Osaka University Raman method offers faster tissue imaging

15 Feb 2023

Multiline illumination detects tissue areas in parallel, could expand biomedical applications.

A project at Osaka University has developed a Raman microscopy platform intended specifically to increase the imaging speed as a route to new clinical uses for the technique.

Described in Biomedical Optics Express, the technique involves multiline illumination confocal Raman microscopy, where the detection of separate regions of the sample is performed in parallel.

"Use of Raman microscopy in life sciences is limited and life scientists predominantly use confocal fluorescence microscopy, although the fluorescence measurement requires pretreatment of a sample with additional chemicals," noted the Osaka team in its paper.

"In spontaneous Raman microscopy, which provides the inherent molecular fingerprint of the sample without using a dye, the weak point is a far slower measurement speed. Although imaging speed has improved dramatically in the latest couple of decades, a high-resolution imaging mode can still typically take an hour or longer."

The project's solution was to irradiate a sample with multiple line illuminations, and detect the resulting emission with a spectrometer equipped with a periodic array of confocal slits. A comb-like Raman hyperspectral image is then formed on a spectrometer's two-dimensional detector, and a hyperspectral Raman image built up by scanning the sample with multiline illumination arrays.

This expands on previous research at Osaka into line-illumination Raman microscopy, which showed the technique was faster than conventional confocal Raman microscopy and enabled dynamic imaging of living cells, but was at that point still too slow for the large-area imaging required for medical diagnosis and tissue analysis.

"With our new technique, the resolution and imaging speed can be adjusted, depending on the application," said Katsumasa Fujita from Osaka University. "In the future, even faster imaging speed might be possible as cameras continue to be developed with more pixels."

Enabling medical diagnoses not previously possible

The team's device irradiated a sample with 21 simultaneous illumination lines, for proof-of-concept trails which confirmed a corresponding 21-fold improvement in image acquisition time over a single-line method. When applied to samples of mouse brain tissue, 66 frames of 903 by 800 pixels were acquired in 11.4 minutes, equating to a total of 1,108,800 spectra sampled in that time.

Live-cell imaging was also carried out using HeLa cells, although for this application the number of line-shaped illuminating beams was dropped from 21 to 11, primarily to lower the background contributions from water in the buffer solutions employed.

Key aspects of the new platform include the cylindrical lens array used to generate multiple line-shaped beams from a single laser source, and a high-sensitivity low-noise CCD camera with a large number of pixels to allow a large number of Raman spectra to be distributed on the sensor chip and detected simultaneously. The project's spectrophotometer was custom made to create a 2D distribution of spectra on the camera without significant distortion.

The next steps will involve increasing the imaging speed further factor while also reducing the cost of camera, laser, and spectrophotometer in order to make commercialization more practical. But the initial results already indicate that a multiline technique could address the inherent issue of sensitivity in Raman applications, offering a way to acquire more signals in a given exposure time.

"We hope that high-throughput Raman imaging will eventually make it possible to perform medical diagnoses more efficiently and accurately while possibly enabling diagnoses that weren’t possible before," commented Katsumasa Fujita. "Label-free molecular analysis with Raman imaging would also be useful for efficiently detecting drug response of cells, aiding in drug development."