World's Smallest LED Transforms Mobile Phone Cameras into High-resolution Microscopes - How?

The world's smallest LED developed by researchers in Singapore has paved the way for converting existing mobile phone cameras into high-resolution microscopes.

Smaller than the Wavelength of Light

A new type of LED, which is smaller than the wavelength of light, has been utilized to construct the world's tiniest holographic microscope. The silicon chip and software in conventional cameras have been modified to create microscopes.

Moreover, the scientists have devised a pioneering neural networking algorithm that rebuilds objects scanned by the holographic microscope. This innovation makes it feasible to inspect microscopic items like cells and bacteria without the need for cumbersome conventional microscopes or supplementary optics.

The researchers made a new tiny LED that is very bright and has the smallest light emission area ever seen for silicon LEDs. They used this LED to build a tiny microscope that does not need a lens or pinhole, called a lensless holographic microscope.

The conventional method of computational reconstruction relies heavily on a precise understanding of the experiment setup for accurate results, which is susceptible to optical aberrations, noise, and the twin-image issue.

To overcome these issues, the researchers created a deep neural network architecture that enhances image reconstruction quality.

This new deep neural network does not need training data, unlike conventional methods. It includes a physics model within the algorithm, providing an increase in contrast through total variation regularization, and takes into account the wide spectral bandwidth of the source.

The researchers showed that the untrained neural network can be used to work with new light sources, even if the spectrum or beam profile is not known beforehand. They achieved this by using the novel and smallest known Si LED fabricated via a fully commercial, unmodified bulk CMOS microelectronics.

Next-gen On-chip Imaging Systems

The researchers see potential for applying the combination of CMOS micro-LEDs and the neural network to other computational imaging applications, such as compact microscopes for live-cell tracking or spectroscopic imaging of living plant tissues.

Moreover, the study demonstrates the potential for developing next-generation on-chip imaging systems.

Various applications, such as particle tracking, environmental monitoring, biological sample imaging, and metrology, have already utilized in-line holography microscopes. In the future, it could be feasible to array these LEDs in CMOS for the creation of programmable coherent illumination, which could further enhance the complexity of systems, according to the team.

"Our breakthrough represents a proof of concept that could be hugely impactful for numerous applications requiring the use of micro-LEDs," Iksung Kang, lead author and research assistant at MIT, said in a statement.

"For instance, this LED could be combined into an array for higher levels of illumination needed for larger-scale applications. In addition, due to the low cost and scalability of microelectronics CMOS processes, this can be done without increasing the system's complexity, cost, or form factor."

The findings of the team were published in the journal Nature Communications.