High-resolution Imaging Research Articles

Mass wasting reveals ongoing asymmetric retreat of the martian north polar ice cap

Ongoing mass wasting through ice block falls is intensive at the north polar ice cap of Mars. We monitored how this activity is currently shaping the marginal steep scarps of the ice cap, which holds a record of the planet’s climate history. With AI-driven change detection between multi-temporal high-resolution satellite images, we created a comprehensive map of mass wasting across the entire North Polar Layered Deposits (NPLD). Our results show a more active erosion process than previously thought, with scarps retreating by up to ~3 m every kiloyear. The distribution of the active scarps indicates an ongoing asymmetric retreat of the already subcircular ice cap. The active scarps and the interior dune fields correlate strongly with exposures of the underlying, sandier Basal Unit (BU), providing evidence that erosion of the BU undermines the base of the NPLD. Moreover, ice block fall activity suggests potential areas where gypsum is released, given that the interior gypsum-bearing dune fields are located adjacent to these active scarps. Here, our study reveals the rates of present-day topographic change of the north polar ice cap, providing a valuable constraint for study of its past evolution.

The clinical-stage drug BTZ-043 accumulates in murine tuberculosis lesions and efficiently acts against Mycobacterium tuberculosis

The development of granulomas with central necrosis harboring Mycobacterium tuberculosis (Mtb) is the hallmark of human tuberculosis (TB). New anti-TB therapies need to effectively penetrate the cellular and necrotic compartments of these lesions and reach sufficient concentrations to eliminate Mtb. BTZ-043 is a novel antibiotic showing good bactericidal activity in humans in a phase IIa trial. Here, we report on lesional BTZ-043 concentrations severalfold above the minimal-inhibitory-concentration and the substantial local efficacy of BTZ-043 in interleukin-13-overexpressing mice, which mimic human TB pathology of granuloma necrosis. High-resolution MALDI imaging further reveals that BTZ-043 diffuses and accumulates in the cellular compartment, and fully penetrates the necrotic center. This is the first study that visualizes an efficient penetration and accumulation of a clinical-stage TB drug in human-like centrally necrotizing granulomas and that also determines its lesional activity. Our results most likely predict a substantial bactericidal effect of BTZ-043 at these hard-to-reach sites in TB patients.

Machine learning-based assessment of morphometric abnormalities distinguishes bipolar disorder and major depressive disorder.

Bipolar disorder (BD) and major depressive disorder (MDD) have overlapping clinical presentations which may make it difficult for clinicians to distinguish them potentially resulting in misdiagnosis. This study combined structural MRI and machine learning techniques to determine whether regional morphological differences could distinguish patients with BD and MDD. A total of 123 participants, including BD (n = 31), MDD (n = 48), and healthy controls (HC, n = 44), underwent high-resolution 3D T1-weighted imaging. Cortical thickness, surface area, and subcortical volumes were measured using FreeSurfer software. Common and classic machine learning models were utilized to identify distinct morphometric alterations between BD and MDD. Significant morphological differences were observed in both common and distinct brain regions between BD, MDD, and HC. Specifically, abnormalities in the amygdala, thalamus, medial orbitofrontal cortex and fusiform were observed in both BD and MDD compared with HC. Relative to HC, unique differences in BD were identified in the lateral occipital and inferior/middle temporal regions, whereas MDD exhibited differences in nucleus accumbens and middle temporal regions. BD exhibited larger surface area in right middle temporal gyrus and greater right nucleus accumbens volume compared to MDD. The integration of two-stage models, including deep neural network (DNN) and support vector machine (SVM), achieved an accuracy rate of 91.2% in discriminating individuals with BD from MDD. These findings demonstrate that structural MRI combined with machine learning techniques can accurately discriminate individuals with BD from MDD, and provide a foundation supporting the potential of this approach to improve diagnostic accuracy.

Simultaneous multislice diffusion imaging using navigator-free multishot spiral acquisitions.

This work aims to raise a novel design for navigator-free multiband (MB) multishot uniform-density spiral (UDS) acquisition and reconstruction, and to demonstrate its utility for high-efficiency, high-resolution diffusion imaging. Our design focuses on the acquisition and reconstruction of navigator-free MB multishot UDS diffusion imaging. For acquisition, radiofrequency-pulse encoding was used to achieve controlled aliasing in parallel imaging in MB imaging. For reconstruction, a new algorithm named slice-projection onto convex sets-enhanced inherent correction of phase errors (slice-POCS-ICE) was proposed to simultaneously estimate diffusion-weighted images and intershot phase variations for each slice. The efficacy of the proposed methods was evaluated in both numerical simulation and in vivo experiments. In both numerical simulation and in vivo experiments, slice-POCS-ICE estimated phase variations more precisely and provided results with better image quality than other methods. The intershot phase variations and MB slice aliasing artifacts were simultaneously resolved using the proposed slice-POCS-ICE algorithm. The proposed navigator-free MB multishot UDS acquisition and reconstruction method is an effective solution for high-efficiency, high-resolution diffusion imaging.

Evaluation of Intensive Statins and Proprotein Convertase Subtilisin/Kexin Type 9 Inhibitors on Intracranial Artery Plaque Stability: A Prospective Single-Arm Study.

Intracranial atherosclerotic stenosis is a leading cause of ischemic stroke and recurrent events due to plaque instability. High-resolution magnetic resonance imaging identifies plaque enhancement as a key marker of instability. This study evaluated the efficacy of combined high-intensity statins and proprotein convertase subtilisin/kexin type 9 inhibitors in plaque stabilization. In this prospective, single-arm study, patients with acute stroke and intracranial atherosclerotic stroke in the M1 segment of the middle cerebral artery or basilar artery were enrolled. After 24 weeks of intensive statin and evolocumab therapy, high-resolution magnetic resonance imaging assessments of intracranial vessels were conducted at baseline and follow-up. The primary end point was the change in stenosis degree; secondary end points included changes in plaque enhancement (contrast ratio) and plaque burden. Thirty-six patients (median age, 58 years) participated. After therapy, stenosis decreased from 75.9% (interquartile range, 69.5%-84.8%) to 65.3% (interquartile range, 53.8%-75.0%; P<0.001). Contrast ratio grades and contrast volume decreased significantly (P<0.001), and lumen area increased (from 1.03 mm2 [interquartile range, 0.59-1.72 mm2] to 2.08 mm2 [interquartile range, 1.00-3.10 mm2]; P<0.001). No significant changes were observed in wall area, wall area index, or remodeling index. A notable correlation between the decrease in low-density lipoprotein cholesterol/apolipoprotein B ratio and reduction in contrast volume was observed. Combining proprotein convertase subtilisin/kexin type 9 inhibitors with high-intensity statins may improve plaque stability and reduce lumen stenosis in patients with intracranial atherosclerotic stenosis. Larger randomized controlled trials are needed to confirm these findings. URL: https://www.chictr.org.cn; Unique identifier: ChiCTR2200058029.

Automated 3D semantic segmentation of PCB X-ray CT images and netlist extraction

Printed Circuit Board (PCB) design reconstruction is essential for addressing part obsolescence, intellectual property recovery, compliance, quality assurance, and enhancing national capabilities. Traditional methods for PCB design extraction, both non-geometry-based and geometry-based, have limitations in accuracy, efficiency, and scalability. This paper presents an automated approach, combining image processing and machine learning, to achieve 3D semantic segmentation of PCB X-ray Computed Tomography (X-ray CT) images and subsequent netlist extraction. By employing a 3D U-Net architecture with a ResNet-18 backbone and training on synthetic data, we introduce a first-of-its-kind method for direct 3D semantic segmentation, significantly improving over previous efforts. Our approach eliminates the need for extensive labeled datasets by using inherently labeled synthetic data. Further, this method enhances ease of segmentation by significantly reducing or eliminating the preprocessing effort required for 2D image stacks. It also improves universality by expanding the scope of application beyond images with specific 2D stack criteria, segmenting the 3D image in its entirety. Additionally, this method enables the processing of images of PCBs that have undergone bending, which is common among PCBs with a thickness below a certain threshold. The implications of this approach extend beyond PCBs, finding applications in various physical and biological sciences where 3D image segmentation is crucial. This methodology includes high-resolution 3D imaging, watershed segmentation, machine learning-based semantic segmentation, and netlist extraction. Validation with both synthetic and real-world PCB datasets shows high accuracy and robustness, offering a scalable solution for PCB design reconstruction.

TagGen: Diffusion-based generative model for cardiac MR tagging super resolution.

The aim of the work is to develop a cascaded diffusion-based super-resolution model for low-resolution (LR) MR tagging acquisitions, which is integrated with parallel imaging to achieve highly accelerated MR tagging while enhancing the tag grid quality of low-resolution images. We introduced TagGen, a diffusion-based conditional generative model that uses low-resolution MR tagging images as guidance to generate corresponding high-resolution tagging images. The model was developed on 50 patients with long-axis-view, high-resolution tagging acquisitions. During training, we retrospectively synthesized LR tagging images using an undersampling rate (R) of 3.3 with truncated outer phase-encoding lines. During inference, we evaluated the performance of TagGen and compared it with REGAIN, a generative adversarial network-based super-resolution model that was previously applied to MR tagging. In addition, we prospectively acquired data from 6 subjects with three heartbeats per slice using 10-fold acceleration achieved by combining low-resolution R = 3.3 with GRAPPA-3 (generalized autocalibrating partially parallel acquisitions 3). For synthetic data (R = 3.3), TagGen outperformed REGAIN in terms of normalized root mean square error, peak signal-to-noise ratio, and structural similarity index (p < 0.05 for all). For prospectively 10-fold accelerated data, TagGen provided better tag grid quality, signal-to-noise ratio, and overall image quality than REGAIN, as scored by two (blinded) radiologists (p < 0.05 for all). We developed a diffusion-based generative super-resolution model for MR tagging images and demonstrated its potential to integrate with parallel imaging to reconstruct highly accelerated cine MR tagging images acquired in three heartbeats with enhanced tag grid quality.

Enhanced Segmentation Accuracy in High-Resolution Remote Sensing Images Using a Multi-Scale Convolutional Network

Enhanced Segmentation Accuracy in High-Resolution Remote Sensing Images Using a Multi-Scale Convolutional Network

Photoelectron Imaging and Photodetachment Spectroscopy for the Cryogenically Cooled Cyanocyclopentadienide Anion.

The cyano-cyclopentadiene molecule (CN-C5H5) has attracted significant interest since its detection in the interstellar medium, but the radical (CN-C5H4) and anionic (CN-C5H4-) forms of cyano-cyclopentadiene have not been studied. The cyano-cyclopentadienyl radical (CN-Cp) has a strong dipole moment, rendering it an ideal system for vibrational and rotational spectroscopy. We report an investigation of the cryogenically cooled cyano-cyclopentadienide anion (CN-Cp-) using high-resolution photoelectron imaging, photodetachment spectroscopy, and resonant photoelectron imaging. The electron affinity of the CN-Cp radical is measured accurately to be 2.7741 ± 0.0003 eV (22,375 ± 2 cm-1). A low-lying excited state is observed for the CN-Cp neutral radical at 151 cm-1 above the ground state. The overlap and vibronic coupling of the ground and low-lying electronic states give rise to complicated and congested photoelectron spectra. A dipole-bound state is observed for the CN-Cp- anion with a binding energy of 94 cm-1, along with 15 vibrational Feshbach resonances. Resonant photoelectron spectra via the vibrational resonances yield well-resolved spectra, allowing 26 vibronic levels to be identified for CN-Cp. The rich spectroscopic information will be valuable to compare with theoretical studies to unravel the vibronic coupling and nonadiabatic effects in the CN-Cp radical.

Protostellar Outflows at the EarliesT Stages (POETS) VI. Evidence of disk-wind in G11.92-0.61 MM1

Magnetohydrodynamic disk-winds are thought to play a key role in the formation of massive stars by providing the fine-tuning between accretion and ejection, where excess angular momentum is redirected away from the disk, allowing further mass growth of a young protostar. However, only a limited number of disk-wind sources have been detected to date. To better constrain the exact mechanism of this phenomenon, expanding the sample is critical. We performed a detailed analysis of the disk-wind candidate G11.92-0.61 MM1 by estimating the physical parameters of the massive protostellar system and constraining the wind-launching mechanism. Atacama Large Millimeter/submillimeter Array (ALMA) Band 6 observations of G11.92-0.61 MM1 were conducted in September 2021 with ALMA's longest baselines, which provided a synthesised beam of ∼30 mas. We obtained high-resolution images of the CH_3CN (varv_8=1 and varv=0), CH_3OH, SO_2, and SO molecular lines, as well as the 1.3 mm continuum. Our high-resolution molecular data allowed us to refine the parameters of the disk-outflow system in MM1. The rotating disk is resolved into two regions with distinct kinematics: the inner region ($&lt;$300 au) is traced by high-velocity emission of high-excitation CH_3CN lines and shows a Keplerian rotation; the outer region ($&gt;$300 au), traced by mid-velocity CH_3CN emission, rotates in a sub-Keplerian regime. The central source is estimated to be ∼20 M_⊙, which is about half the mass estimated in previous lower-resolution studies. A strong collimated outflow is traced by SO and SO_2 emission up to ∼3400 au around MM1a. The SO and SO_2 emissions show a rotation-dominated velocity pattern, a constant specific angular momentum, and a Keplerian profile that suggests a magneto-centrifugal disk-wind origin with launching radii of ∼50-100 au. G11.92-0.61 MM1 appears to be one of the clearest cases of molecular line-traced disk-winds detected around massive protostars.

Fast and high-resolution multimode fiber speckle imaging via optical coherence control

Fast and high-resolution multimode fiber speckle imaging via optical coherence control

Ultra-high-resolution brain MRI at 0.55T: bSTAR and its application to magnetization transfer ratio imaging.

This study aims to evaluate the feasibility of structural sub-millimeter isotropic brain MRI at 0.55 T using a 3D half-radial dual-echo balanced steady-state free precession sequence, termed bSTAR and to assess its potential for high-resolution magnetization transfer imaging. Phantom and in-vivo imaging of three healthy volunteers was performed on a low-field 0.55 T MR-system with isotropic bSTAR resolution settings of 0.87 × 0.87 × 0.87 mm3 and 0.69 × 0.69 × 0.69 mm3. Furthermore, off-resonance mapping was performed using 3D double-echo spoiled gradient imaging. For magnetization transfer (MT) MRI, the RF pulse duration of the 0.87 mm bSTAR scan was modified. Data were reconstructed using a GPU-accelerated compressed sensing algorithm. Magnetization transfer ratio (MTR) maps were calculated from two bSTAR scans with and without RF pulse prolongation. The MTR scan took 5 minutes and the reproducibility was assessed through repeated scans. Off-resonance mapping revealed that bSSFP brain imaging with TR < 5ms is essentially free of off-resonance-related artifacts even near the nasal cavities. Phantom and in-vivo scans demonstrated the feasibility of sub-millimeter isotropic bSTAR imaging. MTR maps obtained with high isotropic resolution bSTAR showed contrast between white and gray matter in agreement with expectations from high-field studies. The MTR measurements were highly reproducible with an average inter-scan MTR peak value of 43.3 ± 0.3 percent units. This study demonstrated the potential of sub-millimeter and artifact-free morphologic brain imaging at 0.55 T using bSTAR leveraging the advantages of low-field MRI, such as reduced susceptibility artifacts and improved radio-frequency field homogeneity. Furthermore, MT-sensitized bSTAR brain MRI enabled whole-brain MTR assessment within clinically feasible times and with high reproducibility.

Amygdala Nuclei Atrophy in Cognitive Impairment and Dementia: Insights from High-Resolution Magnetic Resonance Imaging

Background and Objectives: Cognitive impairment affects memory, reasoning, and problem-solving, with early detection being critical for effective management. The amygdala, a key structure in emotional processing and memory, may play a pivotal role in detecting cognitive decline. This study examines differences in amygdala nuclei volumes in patients with varying levels of cognitive performance to evaluate its potential as a biomarker. Material and methods: This cross-sectional study of 35 participants was conducted and classified into three groups: the normal (≥26), moderate (15–25), and low (≤14) cognitive performance groups based on the Montreal Cognitive Assessment (MoCA) scores. High-resolution magnetic resonance imaging at 3.0 T scanner was used to assess amygdala nuclei volumes. Results: Significant amygdala atrophy was observed in multiple amygdala nuclei across cognitive performance groups, with more pronounced changes in the low-performance group. The right hemisphere nuclei, including the lateral and basal nuclei, showed more significant differences, indicating their sensitivity to cognitive decline. Conclusions: This study highlights the potential of amygdala nuclei atrophy as a biomarker for cognitive impairment. Additional research with larger sample sizes and longitudinal designs is needed to confirm these findings and determine their diagnostic value.

Correlation of intracranial and extracranial carotid atherosclerotic plaque characteristics with ischemic stroke recurrence: a high-resolution vessel wall imaging study

ObjectivesTo evaluate the ability of the plaque characteristics of extracranial carotid and intracranial arteries to predict large atherosclerotic ischemic stroke recurrence via head and neck combined high-resolution vessel wall imaging (HR-VWI).MethodsThis prospective cohort study included 169 patients with large atherosclerotic ischemic stroke who underwent head and neck combined HR-VWI from April 2022 to May 2023. The baseline clinical data and atherosclerotic plaque characteristics of the intracranial and extracranial carotid arteries were collected, and the patients were followed up for 1 year, with the endpoint event defined as recurrent ischemic stroke. Clinical and imaging data were compared between the recurrent and nonrecurrent groups. Independent risk factors associated with stroke recurrence were assessed via multivariate Cox regression analysis. The receiver operating characteristic (ROC) curves of the relevant variables were also plotted, and the area under the curve (AUC) was calculated to assess their ability to predict stroke recurrence. Kaplan–Meier survival curves were used to compare the probability of stroke recurrence.ResultsDuring the 12-month follow-up, stroke recurrence occurred in 35 of the 169 patients. Multivariate Cox regression analysis revealed that the total number of intracranial and extracranial carotid plaques (p = 0.010) and coexisting extracranial carotid plaques and intracranial significantly enhanced plaques (p = 0.047) were independent risk factors for recurrent ischemic stroke. The AUCs for predicting stroke recurrence were 0.787 and 0.710, respectively. The Kaplan–Meier survival curve revealed that the risk of stroke recurrence was significantly greater in patients whose total number of intracranial and extracranial carotid plaques was &amp;gt;4.5 than in patients whose total number of plaques was &amp;lt;4.5 (p &amp;lt; 0.001) and was significantly greater in patients with coexisting extracranial carotid plaques and intracranial significantly enhanced plaques than in patients without coexisting plaques (p &amp;lt; 0.001).ConclusionA greater total number of intracranial and extracranial carotid plaques and the coexistence of extracranial carotid plaques and intracranially significantly enhanced plaques are independent risk factors associated with recurrent ischemic stroke. Head and neck combined HR-VWI may provide new indicators for the prediction of stroke recurrence, thus helping clinicians identify high-risk patients and target therapy to reduce the recurrence of ischemic events.

Augmented Reality Navigation System Enhances the Accuracy of Spinal Surgery Pedicle Screw Placement: A Randomized, Multicenter, Parallel-Controlled Clinical Trial.

The pedicle screw insertion technique has evolved significantly, and despite the challenges of precise placement, advancements like AR-based surgical navigation systems now offer enhanced accuracy and safety in spinal surgery by integrating real-time, high-resolution imaging with virtual models to aid surgeons. This study aims to evaluate the differences in accuracy between novel AR-guided pedicle screw insertion and conventional surgery techniques. A randomized controlled trial was conducted from March 2019 to December 2023 to compare the efficacy of AR-guided pedicle screw fixation with conventional freehand surgery using CT guidance. The study included 150 patients, aged 18-75, with 75 patients in each group. The total number of pedicle screws planned for the clinical trial placement was 351 and 348 in the experimental and control groups. The safety and efficacy of the procedures were evaluated by assessing screw placement accuracy and complication rates. In the full analysis set (FAS) analysis, the difference in the excellent and good rates of screw placement (experimental group - control group) and 95% confidence interval was 6.3% [3.0%-9.8%], with a p value of 0.0003 for the superiority test. In the FAS sensitivity analysis, the success rate was 98.0% (344 out of 351) in the experimental group and 91.7% (319 out of 348) in the control group, with a difference and 95% confidence interval of 6.3% [2.9% and 9.8%, respectively]. In the per-protocol set (PPS) analysis, the difference in the excellent and good rates of screw placement between the experimental and control groups, and the 95% confidence interval was 6.4% [3.3%-9.5%], with a p value of 0.0001 for the superiority test. In the actual treatment set (ATS) analysis, the excellent and good rates of screw placement were 99.1% in the experimental group and 91.7% in the control group. The difference in the excellent and good rates of screw placement (experimental group - control group) and 95% confidence interval was 7.3% [4.1%-10.6%], with a p value of < 0.0001 for the superiority test. The AR surgical navigation system can improve the accuracy of pedicle screw implantation and provide precise guidance for surgeons during pedicle screw insertion.

ADMNet: adaptive deformable convolution large model combining multi-level progressive fusion for Building Change Detection

ABSTRACT Building Change Detection (BCD) based on high-resolution Remote Sensing Images (RSI) simplifies urban surface monitoring. Nevertheless, the mainstream detection methods utilizing traditional convolution and attention mechanisms are often prone to errors due to the loss of edge detail information and underutilization of global context information. To address these issues, this paper presents a large model, namely ADMNet, which is built on adaptive deformable convolution and is designed to handles various types of building change information. First, we propose a Siamese neural network based on adaptive deformable convolution (ADC) modules. The ADC module incorporates spatial offset parameters into convolutional kernel sampling and mapping weights to capture irregularly varying edge features for local adaptive receptive fields. Second, we utilize a large model semantically driven to enhance model context awareness and construct long-range feature dependencies from multi-scale edge information, which are then integrated with locally adaptive edge structure features to achieve accurate edge localization. Furthermore, we design a Multi-Level Progressive Feature Fusion (MLPFF) module that enhances feature characterization capabilities to ensure internal integrity and improves model detection performance by integrating a priori knowledge from large-model transfer learning. To evaluate the effectiveness and generalizability of ADMNet, we conduct comparative experiments with current mainstream methods on two building datasets, LEVIR-CD and WHU-CD, and a land cover dataset, SYSU-CD. The results show that ADMNet outperforms all comparative methods. The source code is available at https://github.com/spaceYu180/ADMNet.

Elemental cryo-imaging reveals SOS1-dependent vacuolar sodium accumulation.

Increasing soil salinity causes significant crop losses globally; therefore, understanding plant responses to salt (sodium) stress is of high importance. Plants avoid sodium toxicity through subcellular compartmentation by intricate processes involving a high level of elemental interdependence. Current technologies to visualize sodium, in particular, together with other elements, are either indirect or lack in resolution. Here we used the newly developed cryo nanoscale secondary ion mass spectrometry ion microprobe1, which allows high-resolution elemental imaging of cryo-preserved samples and reveals the subcellular distributions of key macronutrients and micronutrients in root meristem cells of Arabidopsis and rice. We found an unexpected, concentration-dependent change in sodium distribution, switching from sodium accumulation in the cell walls at low external sodium concentrations to vacuolar accumulation at stressful concentrations. We conclude that, in root meristems, a key function of the NHX family sodium/proton antiporter SALT OVERLY SENSITIVE 1 (also known as Na+/H+ exchanger 7;SOS1/NHX7) is to sequester sodium into vacuoles, rather than extrusion of sodium into the extracellular space. This is corroborated by the use of new genomic, complementing fluorescently tagged SOS1 variants. We show that, in addition to the plasma membrane, SOS1 strongly accumulates at late endosome/prevacuoles as well as vacuoles, supporting a role of SOS1 in vacuolar sodium sequestration.

DualTransAttNet: A Hybrid Model with a Dual Attention Mechanism for Corn Seed Classification

Varietal purity is a critical quality indicator for seeds, yet various production processes can lead to the mixing of seeds from different varieties. Consequently, seed variety classification is an essential step in seed production. Existing classification algorithms often suffer from limitations such as reliance on single information sources, constrained feature extraction capabilities, time consumption, low accuracy, and the potential to cause irreversible damage to seeds. To address these challenges, this paper proposes a fast and non-destructive classification method for corn seeds, named DualTransAttNet, based on multi-source image information and hybrid feature extraction. High-resolution hyperspectral images of various corn varieties were collected, and a sliding sampling approach was employed to capture feature information across all spectral bands, resulting in the construction of a hyperspectral dataset for corn seed classification. Hyperspectral and RGB image data were then integrated to complement one another’s information and mitigate the insufficient feature diversity caused by single-source data. The proposed method leverages the strengths of convolutional neural networks (CNNs) and transformers to extract both local and global features, effectively capturing spectral and image characteristics. The experimental results demonstrate that the DualTransAttNet model can achieve a compact size of only 1.758 MB and an inference time of 0.019 ms. Compared to typical machine learning and deep learning models, the proposed model exhibits superior performance with an overall accuracy, F1-score, and Kappa coefficient of 90.01%, 88.9%, and 88.4%, respectively. The model’s rapid inference capability and low parameter count make it an excellent technical solution for agricultural automation and intelligent systems, thereby enhancing the efficiency and profitability of agricultural production.

Non-line-of-sight virtual modulated range migration imaging based on super-resolution histograms.

High-resolution non-line-of-sight (NLOS) imaging under nanosecond time-resolution conditions is challenging in applications. We propose a novel NLOS imaging method consisting of deconvolution modified iterative back projection and virtual modulated range migration for low time-resolution system, obtaining super-resolution (SR) histogram signal and high-resolution NLOS images sequentially. The proposed method is applicable to both confocal and non-confocal configurations. Experimental results demonstrate that the method can obtain a 50 multiples SR histogram signal from 1 ns low resolution signals. Based on the SR histograms, our method can reconstruct higher resolution NLOS images with several times less data quantity than the related methods. These above advantages indicate that our method has significant application potential in NLOS systems with low time resolution.

Use of Multimodal Artificial Intelligence in Surgical Instrument Recognition

Accurate identification of surgical instruments is crucial for efficient workflows and patient safety within the operating room, particularly in preventing complications such as retained surgical instruments. Artificial Intelligence (AI) models have shown the potential to automate this process. This study evaluates the accuracy of publicly available Large Language Models (LLMs)—ChatGPT-4, ChatGPT-4o, and Gemini—and a specialized commercial mobile application, Surgical-Instrument Directory (SID 2.0), in identifying surgical instruments from images. The study utilized a dataset of 92 high-resolution images of 25 surgical instruments (retractors, forceps, scissors, and trocars) photographed from multiple angles. Model performance was evaluated using accuracy, weighted precision, recall, and F1 score. ChatGPT-4o exhibited the highest accuracy (89.1%) in categorizing instruments (e.g., scissors, forceps). SID 2.0 (77.2%) and ChatGPT-4 (76.1%) achieved comparable accuracy, while Gemini (44.6%) demonstrated lower accuracy in this task. For precise subtype identification of instrument names (like “Mayo scissors” or “Kelly forceps”), all models had low accuracy, with SID 2.0 having an accuracy of 39.1%, followed by ChatGPT-4o (33.69%). Subgroup analysis revealed ChatGPT-4 and 4o recognized trocars in all instances. Similarly, Gemini identified surgical scissors in all instances. In conclusion, publicly available LLMs can reliably identify surgical instruments at the category level, with ChatGPT-4o demonstrating an overall edge. However, precise subtype identification remains a challenge for all models. These findings highlight the potential of AI-driven solutions to enhance surgical-instrument management and underscore the need for further refinements to improve accuracy and support patient safety.