US researchers say they have developed a simple way to capture high-quality 3D images of live cells and organisms with a standard microscope.
The new technique simultaneously acquires images at different depths at high speeds and with high contrast, they write in a paper in the journal Optica, and can be applied to a variety of microscopy methods.
Standard camera-based microscopy systems capture images at a single focal plane. Most attempts to acquire images with different focal depths simultaneously have used multiple cameras or a special diffractive optical element to split images with a single camera.
Jerome Mertz and colleagues at Boston University tried a new approach, using a z-splitter prism to divide detected light to produce several images in a single camera frame.
“We used a z-splitter prism that can be assembled entirely from off-the-shelf components and is easily applied to a variety of imaging modalities such as fluorescence, phase-contrast or darkfield imaging,” says co-author Sheng Xiao.
The prism divides detected light to produce several images in a single camera frame. Each image is focused at a different depth in the sample.
Using a high-speed camera with a large sensor area and high pixel count, the researchers were able to distribute multiple high-resolution images on the same sensor without any overlap.
The multifocal images make it possible to estimate the out-of-focus background from the sample much more accurately than can be done with a single image, the researchers say. They used this information to develop an improved 3D deblurring algorithm that eliminates the out-of-focus background light that is often a problem when using widefield microscopy.
“This improves both the image contrast and signal-to-noise ratio, making it particularly beneficial in fluorescence imaging applications involving thick samples,” says Xiao.
To test the algorithm’s capability, the team imaged various thick samples, including the brain of a living mouse and observed what they say were significant contrast and signal-to-noise ratio improvements compared to both raw multi-focus images and traditional 3D deblurring algorithms.