dim axons and dendrites nearby. Our SLM-inserted setup identified 96% of the dim neuronal fibers seen in confocal images while the widefield microscope only identified 50% of the same fibers. Our SLM-inserted setup represents a very simple (2-component) and inexpensive (<$600) approach to enable confocal capacity on a widefield microscope. This SLM-inserted setup can be broadly employed by labs that are using widefield microscopes, with minimum expense and modification.