Scientists from the Singapore MIT Research and Technology Alliance have developed the world's smallest LED (Light Emitting Diode). This new type of LED can be used to build the smallest holographic microscope to date, allowing existing cameras on mobile phones to be converted into microscopes simply by modifying silicon chips and software. The relevant research was published in the recent journal Optics. This breakthrough is supported by revolutionary neural network algorithms that can reconstruct objects observed by holographic microscopes, enhance the examination of microscopic objects such as cells and bacteria, without the need for bulky traditional microscopes or additional optical devices. Most of the light in photonic chips comes from outside the chip, which leads to low overall energy efficiency and fundamentally limits the scalability of the chip. The smallest silicon emitter developed by the team this time has a light intensity comparable to the most advanced large-area silicon emitter currently available. The new LED exhibits high spatial intensity (102 ± 48 milliwatts/square centimeter) at room temperature and has the smallest emission area (0.09 ± 0.04 square micrometers) among all known silicon emitters. To demonstrate potential practical applications, researchers then integrated this LED into an online, centimeter scale all silicon holographic microscope that does not require lenses or pinholes. They also constructed a novel, untrained deep neural network architecture that enables holographic microscopy to reconstruct images and improve image quality. Unlike traditional methods that require training, the new neural network architecture eliminates the need for training by embedding physical models in the algorithm, allowing researchers to use new light sources without prior knowledge of the light source spectrum or beam profile. This collaborative combination of micro LEDs and neural networks can be used for other computational imaging purposes, such as compact microscopes for tracking living cells or spectral imaging of biological tissues such as living plants. This research also paves the way for significant progress in photonics. (New News Agency)