Researchers at the University of California, Los Angeles (UCLA) have developed an all-optical complex field imager capable of capturing both amplitude and phase information of optical fields without the need for digital processing. The researchers believe that the imager could be used in fields such as biomedical imaging, security, sensing, and material science. Traditional optical imaging technologies have relied on intensity-based sensors that can only capture the amplitude of light, leaving out crucial phase information. Phase information can grant insight into structural properties such as absorption and refractive index distributions, which are essential for detailed sample analysis. Current methods to capture phase information involve complex interferometric or holographic systems supplemented by iterative phase retrieval algorithms, resulting in increased hardware complexity and computational demand. A depiction of the UCLA research team’s complex field imager. Using diffractive processors, the imager acquires both amplitude and phase information without digital processing. Courtesy of the Ozcan Research Lab/UCLA. Led by UCLA professor Aydogan Ozcan, the team created a novel complex field imager that overcomes these limitations by using a series of deep learning-optimized diffractive surfaces to modulate incoming complex fields. The approach takes away the need for any digital reconstruction algorithms, as the surfaces create two independent imaging channels that transform the amplitude and phase of the input fields into intensity distributions on the sensor plane. The complex field imager consists of spatially engineered diffractive surfaces arranged to perform amplitude-to-amplitude and phase-to-intensity transformations. The transformations allow the device to directly measure the amplitude and phase profiles of input complex fields. The imager is also highly integrable into existing optical systems, with a compact optical design that spans approximately 100 wavelengths axially. The results during experimentation of the imager showed a high degree of accuracy when validated using 3D-printed prototypes operating in the terahertz spectrum. This resulted in the output amplitude and phase channel images closely matching previously made numerical simulations. The research team said that it will continue to expand upon its designs and develop them for real world solutions. They have so far identified application areas such as the biomedical field where the imager could be used for real-time, non-invasive imaging of tissues and cells with its compact design making it suitable for integration into endoscopic devices and miniature microscopes. Other applications could include environmental monitoring and industrial applications, where the imager could facilitate the development of portable lab-on-a-chip sensors for rapid detection of microorganisms and pollutants, as well as the rapid inspection of materials. The research was published in Light: Science & Applications (www.doi.org/10.1038/s41377-024-01482-6).