Adaptive and Iterative Learning with Multi-perspective Regularizations for Metal Artifact Reduction.

Metal artifact reduction (MAR) is important for clinical diagnosis with CT images. The existing state-of-the-art deep learning methods usually suppress metal artifacts in sinogram or image domains or both. However, their performance is limited by the inherent characteristics of the two domains, i.e., the errors introduced by local manipulations in the sinogram domain would propagate throughout the whole image during backprojection and lead to serious secondary artifacts, while it is difficult to distinguish artifacts from actual image features in the image domain. To alleviate these limitations, this study analyzes the desirable properties of wavelet transform in-depth and proposes to perform MAR in the wavelet domain. First, wavelet transform yields components that possess spatial correspondence with the image, thereby preventing the spread of local errors to avoid secondary artifacts. Second, using wavelet transform could facilitate identification of artifacts from image since metal artifacts are mainly high-frequency signals. Taking these advantages of the wavelet transform, this paper decomposes an image into multiple wavelet components and introduces multi-perspective regularizations into the proposed MAR model. To improve the transparency and validity of the model, all the modules in the proposed MAR model are designed to reflect their mathematical meanings. In addition, an adaptive wavelet module is also utilized to enhance the flexibility of the model. To optimize the model, an iterative algorithm is developed. The evaluation on both synthetic and real clinical datasets consistently confirms the superior performance of the proposed method over the competing methods.

Oriented bouncing of droplets with a small Weber number on inclined one-dimensional nanoforests.

Asymmetric superhydrophobic structures with anisotropic wettability can achieve directional bouncing of droplets and thus can have applications in directional self-cleaning, liquid transportation, and heat transfer. To achieve convenient large-scale preparation of asymmetric superhydrophobic surfaces, inclined nanoforests are prepared in this work using a technique of competitive ablation polymerization, which allows the control of the inclined angles, diameters, and heights of the nanostructures. In this study, such asymmetric structures with the smallest dimension (230 nm diameter) known are achieved by a simple etching method to guide droplet unidirectional bouncing. With such nanoforests, the mechanism of droplet bouncing on their surface is investigated, and controllable droplet bouncing over a long distance is achieved using droplets with a low Weber number. The proposed structure has a promising future in directional self-cleaning, liquid transportation and heat transfer.

Wavelet-Inspired Multi-channel Score-based Model for Limited-angle CT Reconstruction.

Score-based generative model (SGM) has demonstrated great potential in the challenging limited-angle CT (LA-CT) reconstruction. SGM essentially models the probability density of the ground truth data and generates reconstruction results by sampling from it. Nevertheless, direct application of the existing SGM methods to LA-CT suffers multiple limitations. Firstly, the directional distribution of the artifacts attributing to the missing angles is ignored. Secondly, the different distribution properties of the artifacts in different frequency components have not been fully explored. These drawbacks would inevitably degrade the estimation of the probability density and the reconstruction results. After an in-depth analysis of these factors, this paper proposes a Wavelet-Inspired Score-based Model (WISM) for LA-CT reconstruction. Specifically, besides training a typical SGM with the original images, the proposed method additionally performs the wavelet transform and models the probability density in each wavelet component with an extra SGM. The wavelet components preserve the spatial correspondence with the original image while performing frequency decomposition, thereby keeping the directional property of the artifacts for further analysis. On the other hand, different wavelet components possess more specific contents of the original image in different frequency ranges, simplifying the probability density modeling by decomposing the overall density into component-wise ones. The resulting two SGMs in the image-domain and wavelet-domain are integrated into a unified sampling process under the guidance of the observation data, jointly generating high-quality and consistent LA-CT reconstructions. The experimental evaluation on various datasets consistently verifies the superior performance of the proposed method over the competing method.

Wafer-Level, High-Performance, Flexible Sensors Based on Organic Nanoforests for Human-Machine Interactions.

High-performance flexible sensors are essential for real-time information analysis and constructing noncontact communication modules for emerging human-machine interactions. In these applications, batch fabrication of sensors that exhibit high performance at the wafer level is in high demand. Here, we present organic nanoforest-based humidity sensor (NFHS) arrays on a 6 in. flexible substrate prepared via a facile, cost-effective manufacturing approach. Such an NFHS achieves state-of-the-art overall performance: high sensitivity and fast recovery time; the best properties are at a small device footprint. The high sensitivity (8.84 pF/% RH) and fast response time (5 s) of the as-fabricated organic nanoforests are attributed to the abundant hydrophilic groups, the ultra-large surface area with a huge number of nanopores, and the vertically distributed structures beneficial to the transfer of molecules up and down. The NFHS also exhibits excellent long-term stability (90 days), superior mechanical flexibility, and good performance repeatability after bending. With these superiorities, the NFHS is further applied as a smart noncontact switch, and the NFHS array is used as the motion trajectory tracker. The wafer-level batch fabrication capability of our NFHS provides a potential strategy for developing practical applications of such humidity sensors.

A Thermopile-Based Gas Flow Sensor with High Sensitivity for Noninvasive Respiration Monitoring.

In this work, a N/P polySi thermopile-based gas flow device is presented, in which a microheater distributed in a comb-shaped structure is embedded around hot junctions of thermocouples. The unique design of the thermopile and the microheater effectively enhances performance of the gas flow sensor leading to a high sensitivity (around 6.6 µV/(sccm)/mW, without amplification), fast response (around 35 ms), high accuracy (around 0.95%), and mood long-term stability. In addition, the sensor has the advantages of easy production and compact size. With such characteristics, the sensor is further used in real-time respiration monitoring. It allows detailed and convenient collection of respiration rhythm waveform with sufficient resolution. Information such as respiration periods and amplitudes can be further extracted to predict and alert of potential apnea and other abnormal status. It is expected that such a novel sensor could provide a new approach for respiration monitoring related noninvasive healthcare systems in the future.

A Highly Accurate Method for Measuring Response Time of MEMS Thermopiles.

The response time is an important parameter for thermopiles sensors, which reflects the response speed of the device. The accurate measurement of response time is extremely important to evaluate device characteristics for using them in suitable scenarios. In this work, to accurately measure the response time of thermopile sensors, an Al microheater is integrated in a MEMS thermopile as an in situ heat source. Compared with the traditional chopper measurement method for response time, this approach avoids mechanical delay induced by chopper blades. Accordingly, based on this approach, the response time of the device is measured to be 6.9 ms, while that is 12.7 ms when a chopping system is used, demonstrating that an error of at least 5.8 ms is avoided. Such an approach is quite simple to realize and provides a novel route to accurately measure the response time.

Overview of the MEMS Pirani Sensors.

Vacuum equipment has a wide range of applications, and vacuum monitoring in such equipment is necessary in order to meet practical applications. Pirani sensors work by using the effect of air density on the heat conduction of the gas to cause temperature changes in sensitive structures, thus detecting the pressure in the surrounding environment and thus vacuum monitoring. In past decades, MEMS Pirani sensors have received considerable attention and practical applications because of their advances in simple structures, long service life, wide measurement range and high sensitivity. This review systematically summarizes and compares different types of MEMS Pirani sensors. The configuration, material, mechanism, and performance of different types of MEMS Pirani sensors are discussed, including the ones based on thermistors, thermocouples, diodes and surface acoustic wave. Further, the development status of novel Pirani sensors based on functional materials such as nanoporous materials, carbon nanotubes and graphene are investigated, and the possible future development directions for MEMS Pirani sensors are discussed. This review is with the purpose to focus on a generalized knowledge of MEMS Pirani sensors, thus inspiring the investigations on their practical applications.

A nanoforest-based humidity sensor for respiration monitoring.

Traditional humidity sensors for respiration monitoring applications have faced technical challenges, including low sensitivity, long recovery times, high parasitic capacitance and uncalibrated temperature drift. To overcome these problems, we present a triple-layer humidity sensor that comprises a nanoforest-based sensing capacitor, a thermistor, a microheater and a reference capacitor. When compared with traditional polyimide-based humidity sensors, this novel device has a sensitivity that is improved significantly by 8 times within a relative humidity range of 40-90%. Additionally, the integration of the microheater into the sensor can help to reduce its recovery time to 5 s. The use of the reference capacitor helps to eliminate parasitic capacitance, and the thermistor helps the sensor obtain a higher accuracy. These unique design aspects cause the sensor to have an excellent humidity sensing performance in respiration monitoring applications. Furthermore, through the adoption of machine learning algorithms, the sensor can distinguish different respiration states with an accuracy of 94%. Therefore, this humidity sensor design is expected to be used widely in both consumer electronics and intelligent medical instrument applications.

A Thermopile Infrared Sensor Array Pixel Monolithically Integrated with an NMOS Switch.

In this article, we present the design, fabrication, and characterization of a thermopile infrared sensor array (TISA) pixel. This TISA pixel is composed of a dual-layer p+/n- poly-Si thermopile with a closed membrane and an n-channel metal oxide semiconductor (NMOS) switch. To address the challenges in fabrication through the 3D integration method, the anode of the thermopile is connected to the drain of the NMOS, both of which are fabricated on the same bulk wafer using a CMOS compatible monolithic integration process. During a single process sequence, deposition, etching, lithography, and ion implantation steps are appropriately combined to fabricate the thermopile and the NMOS simultaneously. At the same time as ensuring high thermoelectric characteristics of the dual-layer p+/n- poly-Si thermopile, the basic switching functions of NMOS are achieved. Compared with a separate thermopile, the experimental results show that the thermopile integrated with the NMOS maintains a quick response, high sensitivity and high reliability. In addition, the NMOS employed as a switch can effectively and quickly control the readout of the thermopile sensing signal through the voltage, both on and off, at the gate of NMOS. Thus, such a TISA pixel fabricated by the monolithic CMOS-compatible integration approach is low-cost and high-performance, and can be applied in arrays for high-volume production.

Electrostatic tweezer for droplet manipulation.

Various physical tweezers for manipulating liquid droplets based on optical, electrical, magnetic, acoustic, or other external fields have emerged and revolutionized research and application in medical, biological, and environmental fields. Despite notable progress, the existing modalities for droplet control and manipulation are still limited by the extra responsive additives and relatively poor controllability in terms of droplet motion behaviors, such as distance, velocity, and direction. Herein, we report a versatile droplet electrostatic tweezer (DEST) for remotely and programmatically trapping or guiding the liquid droplets under diverse conditions, such as in open and closed spaces and on flat and tilted surfaces as well as in oil medium. DEST, leveraging on the coulomb attraction force resulting from its electrostatic induction to a droplet, could manipulate droplets of various compositions, volumes, and arrays on various substrates, offering a potential platform for a series of applications, such as high-throughput surface-enhanced Raman spectroscopy detection with single measuring time less than 20 s.

A Thermopile Detector Based on Micro-Bridges for Heat Transfer.

A thermopile detector with their thermocouples distributed in micro-bridges is designed and investigated in this work. The thermopile detector consists of 16 pairs of n-poly-Si/p-poly-Si thermocouples, which are fabricated using a low-cost, high-throughput CMOS process. The micro-bridges are realized by forming micro trenches at the front side first and then releasing the silicon substrate at the back side. Compared with a thermopile device using a continuous membrane, the micro-bridge-based one can achieve an improvement of the output voltage by 13.8% due to a higher temperature difference between the hot and cold junctions as there is a decrease in thermal conduction loss in the partially hollowed structure. This technique provides an effective way for developing high-performance thermopile detectors and other thermal devices.

The Measurement and Analysis of Impedance Response of HeLa Cells to Distinct Chemotherapy Drugs.

Electric cell-substrate impedance sensing exhibits a real-time and label-free feature to monitor the response of cells stimulated by various biochemical and mechanical signals. Alterations in the currents passing through the cell-electrode system characterize the impedance variations of cells. The impedance responses of HeLa cells under distinct chemotherapy drugs combine the effects of cell proliferation and cell-substrate adhesion. Optimal interdigitated electrodes were selected to explore the impedance responses of HeLa cells. Measurements of impedance of cells in response to three widely used chemotherapy drugs in clinical practice, namely cisplatin, doxorubicin, 5-fluorouracil, were performed. The results demonstrated that distinct impedance responses of HeLa cells to drugs were exhibited and a decrease in measured impedance was observed after drug treatment, accompanied by alterations in the distribution and intensity of the adhesion-related protein vinculin and the rate of cell proliferation. The link between the impedance profiles of HeLa cells and their biological functions was developed based on the circuit model. This study demonstrated the weights of cell proliferation and adhesion of HeLa cells under the treatments of DDP, DOX, and 5-FU, resulted in distinct impedance responses of cells, providing an impedance-based evaluation methodology for cervical cancer treatment.

Effect of cell-nanostructured substrate interactions on the capture efficiency of HeLa cells.

Circulating tumor cells (CTCs) are regarded as an effective biomarker for cancer detection, diagnosis and prognosis monitoring. CTCs capture based on nanostructured substrates is a powerful technique. Some specific adhesion molecule antibody coated on the surface of nanostructured substrates, such as EpCAM, is commonly used to enhance the CTCs capture efficiency. Substrate nanotopographies regulate the interaction between the substrates and captured cells, further influencing cell capture efficiency. However, the relationship between cell capture efficiency and cell-substrate interaction remains poorly understood. Here, we explored the relationship between cell capture efficiency and cell-substrate interaction based on two sets of nanostructures with different nanotopographies without antibody conjugation. Given the urgent demand for improving the capture efficiency of EpCAM-negative cells, we used HeLa (EpCAM-negative) cells as the main targets. We demonstrated that HeLa cells could be more effectively captured by two nanostructural substrates, especially by double-layer composite nanoforests. Therefore, the morphological and migrating interaction between HeLa cells and distinct substrates was associated with cell capture efficiency. Our findings demonstrated the potential mechanism for optimizing the nanotopography for higher capture efficiency, and provide a potential foundation for cancer detection, diagnosis and treatment.

The Study of Reactive Ion Etching of Heavily Doped Polysilicon Based on HBr/O2/He Plasmas for Thermopile Devices.

Heavily doped polysilicon layers have been widely used in the fabrication of microelectromechanical systems (MEMS). However, the investigation of high selectivity, anisotropy, and excellent uniformity of heavily doped polysilicon etching is limited. In this work, reactive ion etching of undoped and heavily doped polysilicon-based hydrogen bromide (HBr) plasmas have been compared. The mechanism of etching of heavily doped polysilicon is studied in detail. The final results demonstrate that the anisotropy profile of heavily doped polysilicon can be obtained based on a HBr plasma process. An excellent uniformity of resistance of the thermocouples reached ± 2.11%. This technology provides an effective away for thermopile and other MEMS devices fabrication.

Self-Concentrated Surface-Enhanced Raman Scattering-Active Droplet Sensor with Three-Dimensional Hot Spots for Highly Sensitive Molecular Detection in Complex Liquid Environments.

In this work, a surface-enhanced Raman scattering (SERS)-active droplet with three-dimensional (3D) hot spots prepared from a superhydrophobic SERS substrate, which is inspired by the nut wizard strategy, was developed for ultrasensitive detection in complex liquid environments. The SERS substrate was composed of silver-capped parylene C-coated carbon nanoparticles (Ag-PC@CNPs). Such a SERS substrate was prepared by candle-soot deposition to provide a porous carbon nanoparticle layer followed by deposition of a parylene C film to protect the CNPs and then sputtering of silver nanoparticles. Similar to a nut wizard, a droplet rolling on the Ag-PC@CNP-coated substrate picked up the Ag-PC@CNPs. In this way, a self-concentrated and extremely sensitive SERS-active droplet sensor with 3D hot spots was formed. The sensor did not require precise laser focusing and showed relatively high repeatability and much higher sensitivity than those of a corresponding SERS substrate with two-dimensional hot spots. The sensor also achieved high sensitivity and specificity in complex liquid environments; in addition, bovine serum albumin with a concentration as low as 1 pM can be achieved. Consequently, an extremely simple, flexible, and highly sensitive SERS detection technique applicable to liquid biopsy analysis is anticipated.

Nanotechnology-Assisted Isolation and Analysis of Circulating Tumor Cells on Microfluidic Devices.

Circulating tumor cells (CTCs), a type of cancer cell that spreads from primary tumors into human peripheral blood and are considered as a new biomarker of cancer liquid biopsy. It provides the direction for understanding the biology of cancer metastasis and progression. Isolation and analysis of CTCs offer the possibility for early cancer detection and dynamic prognosis monitoring. The extremely low quantity and high heterogeneity of CTCs are the major challenges for the application of CTCs in liquid biopsy. There have been significant research endeavors to develop efficient and reliable approaches to CTC isolation and analysis in the past few decades. With the advancement of microfabrication and nanomaterials, a variety of approaches have now emerged for CTC isolation and analysis on microfluidic platforms combined with nanotechnology. These new approaches show advantages in terms of cell capture efficiency, purity, detection sensitivity and specificity. This review focuses on recent progress in the field of nanotechnology-assisted microfluidics for CTC isolation and detection. Firstly, CTC isolation approaches using nanomaterial-based microfluidic devices are summarized and discussed. The different strategies for CTC release from the devices are specifically outlined. In addition, existing nanotechnology-assisted methods for CTC downstream analysis are summarized. Some perspectives are discussed on the challenges of current methods for CTC studies and promising research directions.

Wafer-level fabrication of 3D nanoparticles assembled nanopillars and click chemistry modification for sensitive SERS detection of trace carbonyl compounds.

In this work, we develop a new method for fabricating wafer-level gold nanoparticles covered silicon nanopillars (SNPs) combined with surface chemical modification to detect trace level carbonyl compounds based on surface-enhanced Raman scattering (SERS) technique. The SNPs are fabricated with an etching process using nano masks synthesized in oxygen-plasma bombardment of photoresist, and further deposited with gold nanoparticles on the surface, thus forming a 3D 'particles on pillars' nanostructure for sensitive SERS detection. The enhancement factor (EF) of the devices for R6G detection can achieve 1.56 × 106 times compared with a flat Si substrate. We also developed an oximation click chemistry reaction procedure by chemically modifying the nanostructures with aminooxy dodecane thiol (ADT) self-assemble modification. The chip is further integrated with a polydimethylsiloxane (PDMS) microfluidic chamber, which allows fast and convenient detection of trace carbonyl compounds in liquid samples. The SERS detection capability was demonstrated by the dropwise addition of fluorescent carbonyl compounds before and after elution. Furthermore, the device was proved with high surface consistency(<70%) for repeated measurement, which has the potential for ppb(parts per billion) level concentration of carbonyl compounds detection.

Study on a High Performance MEMS Infrared Thermopile Detector.

This paper presents a high-performance micro-electromechanical systems (MEMS) thermopile infrared detector. It consists of a double-end beam and a dual-layer thermocouple structure, which improves the responsivity of the detector. The etch-stop structure is integrated into the detector to prevent isotropic etching-caused damage on the device. The responsivity of the detector achieved 1151.14 V/W, and the measured response time was 14.46 ms. The detector had the potential to work as a high-precision temperature sensor and as a vacuum sensor.

Fabrication and characterization of SiO2/Si heterogeneous nanopillar arrays.

This work presents arrays of heterogeneous nanopillars stacked with Si bodies and SiO2 heads for biomedical applications. Novel crossed and overlapped spacer techniques are proposed to fabricate the nanopillar arrays in controllable dimensions. For the nanopillars in the arrays, the minimum spacing, body diameter and head tip-radius reach 100 nm, 23 nm and 11 nm, respectively. The maximum height is 1.2 µm. In addition, because of hydrophilic/hydrophobic selectivity between the SiO2 heads and Si bodies, localized nanoliter water-droplet condensing, fluorescein solution extraction and protein capturing are observed on the SiO2 pillar heads. These experiments demonstrate the great potential of heterogeneous nanopillars in biomedical applications.

Microfluidic surface-enhanced Raman scattering sensors based on nanopillar forests realized by an oxygen-plasma-stripping-of-photoresist technique.

A novel surface-enhanced Raman scattering (SERS) sensor is developed for real-time and highly repeatable detection of trace chemical and biological indicators. The sensor consists of a polydimethylsiloxane (PDMS) microchannel cap and a nanopillar forest-based open SERS-active substrate. The nanopillar forests are fabricated based on a new oxygen-plasma-stripping-of-photoresist technique. The enhancement factor (EF) of the SERS-active substrate reaches 6.06 × 10(6) , and the EF of the SERS sensor is about 4 times lower due to the influence of the PDMS cap. However, the sensor shows much higher measurement repeatability than the open substrate, and it reduces the sample preparation time from several hours to a few minutes, which makes the device more reliable and facile for trace chemical and biological analysis.