3D Deep Learning Enables Accurate Layer Mapping of 2D Materials

Fast adaptive focusing confocal Raman microscopy for large-area two-dimensional materials

Large-area two-dimensional (2D) materials possess significant potential in the development of next generation semiconductor due to their unique physicochemical properties. Confocal Raman spectroscopy (CRM), a typical 2D material characterization method, has a limited effective measurement area owing to the restricted focus depth of the system and the less-than-ideal level of the substrate. We propose fast adaptive focusing confocal Raman microscopy (FAFCRM) to realize real-time focusing detection for large-area 2D materials. By observing spot changes on the charge coupled device (CCD) caused by placing an aperture in front of the CCD, the methodology gives a focusing resolution up to 100 nm per 60 μm without axial scanning. A graphene was measured over 25.6 mm × 25.6 mm area on focus through all the scanning. The research results provide new perspectives for non-destructive characterization of 2D materials at the inch level.

When Machine Learning Meets 2D Materials: A Review

The availability of an ever-expanding portfolio of 2D materials with rich internal degrees of freedom (spin, excitonic, valley, sublattice, and layer pseudospin) together with the unique ability to tailor heterostructures made layer by layer in a precisely chosen stacking sequence and relative crystallographic alignments, offers an unprecedented platform for realizing materials by design. However, the breadth of multi-dimensional parameter space and massive data sets involved is emblematic of complex, resource-intensive experimentation, which not only challenges the current state of the art but also renders exhaustive sampling untenable. To this end, machine learning, a very powerful data-driven approach and subset of artificial intelligence, is a potential game-changer, enabling a cheaper - yet more efficient - alternative to traditional computational strategies. It is also a new paradigm for autonomous experimentation for accelerated discovery and machine-assisted design of functional 2D materials and heterostructures. Here, the study reviews the recent progress and challenges of such endeavors, and highlight various emerging opportunities in this frontier research area.

Identification of triangular single crystals of transition metal dichalcogenides based on the detection algorithm

The distinctive properties and facile integration of 2D materials hold the potential to offer promising avenues for the on-chip photonic devices, and the expeditious and nondestructive identification and localization of diverse fundamental building blocks become key prerequisites. Here, we present a methodology grounded in digital image processing and deep learning, which effectively achieves the detection and precise localization of four monolayer-thick triangular single crystals of transition metal dichalcogenides with the mean average precision above 90%, and the approach demonstrates robust recognition capabilities across varied imaging conditions encompassing both white light and monochromatic light. This stands poised to serve as a potent data-driven tool enhancing the characterizing efficiency and holds the potential to expedite research initiatives and applications founded on the utilization of 2D materials.

A Universal Approach to Determine the Atomic Layer Numbers in Two-Dimensional Materials Using Dark-Field Optical Contrast

Two-dimensional (2D) materials possess unique properties primarily due to the quantum confinement effect, which highly depends on their thicknesses. Identifying the number of atomic layers in these materials is a crucial, yet challenging step. However, the commonly used optical reflection method offers only very low contrast. Here, we develop an approach that shows unprecedented sensitivity by analyzing the brightness of dark-field optical images. The brightness of the 2D material edges has a linear dependence on the number of atomic layers. The findings are modeled by Rayleigh scattering, and the results agree well with the experiments. The relative contrast of single-layer graphene can reach 70% under white-light incident conditions. Furthermore, different 2D materials were successfully tested. By adjusting the exposure conditions, we can identify the number of atomic layers ranging from 1 to over 100. Finally, this approach can be applied to various substrates, even transparent ones, making it highly versatile.

Correlating Optical Microspectroscopy with 4×4 Transfer Matrix Modeling for Characterizing Birefringent Van der Waals Materials

Van der Waals materials exhibit intriguing properties for future electronic and optoelectronic devices. As those unique features strongly depend on the materials' thickness, it has to be accessed precisely for tailoring the performance of a specific device. In this study, a nondestructive and technologically easily implementable approach for accurate thickness determination of birefringent layered materials is introduced by combining optical reflectance measurements with a modular model comprising a 4×4 transfer matrix method and the optical components relevant to light microspectroscopy. This approach is demonstrated being reliable and precise for thickness determination of anisotropic materials like highly oriented pyrolytic graphite and black phosphorus in a range from atomic layers up to more than 100 nm. As a key feature, the method is well-suited even for encapsulated layers outperforming state of-the-art techniques like atomic force microscopy.

Tumor-to-bone distance and radiomic features on MRI distinguish intramuscular lipomas from well-differentiated liposarcomas

Background

To develop a machine learning model based on tumor-to-bone distance and radiomic features derived from preoperative MRI images to distinguish intramuscular (IM) lipomas and atypical lipomatous tumors/well-differentiated liposarcomas (ALTs/WDLSs) and compared with radiologists.

Methods

The study included patients with IM lipomas and ALTs/WDLSs diagnosed between 2010 and 2022, and with MRI scans (sequence/field strength: T1-weighted (T1W) imaging at 1.5 or 3.0 Tesla MRI). Manual segmentation of tumors based on the three-dimensional T1W images was performed by two observers to appraise the intra- and interobserver variability. After radiomic features and tumor-to-bone distance were extracted, it was used to train a machine learning model to distinguish IM lipomas and ALTs/WDLSs. Both feature selection and classification steps were performed using Least Absolute Shrinkage and Selection Operator logistic regression. The performance of the classification model was assessed using a tenfold cross-validation strategy and subsequently evaluated using the receiver operating characteristic curve (ROC) analysis. The classification agreement of two experienced musculoskeletal (MSK) radiologists was assessed using the kappa statistics. The diagnosis accuracy of each radiologist was evaluated using the final pathological results as the gold standard. Additionally, we compared the performance of the model and two radiologists in terms of the area under the receiver operator characteristic curves (AUCs) using the Delong’s test.

Results

There were 68 tumors (38 IM lipomas and 30 ALTs/WDLSs). The AUC of the machine learning model was 0.88 [95% CI 0.72–1] (sensitivity, 91.6%; specificity, 85.7%; and accuracy, 89.0%). For Radiologist 1, the AUC was 0.94 [95% CI 0.87–1] (sensitivity, 97.4%; specificity, 90.9%; and accuracy, 95.0%), and as to Radiologist 2, the AUC was 0.91 [95% CI 0.83–0.99] (sensitivity, 100%; specificity, 81.8%; and accuracy, 93.3%). The classification agreement of the radiologists was 0.89 of kappa value (95% CI 0.76–1). Although the AUC of the model was lower than of two experienced MSK radiologists, there was no statistically significant difference between the model and two radiologists (all P > 0.05).

Conclusions

The novel machine learning model based on tumor-to-bone distance and radiomic features is a noninvasive procedure that has the potential for distinguishing IM lipomas from ALTs/WDLSs. The predictive features that suggested malignancy were size, shape, depth, texture, histogram, and tumor-to-bone distance.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-023-03718-4.

Robustness of Radiomic Features: Two-Dimensional versus Three-Dimensional MRI-Based Feature Reproducibility in Lipomatous Soft-Tissue Tumors

This retrospective study aimed to compare the intra- and inter-observer manual-segmentation variability in the feature reproducibility between two-dimensional (2D) and three-dimensional (3D) magnetic-resonance imaging (MRI)-based radiomic features. The study included patients with lipomatous soft-tissue tumors that were diagnosed with histopathology and underwent MRI scans. Tumor segmentation based on the 2D and 3D MRI images was performed by two observers to assess the intra- and inter-observer variability. In both the 2D and the 3D segmentations, the radiomic features were extracted from the normalized images. Regarding the stability of the features, the intraclass correlation coefficient (ICC) was used to evaluate the intra- and inter-observer segmentation variability. Features with ICC > 0.75 were considered reproducible. The degree of feature robustness was classified as low, moderate, or high. Additionally, we compared the efficacy of 2D and 3D contour-focused segmentation in terms of the effects of the stable feature rate, sensitivity, specificity, and diagnostic accuracy of machine learning on the reproducible features. In total, 93 and 107 features were extracted from the 2D and 3D images, respectively. Only 35 features from the 2D images and 63 features from the 3D images were reproducible. The stable feature rate for the 3D segmentation was more significant than for the 2D segmentation (58.9% vs. 37.6%, p = 0.002). The majority of the features for the 3D segmentation had moderate-to-high robustness, while 40.9% of the features for the 2D segmentation had low robustness. The diagnostic accuracy of the machine-learning model for the 2D segmentation was close to that for the 3D segmentation (88% vs. 90%). In both the 2D and the 3D segmentation, the specificity values were equal to 100%. However, the sensitivity for the 2D segmentation was lower than for the 3D segmentation (75% vs. 83%). For the 2D + 3D radiomic features, the model achieved a diagnostic accuracy of 87% (sensitivity, 100%, and specificity, 80%). Both 2D and 3D MRI-based radiomic features of lipomatous soft-tissue tumors are reproducible. With a higher stable feature rate, 3D contour-focused segmentation should be selected for the feature-extraction process.

A simple and robust machine learning assisted process flow for the layer number identification of TMDs using optical contrast spectroscopy

Layered transition metal dichalcogenides (TMDs) like tungsten disulphide (WS₂) possess a large direct electronic band gap (∼2 eV) in the monolayer limit, making them ideal candidates for opto-electronic applications. The size and nature of the bandgap is strongly dependent on the number of layers. However, different TMDs require different experimental tools under specific conditions to accurately determine the number of layers. Here, we identify the number of layers of WS₂exfoliated on top of SiO₂/Si wafer from optical images using the variation of optical contrast with thickness. Optical contrast is a universal feature that can be easily extracted from digital images. But fine variations in the optical images due to different capturing conditions often lead to inaccurate layer number determination. In this paper, we have implemented a simple Machine Learning assisted image processing workflow that uses image segmentation to eliminate this difficulty. The workflow developed for WS₂is also demonstrated on MoS₂, graphene and h-BN, showing its applicability across various classes of 2D materials. A graphical user interface is provided to enhance the adoption of this technique in the 2D materials research community.

Opportunities and challenges for 2D heterostructures in battery applications: a computational perspective

With an increasing demand for large-scale energy storage systems, there is a need for novel electrode materials to store energy in batteries efficiently. 2D materials are promising as electrode materials for battery applications. Despite their excellent properties, none of the available single-phase 2D materials offers a combination of properties required for maximizing energy density, power density, and cycle life. This article discusses how stacking distinct 2D materials into a 2D heterostructure may open up new possibilities for battery electrodes, combining favourable characteristics and overcoming the drawbacks of constituent 2D layers. Computational studies are crucial to advancing this field rapidly with first-principles simulations of various 2D heterostructures forming the basis for such investigations that offer insights into processes that are hard to determine otherwise. We present a perspective on the current methodology, along with a review of the known 2D heterostructures as anodes and their potential for Li and Na-ion battery applications. 2D heterostructures showcase excellent tunability with different compositions. However, each of them has distinct properties, with its own set of challenges and opportunities for application in batteries. We highlight the current status and prospects to stimulate research into designing new 2D heterostructures for battery applications.

Universal image segmentation for optical identification of 2D materials

Machine learning methods are changing the way data is analyzed. One of the most powerful and widespread applications of these techniques is in image segmentation wherein disparate objects of a digital image are partitioned and classified. Here we present an image segmentation program incorporating a series of unsupervised clustering algorithms for the automatic thickness identification of two-dimensional materials from digital optical microscopy images. The program identifies mono- and few-layer flakes of a variety of materials on both opaque and transparent substrates with a pixel accuracy of roughly 95%. Contrasting with previous attempts, application generality is achieved through preservation and analysis of all three digital color channels and Gaussian mixture model fits to arbitrarily shaped data clusters. Our results provide a facile implementation of data clustering for the universal, automatic identification of two-dimensional materials exfoliated onto any substrate.

Deep Neural Networks for Analysis of Microscopy Images—Synthetic Data Generation and Adaptive Sampling

The analysis of microscopy images has always been an important yet time consuming process in materials science. Convolutional Neural Networks (CNNs) have been very successfully used for a number of tasks, such as image segmentation. However, training a CNN requires a large amount of hand annotated data, which can be a problem for material science data. We present a procedure to generate synthetic data based on ad hoc parametric data modelling for enhancing generalization of trained neural network models. Especially for situations where it is not possible to gather a lot of data, such an approach is beneficial and may enable to train a neural network reasonably. Furthermore, we show that targeted data generation by adaptively sampling the parameter space of the generative models gives superior results compared to generating random data points.