Raman spectroscopy has become an important single-cell analysis tool for monitoring biochemical changes at the cellular level. However, Raman spectral data, typically presented as continuous data with high-dimensional characteristics, is distinct from discrete sequences, which limits the application of deep learning-based algorithms in data analysis due to the lack of discretization. Herein, a model called fragment-fusion transformer is proposed, which integrates the discrete fragmentation of continuous spectra based on their intrinsic characteristics with the extraction of intrafragment features and the fusion of interfragment features. The model integrates the intrinsic feature-based fragmentation of spectra with transformer, constructing the fragment transformer block for feature extraction within fragments. Interfragment information is combined through the pyramid design structure to improve the model\'s receptive field and fully exploit the spectral properties. During the pyramidal fusion process, the information gain of the final extracted features in the spectrum has been enhanced by a factor of 9.24 compared to the feature extraction stage within the fragment, and the information entropy has been enhanced by a factor of 13. The fragment-fusion transformer achieved a spectral recognition accuracy of 94.5%, which is 4% higher compared to the method without fragmentation and fusion processes on the test set of cell Raman spectroscopy identification experiments. In comparison to common spectral classification models such as KNN, SVM, logistic regression, and CNN, fragment-fusion transformer has achieved 4.4% higher accuracy than the best-performing CNN model. Fragment-fusion transformer method has the potential to serve as a general framework for discretization in the field of continuous spectral data analysis and as a research tool for analyzing the intrinsic information within spectra.

Total volatile basic nitrogen (TVB-N) and total viable count (TVC) are important freshness indicators of meat. Hyperspectral imaging combined with chemometrics has been proven to be effective in meat detection. However, a challenge with chemometrics is the lack of a universally applicable processing combination, requiring trial-and-error experiments with different datasets. This study proposes an end-to-end deep learning model, pyramid attention features fusion model (PAFFM), integrating CNN, attention mechanism and pyramid structure. PAFFM fuses the raw visible and near-infrared range (VNIR) and shortwave near-infrared range (SWIR) spectral data for predicting TVB-N and TVC in chicken breasts. Compared with the CNN and chemometric models, PAFFM obtains excellent results without a complicated processing combinatorial optimization process. Important wavelengths that contributed significantly to PAFFM performance are visualized and interpreted. This study offers valuable references and technical support for the market application of spectral detection, benefiting related research and practical fields.

NH3is widely existed in the environment and is closely associated with various health issues. Additionally, detecting the small amounts of NH3exhaled by patients with liver and kidney diseases offers potential opportunities for painless early disease diagnosis. Therefore, there is an urgent need for a convenient, rapid, and highly sensitive real-time NH3monitoring method. This work presents a high-performance NH3sensor based on olfactory receptor-derived peptides (ORPs) on a pyramid silicon nanowires (SiNWs) structure substrate. First, we successfully fabricated the pyramid-SiNWs structure on a silicon substrate using a chemical etching method. Subsequently, by dehydrative condensation reaction between the amino groups on APTES and the carboxyl groups of ORPs, ORPs were successfully immobilized onto the pyramid-SiNWs structure. This methodology allows the ORPs sensor on the pyramid-SiNWs substrate to detect NH3as low as 1 ppb, which was the reported lowest limit of detection, with a higher response rate compared to ORPs sensors on flat SiNWs substrates. The sensors also exhibit good sensitivity and stability for NH3gas detection. The results show the feasibility and potential applications of ORPs-pyramid-SiNWs structure sensors, in the fields of food safety, disease monitoring, and environmental protection, etc.

Bi2Se3, as a novel 3D topological insulator (TI), is expected to be a strong candidate for next-generation optoelectronic devices due to its intriguing optical and electrical properties. In this study, a series of Bi2Se3 films with different thicknesses of 5-40 nm were successfully prepared on planar-Si substrates and developed as self-powered light position-sensitive detectors (PSDs) by introducing lateral photovoltaic effect (LPE). It is demonstrated that the Bi2Se3/planar-Si heterojunction shows a broad-band response range of 450-1064 nm, and the LPE response is strongly dependent on the Bi2Se3 layer thickness, which can be mainly attributed to the thickness-modulated longitudinal carrier separation and transport. The 15 nm thick PSD shows the best performance with a position sensitivity of up to 89.7 mV/mm, a nonlinearity of lower than 7%, and response time as fast as 62.6/49.4 μs. Moreover, to further enhance the LPE response, a novel Bi2Se3/pyramid-Si heterojunction is built by constructing a nanopyramid structure for the Si substrate. Owing to the improvement of the light absorption capability in the heterojunction, the position sensitivity is largely boosted up to 178.9 mV/mm, which gets an increment of 199% as compared with that of the Bi2Se3/planar-Si heterojunction device. At the same time, the nonlinearity is still kept within 10% as well due to the excellent conduction property of the Bi2Se3 film. In addition, an ultrafast response speed of 173/97.4 μs is also achieved in the newly proposed PSD with excellent stability and reproducibility. This result not only demonstrates the great potential of TIs in PSD but also provides a promising approach for tuning its performance.

Microwave-absorbing materials with wide bandwidth and high absorptivity are increasingly playing an important role in over-the-air (OTA) testing. In this work, a kind of pyramid absorbing material was prepared using flame-retardant absorbers as the filler. In addition, a coating was used to further improve the flame-retardant properties of the microwave-absorbing material. To obtain excellent microwave absorption performance (MWAP), a high-frequency structure simulator (HFSS) was adopted to design structural materials. Here, the total height, the base height, the decapitation height of the pyramid tip, the distance between the pyramids, and other parameters were analyzed; then, the actual processing and molding were realized. The MWAP of -30 dB was achieved at 2.7-18 GHz, and the MWAP of -10 dB was also met at 2-18 GHz. In particular, the study also investigated the MWAP of large angle, which can meet the MWAP of -10 dB at 2-18 GHz and MWAP of -30 dB at 4-18 GHz. Most importantly, the absorption mechanism of the pyramid structure was explored. The influence of the tip was proved by the distribution of the electromagnetic field in the pyramid. It can be regarded as a multilayer microwave-absorbing material due to the impedance gradient of the pyramid, which can provide an effective research idea and method for future engineering applications.

Ultrasound (US) is an important imaging modality used to assess breast lesions for malignant features. In the past decade, many machine learning models have been developed for automated discrimination of breast cancer versus normal on US images, but few have classified the images based on the Breast Imaging Reporting and Data System (BI-RADS) classes. This work aimed to develop a model for classifying US breast lesions using a BI-RADS classification framework with a new multi-class US image dataset. We proposed a deep model that combined a novel pyramid triple deep feature generator (PTDFG) with transfer learning based on three pre-trained networks for creating deep features. Bilinear interpolation was applied to decompose the input image into four images of successively smaller dimensions, constituting a four-level pyramid for downstream feature generation with the pre-trained networks. Neighborhood component analysis was applied to the generated features to select each network\'s 1,000 most informative features, which were fed to support vector machine classifier for automated classification using a ten-fold cross-validation strategy. Our proposed model was validated using a new US image dataset containing 1,038 images divided into eight BI-RADS classes and histopathological results. We defined three classification schemes: Case 1 involved the classification of all images into eight categories; Case 2, classification of breast US images into five BI-RADS classes; and Case 3, classification of BI-RADS 4 lesions into benign versus malignant classes. Our PTDFG-based transfer learning model attained accuracy rates of 79.29%, 80.42%, and 88.67% for Case 1, Case 2, and Case 3, respectively.

In this study, we developed a novel three-dimensional (3D) cancer cell chip using a three-floor hierarchical 3D pyramid structure (3D pyramid) to simulate 3D tumor cell growth in vitro and to detect anticancer drugs. The proposed 3D pyramidbased cancer cell chip offered substantial advantages for the agglomerate formation of tumor cells, in which cells could be maintained as tumor spheroids for up to 3 weeks. Soon after HeLa tumor cells adhered to the micropatterned pillar sidewalls, they were suspended between the pillars based on scanning electron microscopy images. Treatment with the anticancer drug oleanolic acid resulted in 46.33% and 5.86% apoptotic cells on the 2D plate and 3D pyramid-based cell chip, respectively, compared with only 0.06% apoptotic cells in the control. The increase in chemoresistance to anticancer drugs in the 3D pyramid-based cell chip might be a result of cell confluence and hypoxia due to the spheroid formation of tumor cells in the 3D pyramid structure. These results indicated that the proposed cell chip could potentially be used for anticancer drug screening or can be incorporated into other models aimed at prolonging various cell functions in culture.