Journal of Imaging最新文献

This paper presents a novel hybrid approach to feature detection designed specifically for enhancing Feature-Based Image Registration (FBIR). Through an extensive evaluation involving state-of-the-art feature detectors such as BRISK, FAST, ORB, Harris, MinEigen, and MSER, the proposed hybrid detector demonstrates superior performance in terms of keypoint detection accuracy and computational efficiency. Three image acquisition methods (i.e., rotation, scene-to-model, and scaling transformations) are considered in the comparison. Applied across a diverse set of remote-sensing images, the proposed hybrid approach has shown marked improvements in match points and match rates, proving its effectiveness in handling varied and complex imaging conditions typical in satellite and aerial imagery. The experimental results have consistently indicated that the hybrid detector outperforms conventional methods, establishing it as a valuable tool for advanced image registration tasks.

Video summarization aims to select the most informative subset of frames in a video to facilitate efficient video browsing. Past efforts have invariantly involved training summarization models with annotated summaries or heuristic objectives. In this work, we reveal that features pre-trained on image-level tasks contain rich semantic information that can be readily leveraged to quantify frame-level importance for zero-shot video summarization. Leveraging pre-trained features and contrastive learning, we propose three metrics featuring a desirable keyframe: local dissimilarity, global consistency, and uniqueness. We show that the metrics can well-capture the diversity and representativeness of frames commonly used for the unsupervised generation of video summaries, demonstrating competitive or better performance compared to past methods when no training is needed. We further propose a contrastive learning-based pre-training strategy on unlabeled videos to enhance the quality of the proposed metrics and, thus, improve the evaluated performance on the public benchmarks TVSum and SumMe.

Irritable bowel syndrome (IBS) is a gastrointestinal functional disorder characterized by unclear underlying mechanisms. Several theories propose that hyperactivation of the hypothalamic-pituitary-adrenal (HPA) axis leads to elevated cortisol levels and increased sensitivity of gut wall receptors. Given the absence of prior literature on this topic, our study aimed to investigate the potential for diagnosing IBS based on morphological changes in the pituitary gland, specifically its volume and grayscale intensity. Additionally, we aimed to assess whether factors such as gender, age, and body mass index influence these parameters. This retrospective study involved 60 patients, examining the volume and grayscale characteristics of their pituitary glands in the presence of IBS. Our findings revealed a positive correlation between pituitary gland volume and IBS diagnosis, although no significant correlation was observed with grayscale intensity. Due to the limited existing research and the small sample size of our study, further investigation with a larger cohort is warranted to validate these results.

Identifying the effects of pain catastrophizing on movement patterns in people with chronic low back pain (CLBP) has important clinical implications for treatment approaches. Prior research has shown people with CLBP have decreased lumbar-hip ratios during trunk flexion movements, indicating a decrease in the contribution of lumbar flexion relative to hip flexion during trunk flexion. In this study, we aim to explore the relationship between pain catastrophizing and movement patterns during trunk flexion in a CLBP population. Participants with CLBP (N = 98, male = 59, age = 39.1 ± 13.0) completed a virtual reality standardized reaching task that necessitated a progressively larger amount of trunk flexion. Specifically, participants reached for four virtual targets to elicit 15°, 30°, 45°, and 60° trunk flexion in the mid-sagittal plane. Lumbar flexion was derived from the motion data. Self-report measures of numerical pain ratings, kinesiophobia, and pain catastrophizing were obtained. Pain catastrophizing leads to decreased lumbar flexion angles during forward reaching. This effect is greater in females than males.

Nutrition in early life has an impact on white matter (WM) development in preterm-born babies. Quantitative analysis of pixel brightness intensity (PBI) on cranial ultrasound (CUS) scans has shown a great potential in the evaluation of periventricular WM echogenicity in preterm newborns. We aimed to investigate the employment of this technique to objectively verify the effects of parenteral nutrition (PN) on periventricular WM damage in preterm infants. Prospective observational study including newborns with gestational age at birth ≤32 weeks and/or birth weight ≤1500 g who underwent CUS examination at term-equivalent age. The echogenicity of parieto-occipital periventricular WM relative to that of homolateral choroid plexus (RE_CP) was calculated on parasagittal scans by means of quantitative analysis of PBI. Its relationship with nutrient intake through enteral and parenteral routes in the first postnatal week was evaluated. The study included 42 neonates for analysis. We demonstrated that energy and protein intake administered through the parenteral route positively correlated with both right and left RE_CP values (parenteral energy intake vs. right RE_CP: r = 0.413, p = 0.007; parenteral energy intake vs. left RE_CP: r = 0.422, p = 0.005; parenteral amino acid intake vs. right RE_CP: r = 0.438, p = 0.004; parenteral amino acid intake vs. left RE_CP: r = 0.446, p = 0.003). Multivariate linear regression analysis confirmed these findings. Quantitative assessment of PBI could be considered a simple, risk-free, and repeatable method to investigate the effects of PN on WM development in preterm neonates.

Non-contrast magnetic resonance angiography (NC-MRA), including fresh blood imaging (FBI), is a suitable choice for evaluating patients with peripheral artery disease (PAD). We evaluated standard FBI (sFBI) and centric ky-kz FBI (cFBI) acquisitions, using 1D and 2D parallel imaging factors (PIFs) to assess the trade-off between scan time and image quality due to blurring. The bilateral legs of four volunteers (mean age 33 years, two females) were imaged in the coronal plane using a body array coil with a posterior spine coil. Two types of sFBI and cFBI sequences with 1D PIF factor 5 in the phase encode (PE) direction (in-plane) and 2D PIF 3 (PE) × 2 (slice encode (SE)) (in-plane, through-slice) were studied. Image quality was evaluated by a radiologist, the vessel's signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were measured, and major vessel width was measured on the coronal maximum intensity projection (MIP) and 80-degree MIP. Results showed significant time reductions from 184 to 206 s on average when using sFBI down to 98 to 162 s when using cFBI (p = 0.003). Similar SNRs (averaging 200 to 370 across all sequences and PIF) and CNRs (averaging 190 to 360) for all techniques (p > 0.08) were found. There was no significant difference in the image quality (averaging 4.0 to 4.5; p > 0.2) or vessel width (averaging 4.1 to 4.9 mm; p > 0.1) on coronal MIP due to sequence or PIF. However, vessel width measured using 80-degree MIP demonstrated a significantly wider vessel in cFBI (5.6 to 6.8 mm) compared to sFBI (4.5 to 4.7 mm) (p = 0.022), and in 1D (4.7 to 6.8 mm) compared to 2D (4.5 to 5.6 mm) (p < 0.05) PIF. This demonstrated a trade-off in T2 blurring between 1D and 2D PIF: 1D using a PIF of 5 shortened the acquisition window, resulting in sharper arterial blood vessels in coronal images but significant blur in the 80-degree MIP. Two-dimensional PIF for cFBI provided a good balance between shorter scan time (relative to sFBI) and good sharpness in both in- and through-plane, while no benefit of 2D PIF was seen for sFBI. In conclusion, this study demonstrated the usefulness of FBI-based techniques for peripheral artery imaging and underscored the need to strike a balance between scan time and image quality in different planes through the use of 2D parallel imaging.

In this study, a novel method is proposed to estimate surface-spectral reflectance from camera responses that combine model-based and training-based approaches. An imaging system is modeled using the spectral sensitivity functions of an RGB camera, spectral power distributions of multiple light sources, unknown surface-spectral reflectance, additive noise, and a gain parameter. The estimation procedure comprises two main stages: (1) selecting the local optimal reflectance dataset from a reflectance database and (2) determining the best estimate by applying a neural network to the local optimal dataset only. In stage (1), the camera responses are predicted for the respective reflectances in the database, and the optimal candidates are selected in the order of lowest prediction error. In stage (2), most reflectance training data are obtained by a convex linear combination of local optimal data using weighting coefficients based on random numbers. A feed-forward neural network with one hidden layer is used to map the observation space onto the spectral reflectance space. In addition, the reflectance estimation is repeated by generating multiple sets of random numbers, and the median of a set of estimated reflectances is determined as the final estimate of the reflectance. Experimental results show that the estimation accuracies exceed those of other methods.

In recent years, improvements in display image quality have made it easier to perceive rich object information, such as gloss and transparency, from images, known as shitsukan. Do the different display specifications in the world affect their appearance? Clarifying the effects of differences in pixel structure on shitsukan perception is necessary to realize shitsukan management for displays with different hardware structures, which has not been fully clarified before. In this study, we experimentally investigated the effects of display pixel arrays on the perception of glossiness and transparency. In a visual evaluation experiment, we investigated the effects of three types of sub-pixel arrays (RGB, RGBW, and PenTile) on the perception of glossiness and transparency using natural images. The results confirmed that sub-pixel arrays affect the appearance of glossiness and transparency. A general relationship of RGB > PenTile > RGBW for glossiness and RGB > RGBW > PenTile for transparency was found; however, detailed analysis, such as cluster analysis, confirmed that the relative superiority of these sub-pixel arrays may vary depending on the observer and image content.

The utilization of robust, pre-trained foundation models enables simple adaptation to specific ongoing tasks. In particular, the recently developed Segment Anything Model (SAM) has demonstrated impressive results in the context of semantic segmentation. Recognizing that data collection is generally time-consuming and costly, this research aims to determine whether the use of these foundation models can reduce the need for training data. To assess the models' behavior under conditions of reduced training data, five test datasets for semantic segmentation will be utilized. This study will concentrate on traffic sign segmentation to analyze the results in comparison to Mask R-CNN: the field's leading model. The findings indicate that SAM does not surpass the leading model for this specific task, regardless of the quantity of training data. Nevertheless, a knowledge-distilled student architecture derived from SAM exhibits no reduction in accuracy when trained on data that have been reduced by 95%.

Echocardiography is a reliable and non-invasive method for assessing cardiac structure and function in both clinical and experimental settings, offering valuable insights into disease progression and treatment efficacy. The successful application of echocardiography in murine models of disease has enabled the evaluation of disease severity, drug testing, and continuous monitoring of cardiac function in these animals. However, there is insufficient standardization of echocardiographic measurements for smaller animals. This article aims to address this gap by providing a guide and practical tips for the appropriate acquisition and analysis of echocardiographic parameters in adult rats, which may also be applicable in other small rodents used for scientific purposes, like mice. With advancements in technology, such as ultrahigh-frequency ultrasonic transducers, echocardiography has become a highly sophisticated imaging modality, offering high temporal and spatial resolution imaging, thereby allowing for real-time monitoring of cardiac function throughout the lifespan of small animals. Moreover, it allows the assessment of cardiac complications associated with aging, cancer, diabetes, and obesity, as well as the monitoring of cardiotoxicity induced by therapeutic interventions in preclinical models, providing important information for translational research. Finally, this paper discusses the future directions of cardiac preclinical ultrasound, highlighting the need for continued standardization to advance research and improve clinical outcomes to facilitate early disease detection and the translation of findings into clinical practice.