Large-scale, highly integrated and low-power-consuming hardware is becoming progressively more important for realizing optical neural networks (ONNs) capable of advanced optical computing. Traditional experimental implementations needN²units such as Mach-Zehnder interferometers (MZIs) for an input dimensionNto realize typical computing operations (convolutions and matrix multiplication), resulting in limited scalability and consuming excessive power. Here, we propose the integrated diffractive optical network for implementing parallel Fourier transforms, convolution operations and application-specific optical computing using two ultracompact diffractive cells (Fourier transform operation) and onlyNMZIs. The footprint and energy consumption scales linearly with the input data dimension, instead of the quadratic scaling in the traditional ONN framework. A ~10-fold reduction in both footprint and energy consumption, as well as equal high accuracy with previous MZI-based ONNs was experimentally achieved for computations performed on theMNISTandFashion-MNISTdatasets. The integrated diffractive optical network (IDNN) chip demonstrates a promising avenue towards scalable and low-power-consumption optical computational chips for optical-artificial-intelligence.

The apparent increase in hormone-induced cancers and disorders of the reproductive tract has led to a growing demand for new technologies capable of detecting endocrine disruptors. However, a long-lasting challenge unaddressed is how to achieve ultrahigh sensitive, continuous, and in situ measurement with a portable device for in-field and remote environmental monitoring. Here we demonstrate a simple-to-implement plasmonic optical fiber biosensing platform to achieve an improved light–matter interaction and advanced surface chemistry for ultrasensitive detection of endocrine disruptors. Our platform is based on a gold-coated highly tilted fiber Bragg grating that excites high-density narrow cladding mode spectral combs that overlap with the broad absorption of the surface plasmon for high accuracy interrogation, hence enabling the ultrasensitive monitoring of refractive index changes at the fiber surface. Through the use of estrogen receptors as the model, we design an estradiol–streptavidin conjugate with the assistance of molecular dynamics, converting the specific recognition of environmental estrogens (EEs) by estrogen receptor into surface-based affinity bioassay for protein. The ultrasensitive platform with conjugate-induced amplification biosensing approach enables the subsequent detection for EEs down to 1.5 × 10⁻³ ng ml⁻¹ estradiol equivalent concentration level, which is one order lower than the defined maximal E₂ level in drinking water set by the Japanese government. The capability to detect EEs down to nanogram per liter level is the lowest limit of detection for any estrogen receptor-based detection reported thus far. Its compact size, flexible shape, and remote operation capability open the way for detecting other endocrine disruptors with ultrahigh sensitivity and in various hard-to-reach spaces, thereby having the potential to revolutionize environment and health monitoring.