Raw Images Research Articles

GobletNet: Wavelet-Based High-Frequency Fusion Network for Semantic Segmentation of Electron Microscopy Images.

Semantic segmentation of electron microscopy (EM) images is crucial for nanoscale analysis. With the development of deep neural networks (DNNs), semantic segmentation of EM images has achieved remarkable success. However, current EM image segmentation models are usually extensions or adaptations of natural or biomedical models. They lack the full exploration and utilization of the intrinsic characteristics of EM images. Furthermore, they are often designed only for several specific segmentation objects and lack versatility. In this study, we quantitatively analyze the characteristics of EM images compared with those of natural and other biomedical images via the wavelet transform. To better utilize these characteristics, we design a high-frequency (HF) fusion network, GobletNet, which outperforms state-of-the-art models by a large margin in the semantic segmentation of EM images. We use the wavelet transform to generate HF images as extra inputs and use an extra encoding branch to extract HF information. Furthermore, we introduce a fusion-attention module (FAM) into GobletNet to facilitate better absorption and fusion of information from raw images and HF images. Extensive benchmarking on seven public EM datasets (EPFL, CREMI, SNEMI3D, UroCell, MitoEM, Nanowire and BetaSeg) demonstrates the effectiveness of our model. The code is available at https://github.com/Yanfeng-Zhou/GobletNet.

Reinforcement Learning with Decoupled State Representation for Robot Manipulations

Deep reinforcement learning has significantly advanced robot manipulations by providing an alternative solution for designing control strategies using raw images as direct inputs. While images offer additional environmental information, the end-to-end policy training manner (from image to action) requires simultaneous representation and task learning by the agent. This often necessitates a substantial number of interaction samples to achieve satisfactory policy performance. Previous works has attempted to address this challenge by learning a visual representation model that encodes the entire image into a low-dimensional vector before the policy training. However, since this vector contains both robot and object information, it inevitably introduces coupling within the state, which can mislead the policy training process. In this study, a novel method called Reinforcement Learning with Decoupled State Representation is proposed to effectively decouple robot and object information within the state representation. Experimental results demonstrate that the proposed method exhibits faster learning speed and achieves superior performance compared to previous methods across various robot manipulation tasks. Moreover, with only 3096 offline images, the proposed method successfully applies to real-world robot pushing tasks, which demonstrates its high practicability.

Fourier Ptychographic Coherent Anti-Stokes Raman Scattering Microscopy with Point-Scanning for Super-Resolution Imaging.

Fourier ptychography (FP) is a high resolution wide-field imaging method based on the extended aperture in the Fourier space, which is synthesized from raw images with varying illumination angles. If FP is extended to coherent nonlinear optical imaging, the resolution could be further improved due to the increase of the cutoff frequency of the synthesized coherent optical transfer function (C-OTF) with respect to the order of nonlinear optical processes. However, there is a fundamental conflict between wide-field FP and nonlinear optical imaging, whereby the nonlinear optical imaging typically requires a focused excitation laser beam with high power density. To tackle the problem, in this work, a unique point-scanning FP (PS-FP) method is presented for super-resolution nonlinear optical imaging, in which the nonlinear optical signal is obtained by using focused laser beam, while the conventional FP algorithm can still be used to retrieve the super-resolution image. PS-FP coherent anti-Stokes Raman scattering (PS-FP-CARS) imaging on a variety of samples, where a 1.8-fold expansion of the OTF is achieved experimentally for enhancing vibrational imaging. Further theoretical calculation shows that the C-OTF of PS-FP higher-order CARS (PS-FP-HO-CARS) can be expanded up to ≈4.9-fold, thereby improving the spatial resolution by ≈3-fold in comparison with conventional point-scanning CARS with under tightly focused beams. The generality of PS-FP method developed in this work can be adapted to other coherent nonlinear optical imaging modalities for super-resolution imaging in tissue and cells.

Android-Based App Guava Leaf Diseases Identification using Convolution Neural Network

Guava is an economically significant fruit crop in many regions. Conventional plant disease identification depends on skilled workers visually inspecting samples, which can be laborious and error-prone. Therefore, the goal of this study is to create an Android app that uses InceptionV3 to precisely determine the health of guava plants, with an emphasis on hyperparameter tuning and ideal model selection. Inception V3 was chosen as the base model for this study as it outperformed the 25 pre-trained models from Keras API. The Guava leaves dataset consists of 120 raw images collected from real-time capture and the Google website. The images belong to four categories: Algal Spot, Red Rust, Whitefly, and Healthy leaf. The dataset was split into 60% for training and 40% for validation. To increase the number of images, augmentation was performed to multiply each training by 5 times. The 2048 features were extracted from the second last layer of Inception V3. These features were saved in the Numpy format to facilitate hyperparameter tuning. The best hyperparameter tuning results were achieved with a batch size of 4, a learning rate of 0.001, and the Adamax optimizer for 50 epochs, resulting in a validation accuracy of 0.99. This set of hyperparameters was used to train the model, yielding in a training accuracy of 0.98 and a validation accuracy of 0.9. The trained model was exported in the tflite file format and integrated into an Android app developed using Android Studio. To evaluate the performance evaluation of the app, 15 unseen images per class were used and the app achieved an overall F1-Score of 0.85, a recall of 0.85, a precision of 0.89, and a test accuracy of 0.85.

B-146 Unlocking the potential of large-cohort proteomics studies with the Orbitrap Astral mass platform

Abstract Background Large-cohort proteomics analysis using mass spectrometry is a powerful approach to discover and validate new biomarkers. In combination with clinical data and computational analysis, large-cohort proteomics study brings opportunities in improving early diagnosis, refining patient stratification, and predicting/monitoring treatment response. Yet, to achieve meaningful biological insights in large-cohort studies, robust, reproducible, and comprehensive proteome profiling in a high-throughput manner still remains challenging. Here, we use a high-resolution accurate mass (HRAM) Oribtrap Astral platform to enable high-quality and robust protein quantification across thousands of LC-MS/MS analyses. With a throughput of 100 samples/day, we can reproducibly profile ∼9000 proteins from human cell line and ∼800 proteins from undepleted plasma across multiple instruments and more than 10 consecutive days in a 24/7 operation mode. Methods Experiments were performed on multiple Orbitrap Astral mass spectrometers with and without FAIMS Pro interface. Chromatographic separations were performed using a trap/elute method on Vanquish Neo system at a 11-minute gradient and a flow rate of 1ul/min, resulting in a throughput of 100 samples/day. Undepleted plasma digest was analyzed in a 24/7 mode for > 1000 injections on each LC-MS/MS setup. In addition, HeLa digest as quality control (QC) was analyzed every 12hours. The Oribtrap Astral platform was operated in data-independent acquisition (DIA) mode using a full scan with m/z 380-980, and DIA MS/MS scans with an isolation window of 2Th. Resulting DIA raw files were automatically transferred and processed using Chimerys in a beta version of Proteome Discoverer software. Results Multiple LC-MS/MS systems were operated in DIA mode either with or without FAIMS Pro device in a 24/7 operation mode at a throughput of 100SPD. Undepleted plasma digest was analyzed with >1000 injections on each LC-MS/MS setup. To monitor the baseline performance, HeLa digest served as QC were analyzed periodically every 12 hours with 3 technical replicates each time. To effectively manage and analyze these thousands of data files generated, we developed an automated data transfer and analysis pipeline. Automatically, the resulting DIA raw data files were immediately transferred to a server, then processed by state-of-the-art intelligent search algorithm Chimerys. Benefiting from the ultra-high scan speed of up to 200Hz on the Orbitrap Astral mass spectrometer, a much narrower isolation window width of 2Th was applied in the DIA method, comparing to 10-20Th on the classic DIA method. As a result, ∼9000 proteins from HeLa digest and ∼800 proteins from undepleted plasma digest were identified within only a 11-minutes gradient, respectively. More than 80% of the plasma proteins were reproducibly identified and quantified from all the runs on each LC-MS/MS setup, indicating a great reproducibility from run-to-run longitudinally. Stable and robust peptide quantitation was observed by extracting peptides with high, medium, and low abundant across the runs. Importantly, QC showed no performance degradation throughout the entire study, indicating high robustness of the entire LC-MS/MS setup. Conclusions Orbitrap Astral mass platform can comprehensively analyze the proteome of >1000s of sample robustly and reproducibly in a high-throughput manner, addressing the needs in large-cohort studies.

Optimization and correction of breast dynamic optical imaging projection data based on deep learning

Breast cancer poses a significant health threat to women, necessitating advancements in diagnostic technologies. Breast dynamic optical imaging (DOI) technology, recognized for its non-invasive and radiation-free properties, is extensively utilized for the early screening and quantitative analysis of breast tumors. The integration of deep learning, a robust technology for automatic image feature extraction, with breast DOI has the potential to enhance tumor detection and diagnosis significantly. This paper introduces a deep learning-enhanced image optimization approach to overcome challenges such as poor image quality and distorted projection data commonly encountered in existing DOI methods. The approach utilizes convolutional neural networks (CNNs) to extract features from raw images and employs generative adversarial networks (GANs) to enhance these images, thereby improving their quality and contrast. Additionally, a novel correction algorithm is developed to address projection data distortion, enabling the reconstruction and correction of this data for more accurate and reliable imaging results. Experimental findings confirm that the proposed method markedly enhances both image quality and projection data accuracy in breast DOI, offering a reliable foundation for clinical diagnosis. This study not only provides a new perspective and methodology for the early screening and diagnosis of breast cancer but also holds substantial clinical importance and prospective applications.

Automated multiple sclerosis progression rate computation of a patient from 2D FLAIR images with Rayleigh-Weibull-Fuzzy imaging and augmented morphing method

Multiple sclerosis (MS) is a neurological demyelinating disorder affecting brain and spinal cord by attacking the myelin sheaths of nerves. Estimation of the volumetric changes in MS lesions is a challenging and specialized task which is executed and judged by medical experts. The change in the volume of the lesions provides crucial information on MS progression or regression by comparing the magnetic resonance images (MRI) taken in successive scans. However, visual comparison of the images, even with an expert eye, would not always lead to a conclusive decision nor a consensus on progression or regression. Therefore, we present an automated expert system for estimating MS progression rate by automatic lesion segmentation and volume estimation using two-dimensional MRIs, which is also adaptable to various parameters, slice thickness and increment. A clinical dataset is specially formed for this research which contains three sets of 135 MR images of an MS patient generated within approximately 23- and 6-month periods consecutively with identical device parameters. The lesions are segmented by a novel Rayleigh-Weibull-Fuzzy (RWF) imaging method based on the Nakagami distribution and specialized fuzzy 2-means. Subsequently, the segmentation module is trained to fit the ground truths images created by experts to achieve the highest dice score possible for a total number of 56 images containing lesions, which is found as 93.76%. Afterwards, several imaginary image sequences are generated by augmented linear and nonlinear morphing for re-segmentation of imaginary lesions by RWF. Finally, we estimated the volumetric change between the first two MRI sequences to adjust the morphing module and to predict the progression rate of the lesions in time. The framework automatically selected the highest accuracy, which is 99.9% in the training session and estimated the progression rate in the testing phase with 99.69% accuracy, which are not achievable without augmented morphing methodology. For the first time in the literature, an automated framework could estimate the MS progression rate from the raw MR images, which is also the main innovation of this paper and the outputs would be beneficial for the experts working on this field.

Complex-valued image reconstruction for compressed sensing MRI using hierarchical constraint

In Magnetic Resonance Imaging (MRI), the sequential acquisition of raw complex-valued image data in Fourier space, also known as k-space, results in extended examination times. To speed up the MRI scans, k-space data are usually undersampled and processed using numerical techniques such as compressed sensing (CS). While the majority of CS-MRI algorithms primarily focus on magnitude images due to their significant diagnostic value, the phase components of complex-valued MRI images also hold substantial importance for clinical diagnosis, including neurodegenerative diseases. In this work, complex-valued MRI reconstruction is studied with a focus on the simultaneous reconstruction of both magnitude and phase images. The proposed algorithm is based on the nonsubsampled contourlet transform (NSCT) technique, which offers shift invariance in images. Instead of directly transforming the complex-valued image into the NSCT domain, we introduce a wavelet transform within the NSCT domain, reducing the size of the sparsity of coefficients. This two-level hierarchical constraint (HC) enforces sparse representation of complex-valued images for CS-MRI implementation. The proposed HC is seamlessly integrated into a proximal algorithm simultaneously. Additionally, to effectively minimize the artifacts caused by sub-sampling, thresholds related to different sub-bands in the HC are applied through an alternating optimization process. Experimental results show that the novel method outperforms existing CS-MRI techniques in phase-regularized complex-valued image reconstructions.

Intelligent fatigue damage tracking and prognostics of composite structures utilizing raw images via interpretable deep learning

In recent years, prognostics gained attention in various industries by optimizing maintenance, boosting operational efficiency, and preventing costly downtime. Central to prognostics is the Remaining Useful Life (RUL), representing the critical time before system failure. Deep learning advancements facilitate RUL forecasting by extracting features from diverse data formats such as time series, images, or sequences thereof, in one, two, or three dimensions, respectively. Yet, predicting RUL from image sequences often relies heavily on resource-intensive techniques like digital image correlation, complicating data acquisition. To address challenges with high-dimensional data and unreliable models, this study introduces ISTRUST, an innovative Transformer-based architecture. ISTRUST (Interpretable Spatiotemporal TRansformer for Understanding STructures) tackles the dual challenges posed by high-dimensional data and the black-box nature of existing models. Leveraging Transformers’ attention mechanism, ISTRUST breaks down the spatiotemporal domain, effectively realizing interpretable RUL predictions under uncertainty using only sparse raw image sequences as input. Evaluated on fatigue-loaded composite samples showcasing crack propagation, ISTRUST interprets the relation between cracks and RUL via the attention mechanism. The results substantiate its capacity to interpret and clarify instances in which predictions may exhibit variability in accuracy. Through the attention mechanism, a strong correlation between the model’s spatiotemporal focus and the RUL predictions is established, making it, to the best of our knowledge, the first model to provide interpretable stochastic RUL predictions directly from sequential images of this nature.

Coseismic surface ruptures of 20 strike-slip earthquakes measured from geodetic imaging data

Observations of coseismic surface ruptures are widely used for examining faulting mechanics and kinematics, and in fault displacement hazard analysis. Here, first, we provide new remote sensing observations of 20 historic surface rupturing continental strike-slip earthquakes (6.1 ≤ Mw≤ 7.9). To measure the near-field surface deformation pattern, we have processed a range of raw radar and optical images from satellite and aerial platforms with pixel tracking techniques, and in two cases using standard interferometric synthetic aperture radar (InSAR). We provide a range of observations, including surface displacement maps that constrain at least the horizontal component of surface motion (2D), and for three events the full 3D components. Second, we provide strain invariants for each event, which includes the finite dilatational strain, strain magnitudes, the vorticity, and maximum shear strain. Third, we provide a total of 134 surface fault rupture traces that are mapped manually from the displacement maps. Finally, we provide 2648 measurements of coseismic horizontal fault-parallel surface fault slip that are measured objectively using swath profiles with a function fitting method that we have developed.

An international film dosimetry intercomparison to establish a multi-center audit framework.

In 2021, a Technical Meeting was hosted by the International Atomic Energy Agency (IAEA) where it was recommended that a standardized method for assessing the accuracy of film dose calculations should be established. To design an audit that evaluates the accuracy of film dosimetry processes. To propose a framework for identifying out-of-tolerance results and to perform an international pilot study to test the audit design. Six members of an international Dosimetry Audit Network (DAN) developed an audit for radiochromic film dosimetry. A single host center provided the materials to each participating DAN member to conduct the audits. Materials included: (1) a set of two irradiated audit films (10 Sq: 10cm × 10cm, 15 Sq: 15cm × 15cm), (2) a reference calibration film set, and (3) a blank sheet of film. The participants were blinded to the dose and tasked with producing dose maps using their standard film dosimetry process. Average Region-Of-Interest (ROI: 2cm × 2cm) dose was measured from the dose maps and compared to the known dose. In the audit, all participants used their local scanning and software protocols. Film calibration was performed in two distinct ways: (1) using a calibration film set which was provided by the host center and (2) using a calibration film set which was locally irradiated. Several variations of the audit were also performed to examine how scanning and software processing can affect film dosimetry results. In the first variation of the audit (VariantA), a set of film scans was processed using five different software solutions. In the second variation of the audit (VariantB), all films were scanned on the same scanner and processed using two in-house software solutions. Taking one film scan from each participant, the standard deviations (1σ) (SD) in the dose returned from the host calibration and returned from the local calibration were±7.2% and±6.5% respectively, with variations from -12.4% to 12.9% of the known dose. The larger dose variations in the data set were attributed to the corrections applied for variations in scanner brightness during processing and incorrectly assigned calibration doses. When the raw image data set was processed by an expert user of each software solution (VariantA) the SDs were±2.7% and±3.7% for in-house and vendor-based software, respectively. When the films were scanned on a single scanner and processed with the two in-house software solutions (VariantB) the results had a SD of±2.3%. An audit has been designed and tested for radiotherapy film dosimetry at an international level. A framework for diagnosing issues within a film dosimetry process has been proposed that could be used to audit centers that use film as a dosimeter. Incorporating quality assurance throughout the film process is important in obtaining accurate and consistent film dosimetry. A better understanding of vendor-based software systems is necessary for users to process accurate and consistent film dosimetry.

A novel low-quality Raman hyperspectral image reconstruction method for corn component mapping

Raman imaging used to obtain spatial distribution information of chemical components on the sample surface has been widely applied in the nutritional assessment and quality control of cereal grains. However, obtaining high-quality Raman hyperspectral images of cereal grain based on point scanning is time-consuming due to the limitations of the instrumental imaging principle and Raman hyperspectral images acquired in a short time are low-quality with noise interference, limiting its practical application in the visualization of cereal grain composition. To rapidly and effectively achieve high-quality component distribution mappings of cereal grains based on Raman hyperspectral images, we propose a novel Swin Transformer-based method combining spatial-spectral denoising and image reconstruction (SwinSSID-IR) for low-quality Raman hyperspectral image reconstruction. The novel denoising network named SwinSSID innovatively introduces Swin Transformer blocks (STB) to effectively eliminate noise from raw low-quality Raman hyperspectral images of corn kernels. Next, the super-resolution network named SwinIR rapidly generates high-quality component distribution mappings of corn with exceptional clarity and detail from the denoised low-quality Raman hyperspectral image at characteristic peak. Experimental results show the competitiveness of the proposed method compared with other similar advanced methods and its effective application in non-destructive visual analysis of corn components.

Effective removal of global tilt from atomically-resolved topography images of vicinal surfaces with narrow terraces

The main feature of vicinal surfaces of crystals characterized by the Miller indices (hhm) is rather small width (less than 10 nm) and substantially large length (more than 200 nm) of atomically-flat terraces. This makes difficult to apply standard methods of image processing and correct visualization of crystalline lattices at the terraces and multiatomic steps. Here we consider two procedures allowing us to minimize effects of both small-scale noise and global tilt of sample: (i) analysis of the difference of two Gaussian blurred images, and (ii) subtraction of the plane, whose parameters are determined by optimization of the histogram of the visible heights, from raw topography image. It is shown that both methods provide nondistorted images demonstrating atomic structures on vicinal Si(556) and Si(557) surfaces.

The Effect of Monitored Walking on Extracardiac Intestinal Activity in Myocardial Perfusion Imaging.

Various techniques have been used in attempts to reduce interfering gastrointestinal activity in myocardial perfusion imaging (MPI); however, these approaches have yielded inconsistent results. The goal of this study was to investigate the efficacy of monitored walking, a previously unexplored technique, in reducing subdiaphragmatic activity-related artifacts during pharmacologic stress 99mTc-tetrofosmin MPI with SPECT to improve the overall image quality. Methods: The study included patients who underwent MPI with pharmacologic stress. They were given a step counter immediately after the radiotracer injection and were randomized into a group A, with a request to walk at least 1,000 steps before imaging, and a group B, with no specific instructions about walking. The reconstructed SPECT images were assessed visually. Moderate and severe levels of subdiaphragmatic tracer activity were considered relevant for the interpretation of the scans. Additionally, myocardial and abdominal activity was semiquantitatively assessed on raw planar images, and the mean myocardium-to-abdomen count ratios were calculated. Results: We enrolled 199 patients (95 patients in group A and 104 patients in group B). Clinical characteristics did not differ significantly between the 2 groups. Patients in group A walked more steps than patients in group B (P < 0.001), but there were no differences in the proportion of accepted scans between the 2 groups (P = 0.41). Additionally, there were no differences in the proportion of relevant subdiaphragmatic activity between the groups (P = 0.91). The number of steps did not impact the acceptance rate (P = 0.29). Conclusion: A higher number of steps walked during the waiting period between pharmacologic stress and acquisition does not affect subdiaphragmatic activity-related artifacts or the proportion of accepted scans after pharmacologic stress. However, pedometer use and clear instructions motivate patients to walk while awaiting imaging. Larger studies are required to compare a higher-step-count group with a sedentary control group to assess the influence of walking on gastrointestinal artifacts in MPI.

Leveraging Perspective Transformation for Enhanced Pothole Detection in Autonomous Vehicles.

Road conditions, often degraded by insufficient maintenance or adverse weather, significantly contribute to accidents, exacerbated by the limited human reaction time to sudden hazards like potholes. Early detection of distant potholes is crucial for timely corrective actions, such as reducing speed or avoiding obstacles, to mitigate vehicle damage and accidents. This paper introduces a novel approach that utilizes perspective transformation to enhance pothole detection at different distances, focusing particularly on distant potholes. Perspective transformation improves the visibility and clarity of potholes by virtually bringing them closer and enlarging their features, which is particularly beneficial given the fixed-size input requirement of object detection networks, typically significantly smaller than the raw image resolutions captured by cameras. Our method automatically identifies the region of interest (ROI)-the road area-and calculates the corner points to generate a perspective transformation matrix. This matrix is applied to all images and corresponding bounding box labels, enhancing the representation of potholes in the dataset. This approach significantly boosts detection performance when used with YOLOv5-small, achieving a 43% improvement in the average precision (AP) metric at intersection-over-union thresholds of 0.5 to 0.95 for single class evaluation, and notable improvements of 34%, 63%, and 194% for near, medium, and far potholes, respectively, after categorizing them based on their distance. To the best of our knowledge, this work is the first to employ perspective transformation specifically for enhancing the detection of distant potholes.

Enhancing visual reinforcement learning with State–Action Representation

Despite the remarkable progress made in visual reinforcement learning (RL) in recent years, sample inefficiency remains a major challenge. Many existing approaches attempt to address this by extracting better representations from raw images using techniques like data augmentation or introducing some auxiliary tasks. However, these methods overlook the environmental dynamic information embedded in the collected transitions, which can be crucial for efficient control. In this paper, we present STAR: State-Action Representation Learning, a simple yet effective approach for visual continuous control. STAR learns a joint state–action representation by modeling the dynamics of the environment in the latent space. By incorporating the learned joint state–action representation into the critic, STAR enhances the value estimation with latent dynamics information. We theoretically show that the value function can still converge to the optima when involving additional representation inputs. On various challenging visual continuous control tasks from DeepMind Control Suite, STAR achieves significant improvements in sample efficiency compared to strong baseline algorithms.

Multi-view neural 3D reconstruction of micro- and nanostructures with atomic force microscopy

Atomic Force Microscopy (AFM) is a widely employed tool for micro- and nanoscale topographic imaging. However, conventional AFM scanning struggles to reconstruct complex 3D micro- and nanostructures precisely due to limitations such as incomplete sample topography capturing and tip-sample convolution artifacts. Here, we propose a multi-view neural-network-based framework with AFM, named MVN-AFM, which accurately reconstructs surface models of intricate micro- and nanostructures. Unlike previous 3D-AFM approaches, MVN-AFM does not depend on any specially shaped probes or costly modifications to the AFM system. To achieve this, MVN-AFM employs an iterative method to align multi-view data and eliminate AFM artifacts simultaneously. Furthermore, we apply the neural implicit surface reconstruction technique in nanotechnology and achieve improved results. Additional extensive experiments show that MVN-AFM effectively eliminates artifacts present in raw AFM images and reconstructs various micro- and nanostructures, including complex geometrical microstructures printed via two-photon lithography and nanoparticles such as poly(methyl methacrylate) (PMMA) nanospheres and zeolitic imidazolate framework-67 (ZIF-67) nanocrystals. This work presents a cost-effective tool for micro- and nanoscale 3D analysis.

Twenty Years of Thermal Infrared Observations (2004–2024) at Campi Flegrei Caldera (Italy) by the Permanent Surveillance Ground Network of INGV-Osservatorio Vesuviano

Thermal infrared (TIR) time series images acquired by ground, proximal TIR stations provide valuable data to study evolution of surface temperature fields of diffuse degassing volcanic areas. This paper presents data processing results related to TIR images acquired since 2004 by six ground stations in the permanent thermal infrared surveillance network at Campi Flegrei (TIRNet) set up by INGV-Osservatorio Vesuviano. These results are reported as surface temperature and heat flux time series. The processing methodologies, also discussed in this paper, allow for presentation of the raw TIR image data in a more comprehensible form, suitable for comparisons with other geophysical parameters. A preliminary comparison between different trends in the surface temperature and heat flux values recorded by the TIRNet stations provides evidence of peculiar changes corresponding to periods of intense seismicity at the Campi Flegrei caldera. During periods characterized by modest seismicity, no remarkable evidence of common temperature variations was recorded by the different TIRNet stations. Conversely, almost all the TIRNet stations exhibited common temperature variations, even on a small scale, during periods of significant seismic activity. The comparison between the seismicity and the variations in the surface temperature and heat flux trends suggests an increase in efficiency of heat transfer between the magmatic system and the surface when an increase in seismic activity was registered. This evidence recommends a deeper, multidisciplinary study of this correlation to improve understanding of the volcanic processes affecting the Campi Flegrei caldera.

Lens-free on-chip 3D microscopy based on wavelength-scanning Fourier ptychographic diffraction tomography

Lens-free on-chip microscopy is a powerful and promising high-throughput computational microscopy technique due to its unique advantage of creating high-resolution images across the full field-of-view (FOV) of the imaging sensor. Nevertheless, most current lens-free microscopy methods have been designed for imaging only two-dimensional thin samples. Lens-free on-chip tomography (LFOCT) with a uniform resolution across the entire FOV and at a subpixel level remains a critical challenge. In this paper, we demonstrated a new LFOCT technique and associated imaging platform based on wavelength scanning Fourier ptychographic diffraction tomography (wsFPDT). Instead of using angularly-variable illuminations, in wsFPDT, the sample is illuminated by on-axis wavelength-variable illuminations, ranging from 430 to 1200 nm. The corresponding under-sampled diffraction patterns are recorded, and then an iterative ptychographic reconstruction procedure is applied to fill the spectrum of the three-dimensional (3D) scattering potential to recover the sample’s 3D refractive index (RI) distribution. The wavelength-scanning scheme not only eliminates the need for mechanical motion during image acquisition and precise registration of the raw images but secures a quasi-uniform, pixel-super-resolved imaging resolution across the entire imaging FOV. With wsFPDT, we demonstrate the high-throughput, billion-voxel 3D tomographic imaging results with a half-pitch lateral resolution of 775 nm and an axial resolution of 5.43 μm across a large FOV of 29.85 mm2 and an imaging depth of >200 μm. The effectiveness of the proposed method was demonstrated by imaging various types of samples, including micro-polystyrene beads, diatoms, and mouse mononuclear macrophage cells. The unique capability to reveal quantitative morphological properties, such as area, volume, and sphericity index of single cell over large cell populations makes wsFPDT a powerful quantitative and label-free tool for high-throughput biological applications.

Local-Peak Scale-Invariant Feature Transform for Fast and Random Image Stitching.

Image stitching aims to construct a wide field of view with high spatial resolution, which cannot be achieved in a single exposure. Typically, conventional image stitching techniques, other than deep learning, require complex computation and are thus computationally expensive, especially for stitching large raw images. In this study, inspired by the multiscale feature of fluid turbulence, we developed a fast feature point detection algorithm named local-peak scale-invariant feature transform (LP-SIFT), based on the multiscale local peaks and scale-invariant feature transform method. By combining LP-SIFT and RANSAC in image stitching, the stitching speed can be improved by orders compared with the original SIFT method. Benefiting from the adjustable size of the interrogation window, the LP-SIFT algorithm demonstrates comparable or even less stitching time than the other commonly used algorithms, while achieving comparable or even better stitching results. Nine large images (over 2600 × 1600 pixels), arranged randomly without prior knowledge, can be stitched within 158.94 s. The algorithm is highly practical for applications requiring a wide field of view in diverse application scenes, e.g., terrain mapping, biological analysis, and even criminal investigation.