Direct Imaging of Integrated Circuits in CPU with 60 nm Super-Resolution Optical Microscope.

Far-field super-resolution optical microscopies have achieved incredible success in life science for visualization of vital nanostructures organized in single cells. However, such resolution power has been much less extended to material science for inspection of human-made ultrafine nanostructures, simply because the current super-resolution optical microscopies modalities are rarely applicable to nonfluorescent samples or unlabeled systems. Here, we report an antiphase demodulation pump-probe (DPP) super-resolution microscope for direct optical inspection of integrated circuits (ICs) with a lateral resolution down to 60 nm. Because of the strong pump-probe (PP) signal from copper, we performed label-free super-resolution imaging of multilayered copper interconnects on a small central processing unit (CPU) chip. The label-free super-resolution DPP optical microscopy opens possibilities for easy, fast, and large-scale electronic inspection in the whole pipeline chain for designing and manufacturing ICs.

DDeep3M+: adaptive enhancement powered weakly supervised learning for neuron segmentation.

Significance: Robust segmentations of neurons greatly improve neuronal population reconstruction, which could support further study of neuron morphology for brain research. Aim: Precise segmentation of 3D neuron structures from optical microscopy (OM) images is crucial to probe neural circuits and brain functions. However, the high noise and low contrast of images make neuron segmentation challenging. Convolutional neural networks (CNNs) can provide feasible solutions for the task but they require large manual labels for training. Labor-intensive labeling is highly expensive and heavily limits the algorithm generalization. Approach: We devise a weakly supervised learning framework Docker-based deep network plus (DDeep3M+) for neuron segmentation without any manual labeling. A Hessian analysis based adaptive enhancement filter is employed to generate pseudo-labels for segmenting neuron images. The automated segmentation labels are input for training a DDeep3M to extract neuronal features. We mine more undetected weak neurites from the probability map based on neuronal structures, thereby modifying the pseudo-labels. We iteratively refine the pseudo-labels and retrain the DDeep3M model with the pseudo-labels to obtain a final segmentation result. Results: The proposed method achieves promising results with the F1 score of 0.973, which is close to that of the CNN model with manual labels and superior to several segmentation algorithms. Conclusions: We propose an accurate weakly supervised neuron segmentation method. The high precision results achieved on 3D OM datasets demonstrate the superior generalization of our DDeep3M+.

Imaging Chemical Kinetics of Radical Polymerization with an Ultrafast Coherent Raman Microscope.

Numerous mechanisms have been proposed for polymerization to provide qualitative and quantitative prediction of how monomers spatially and temporally arrange into the polymeric chains. However, less is known about this process at the molecular level because the ultrafast chemical reaction is inaccessible for any form of microscope so far. Here, to address this unmet challenge, a stimulated Raman scattering microscope based on collinear multiple beams (COMB-SRS) is demonstrated, which allows label-free molecular imaging of polymer synthesis in action at speed of 2000 frames per second. The field of view of the developed 2 kHz SRS microscope is 30 × 28 µm2 with 50 × 46 pixels and 7 µs dwell time. By catching up the speed of chemical reaction, COMB-SRS is able to quantitatively visualize the ultrafast dynamics of molecular vibrations with submicron spatial resolution and sub-millisecond temporal resolution. The propagating polymer waves driven by reaction rate and persistent UV initiation are observed in situ. This methodology is expected to permit the development of novel functional polymers, controllable photoresists, 3D printing, and other new polymerization technologies.