Standardizing image assessment in optical diffraction tomography.

Optical diffraction tomography (ODT) has gradually become a popular label-free imaging technique that offers diffraction-limited resolution by mapping an object's three-dimensional (3D) refractive index (RI) distribution. However, there is a lack of comprehensive quantitative image assessment metrics in ODT for studying how various experimental conditions influence image quality, and subsequently optimizing the experimental conditions. In this Letter, we propose to standardize the image assessment in ODT by proposing a set of metrics, including signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and structural distinguishability (SD). To test the feasibility of the metrics, we performed experiments on angle-scanning ODT by varying the number of illumination angles, RI contrast of samples, sample feature sizes, and sample types (e.g., standard polystyrene beads and 3D printed structures) and evaluated the RI tomograms with SNR, CNR, and SD. We further quantitatively studied how image quality can be improved, and tested the image assessment metrics on subcellular structures of living cells. We envision the proposed image assessment metrics may greatly benefit end-users for assessing the RI tomograms, as well as experimentalists for optimizing ODT instruments.

Biomimetic thermoresponsive superstructures by colloidal soft-and-hard co-assembly.

Soft-and-hard hybrid structures are ubiquitous in biological systems and have inspired the design of man-made mechanical devices, actuators, and robots. The realization of these structures, however, has been challenging at microscale, where material integration and actuation become exceedingly less practical. Here, through simple colloidal assembly, we create microscale superstructures consisting of soft and hard materials, which, serving as microactuators, have thermoresponsive shape-transforming properties. In this case, anisotropic metal-organic framework (MOF) particles as the hard components are integrated with liquid droplets, forming spine-mimicking colloidal chains via valence-limited assembly. The chains, with alternating soft and hard segments, are referred to as MicroSpine and can reversibly change shape, switching between straight and curved states through a thermoresponsive swelling/deswelling mechanism. By solidification of the liquid parts within a chain with prescribed patterns, we design various chain morphologies, such as "colloidal arms," with controlled actuating behaviors. The chains are further used to build colloidal capsules, which encapsulate and release guests by the temperature-programmed actuation.