Harvard SEAS metasurface improves tomographic image quality

23 Mar 2022

New method acts as add-on to OCT and other platforms to counteract effects of diffraction.

A project from the Havard group of metamaterials pioneer Federico Capasso offers a way to improve the high-resolution performance of tomography-based imaging modalities.

OCT, as one common example, has made great strides in axial resolution, but still faces challenges in lateral resolution and depth of focus, especially at larger imaging depths. Focusing light onto a single-depth point inhibits a system's depth imaging capability, while distributing light along the axial direction to counter this introduces diffraction problems and affects lateral resolution.

The Capasso team has developed a concept termed bijective illumination collection imaging (BICI), designed to accomplish high-resolution imaging in three dimensions in a relatively large depth range, effectively evading the restrictions imposed by diffraction.

As reported in Nature Photonics, the team has initially designed a solution for OCT imaging as proof-of-concept, but the same principle could also be employed as an add-on for other imaging modalities.

In BICI, the illumination and collection paths are separated using two metasurfaces consisting of arrays of nanoscale subwavelength-spaced optical elements. Using a set of surface structures specifically designed for the task can produce a one-to-one correspondence - or a "bijective" relation - between the focal points of the illumination and collection paths, effectively eliminating out-of-focus signals.

Existing high-resolution systems typically reduce the effect of out-of-focus signals by using tightly focused light to comparatively increase the signal from the focal point, noted the project team, but BICI rejects out-of-focus signals entirely because of the direct correlation created between the two optical paths, without compromising the depth range.

"BICI can extend the range of high-resolution imaging by over 12-fold compared to the state-of-the-art imaging techniques," said Majid Pahlevani of project partners Queen's University. "Unlike conventional imaging techniques, in BICI the light which illuminates the target and the light collected from the target are distributed along the depth by the nanostructures, making it possible to preserve high resolution imaging along a large depth into the tissue."

Cancer cell imaging for real-time diagnosis

With BICI incorporated into a Fourier-domain near-IR OCT system, images of tissue structures in the pulmonary airways of swine were taken ex vivo. Results showed "a lateral resolution of 3.2 microns that is maintained nearly intact over an imaging depth of 1.25 millimeters, giving an imaging depth of focus that is around 12-fold larger compared with that obtained using an ideal Gaussian beam with the same lateral resolution," according to the Nature paper.

A key advantage for BICI in real-world use should be that it does not impose any fresh computational challenges; a particular factor in live animal imaging according to Majid Pahlevani.

"Computationally intensive techniques result in slow imaging, which is not suitable for in vivo imaging," commented Pahlevani. "Organs in live patients are not stationary and move, which give rise to artifacts in imaging. Therefore, in vivo imaging requires fast techniques."

Having proved its worth in OCT, the BICI enhancement should also be applicable to other modalities, potentially assisting the imaging of cancer cells and related intercellular mechanisms for real-time cancer diagnosis.

"Although BICI is applied to OCT in this work, the underlying concept is general and might be adapted across various imaging modalities such as confocal microscopy and two-photon microscopy," said the project team.