Scan Pattern Research Articles

Effect of preparation design and scan pattern on the accuracy of intraoral scans for complete arch laminate veneers under simulated intraoral variables.

The purpose of the study was to investigate the influence of different preparation designs and scan patterns on the accuracy of intraoral scans for complete-arch maxillary laminate veneers. Three maxillary typodonts were used to obtain reference models with three different laminate veneer preparation designs: windows (W), beveled (B), and incisal overlap (IO). Reference scans were obtained with a desktop scanner. A total of 90 complete arch intraoral scans were made with an intraoral scanner (Medit i700) following three different scan patterns: straight motion (SM), zigzag motion (ZM), and combined motion (CM). Ten scans were made in each subgroup and exported as standard tessellation language (STL) files. Assessment of accuracy was conducted with a 3D software analysis program (Geomagic Control X). Each STL file was individually aligned with the reference scan using the best fit algorithm tool, and 3D differences were calculated using the root mean square (RMS) value. Two-way analysis of variance (ANOVA) followed by post hoc comparison tests were applied to analyze precision and trueness data (α = 0.05). Two-way ANOVA and post hoc comparison tests revealed significant differences among different preparation designs and scan patterns (p < 0.05). Regarding trueness, the IO when scanned with SM presented higher mean RMS than the other preparation designs (W and B) scanned with the same scanning pattern (p < 0.05). Regarding precision, the groups of W and IO presented significantly higher mean RMS than the group of B when scanned with ZM (p < 0.05). Accuracy of intraoral scans for complete-arch laminate veneers was affected by different laminate veneer preparation designs and scan patterns. Modifying scan pattern according to preparation design helps to improve scan accuracy for complete-arch laminate veneers.

Abnormal Scanning Patterns Based on Eye Movement Entropy in Early Psychosis

BackgroundRestricted scan path mode is hypothesized to explain abnormal scanning patterns in patients with schizophrenia. Here, we calculated entropy scores (drawing on gaze data to measure the statistical randomness of eye movements) to quantify how strategical and random participants were when processing image stimuli. MethodsEighty-six patients with first-episode schizophrenia (FES), 124 individuals at clinical high risk (CHR) for psychosis, and 115 healthy control participants (HCs) completed an eye-tracking examination while freely viewing 35 static images (each presented for 10 seconds) and cognitive assessments. We compared group differences in the overall entropy score, as well as entropy scores under various conditions. We also investigated the correlations between entropy scores and symptoms and cognitive function. ResultsIncreased overall entropy scores were noted in the FES and CHR groups compared with the HC group, and these differences were already apparent within 0 to 2.5 seconds. In addition, the CHR group exhibited higher entropy than the HC group when viewing low-meaning images. Moreover, the entropy within 0 to 2.5 seconds showed significant correlations with negative symptoms in the FES group, attention/vigilance scores in the CHR group, and speed of processing and attention/vigilance scores across all 3 groups. ConclusionsThe results indicate that individuals with FES and those at CHR scanned pictures more randomly and less strategically than HCs. These patterns also correlated with clinical symptoms and neurocognition. The current study highlights the potential of the eye movement entropy measure as a neurophysiological marker for early psychosis.

A Novel Dose Rate Optimization Method to Maximize Ultrahigh-Dose-Rate Coverage of Critical Organs at Risk Without Compromising Dosimetry Metrics in Proton Pencil Beam Scanning FLASH Radiation Therapy

PurposeTo investigate a dose rate optimization framework based on the spot scanning patterns to improve ultra-high dose rate coverage of critical organs-at-risk (OARs) for proton PBS FLASH radiotherapy, and to present implementation of a genetic algorithm (GA) method for spot sequence optimization to achieve PBS FLASH dose rate optimization under relatively low nozzle beam currents. Materials and MethodsFirstly, a multi-field FLASH plan was developed to meet all the dosimetric goals and optimal FLASH dose rate coverage by considering the deliverable minimum MU (MMU) constraint. Then, a GA method was implemented into the in-house treatment platform to maximize the dose rate by exploring the best spot delivery sequence. A phantom study was performed to evaluate the effectiveness of the dose rate optimization. Then, 10 consecutive plans for lung cancer patients previously treated using PBS intensity-modulated proton therapy (IMPT) were optimized using 45 GyRBE in 3 fractions both transmission and Bragg peak FLASH-RT for further validation. The spot delivery sequence of each treatment field was optimized using this GA. The ultra-high dose rate volume histogram (DRVH) and dose rate coverage V40GyRBE/s were investigated to assess the efficacy of dose rate optimization quantitatively. ResultsUsing a relatively low MU/spot of 150, corresponding to nozzle beam current of 65 nA, the FLASH dose rate ratio V40GyRBE/s of the OAR contour of the core was increased from 0 to ∼60% in the phantom study. In the lung cancer patients, the ultra-high dose rate coverage V40GyRBE/s were improved from 15.2%, 15.5%, 17.6%, and 16.0% before the delivery sequence optimization, to 31.8%, 43.5%, 47.6%, and 30.5% after delivery sequence optimization, in the lungs-GTV, spinal cord, esophagus, and heart (p-values all<0.001). When beam current increased to 130 nA, V40GyRBE/s was improved from 45.1%, 47.1%, 51.2%, and 51.4%, to 65.3%, 83.5%, 88.1%, and 69.4% (p-values<0.05). The averaged V40GyRBE/s for the target and OARs was increased from 12.9% to 41.6%, and 46.3% to 77.5% for 65 and 130 nA, respectively, showing significant improvements based on a clinical proton system. After optimizing the dose rate for the Bragg peak FLASH technique with a beam current of 340 nA, the V40GyRBE/s for the lung-GTV, spinal cord, esophagus, and heart were increased by 8.9%, 15.8%, 22%, and 20.8%, respectively. ConclusionAn optimal plan quality can be maintained as the spot delivery sequence optimization is a separate independent process after the plan optimization. Both the phantom and patient results demonstrated that novel spot delivery sequence optimization can effectively improve the ultra-high dose rate coverage for critical OARs, which can potentially be applied in clinical practice for better OARs sparing efficacy.

Directional TV algorithm for image reconstruction from sparse-view projections in EPR imaging

Objective. Electron paramagnetic resonance (EPR) imaging is an advanced in vivo oxygen imaging modality. The main drawback of EPR imaging is the long scanning time. Sparse-view projections collection is an effective fast scanning pattern. However, the commonly-used filtered back projection (FBP) algorithm is not competent to accurately reconstruct images from sparse-view projections because of the severe streak artifacts. The aim of this work is to develop an advanced algorithm for sparse reconstruction of 3D EPR imaging. Methods. The optimization based algorithms including the total variation (TV) algorithm have proven to be effective in sparse reconstruction in EPR imaging. To further improve the reconstruction accuracy, we propose the directional TV (DTV) model and derive its Chambolle–Pock solving algorithm. Results. After the algorithm correctness validation on simulation data, we explore the sparse reconstruction capability of the DTV algorithm via a simulated six-sphere phantom and two real bottle phantoms filled with OX063 trityl solution and scanned by an EPR imager with a magnetic field strength of 250 G. Conclusion. Both the simulated and real data experiments show that the DTV algorithm is superior to the existing FBP and TV-type algorithms and a deep learning based method according to visual inspection and quantitative evaluations in sparse reconstruction of EPR imaging. Significance. These insights gained in this work may be used in the development of fast EPR imaging workflow of practical significance.

The other-ethnicity effect in facial recognition of Tibetan and Han individuals: Evidence from behavioural and eye-movement data.

Studies have shown that facial recognition among racial groups exhibits not only an other-race effect but also an other-ethnicity effect within the same racial group. To explore differences in facial recognition and visual scanning patterns due to the other-ethnicity effect, behavioural and eye-movement data were used to investigate the other-ethnicity effect in the facial perception of Tibetan and Han Chinese individuals and whether the visual scanning patterns varied between them. Behavioural data revealed an other-ethnicity effect on facial recognition of Tibetan and Han individuals. Eye-movement data indicated that Tibetan and Han individuals fixated more on the eye and mouth regions when recognising Han faces and on the eye and nose regions when recognising Tibetan faces. The other-ethnicity effect appeared to influence facial recognition in Tibetan and Han individuals, who adopted similar visual scanning patterns when scanning the faces of individuals of their own ethnicity and those of other ethnicities.

Visual scanning patterns of a talking face when evaluating phonetic information in a native and non-native language.

When comprehending speech, listeners can use information encoded in visual cues from a face to enhance auditory speech comprehension. For example, prior work has shown that the mouth movements reflect articulatory features of speech segments and durational information, while pitch and speech amplitude are primarily cued by eyebrow and head movements. Little is known about how the visual perception of segmental and prosodic speech information is influenced by linguistic experience. Using eye-tracking, we studied how perceivers' visual scanning of different regions on a talking face predicts accuracy in a task targeting both segmental versus prosodic information, and also asked how this was influenced by language familiarity. Twenty-four native English perceivers heard two audio sentences in either English or Mandarin (an unfamiliar, non-native language), which sometimes differed in segmental or prosodic information (or both). Perceivers then saw a silent video of a talking face, and judged whether that video matched either the first or second audio sentence (or whether both sentences were the same). First, increased looking to the mouth predicted correct responses only for non-native language trials. Second, the start of a successful search for speech information in the mouth area was significantly delayed in non-native versus native trials, but just when there were only prosodic differences in the auditory sentences, and not when there were segmental differences. Third, (in correct trials) the saccade amplitude in native language trials was significantly greater than in non-native trials, indicating more intensely focused fixations in the latter. Taken together, these results suggest that mouth-looking was generally more evident when processing a non-native versus native language in all analyses, but fascinatingly, when measuring perceivers' latency to fixate the mouth, this language effect was largest in trials where only prosodic information was useful for the task.

LidPose: Real-Time 3D Human Pose Estimation in Sparse Lidar Point Clouds with Non-Repetitive Circular Scanning Pattern.

In this paper, we propose a novel, vision-transformer-based end-to-end pose estimation method, LidPose, for real-time human skeleton estimation in non-repetitive circular scanning (NRCS) lidar point clouds. Building on the ViTPose architecture, we introduce novel adaptations to address the unique properties of NRCS lidars, namely, the sparsity and unusual rosetta-like scanning pattern. The proposed method addresses a common issue of NRCS lidar-based perception, namely, the sparsity of the measurement, which needs balancing between the spatial and temporal resolution of the recorded data for efficient analysis of various phenomena. LidPose utilizes foreground and background segmentation techniques for the NRCS lidar sensor to select a region of interest (RoI), making LidPose a complete end-to-end approach to moving pedestrian detection and skeleton fitting from raw NRCS lidar measurement sequences captured by a static sensor for surveillance scenarios. To evaluate the method, we have created a novel, real-world, multi-modal dataset, containing camera images and lidar point clouds from a Livox Avia sensor, with annotated 2D and 3D human skeleton ground truth.

Manipulating scanning strategies towards controlled microstructure of laser remelted Mg–3Al–1Zn alloy

Various laser scanning strategies of multitrack products, consisting of a default Raster pattern (R1), a Raster pattern with a changed line order (R2), and a default Zigzag pattern (Z1), with different overlap rates ranged from 0.7 to 0.5, have been conducted to investigate microstructure evolution for Mg–3Al–1Zn alloy. As the overlap rate changes from 0.7 to 0.5, the average grain size decreases and the texture is slightly concentrated, except for R2 pattern. For effect of different scanning pattern, a relatively small average grain size can be obtained using the R1, R2 patterns compared to using the Z1 pattern; a relatively weak texture can be obtained using the R1, Z1 patterns compared to using the R2 pattern. The utilization of Z1 pattern results in a further texture weakening compared to that of the R1 pattern. Microstructure evolution of scanning strategy is mainly determined by low-undercooling-required epitaxial growth along temperature gradient directions at the solid/liquid interface. Combined with computational thermal-fluid dynamics simulation, a strong fluid flow within the molten pool promotes side expansion of the molten pool, resulting in the larger overlap area, the corresponding three-times remelting domain and continuous deflections of grain major axes.

Reactive UAV-based automatic tunnel surface defect inspection with a field test

This work addresses the problem of automatic tunnel surface defect inspection using unmanned aerial vehicles (UAVs). The research aims at proposing a robust and efficient monitoring method for image data acquisition and processing in complex and dark tunnel environments. A method, called Proximity Move-Pause-Photo for Surface Defect Inspection (PMPP-SDI), is proposed by combining reactive flying control strategies with a grid scanning pattern to capture high-quality image data from multiple views and angles. The image data is then used to generate a 3D point cloud model of the tunnel surface for structural condition assessment. The method is tested in a field experiment in a railway tunnel in Ireland, and the results show that it can achieve stable navigation, high-resolution reconstruction, and accurate defect detection. The paper discusses the advantages and limitations of the method, and suggests improving the control/navigation intelligence, data quality, and defect analysis as the future research directions.

Rediscovering Age-Related Macular Degeneration with Swept-Source OCT Imaging: The 2022 Charles L. Schepens, MD, Lecture

Swept-source OCT angiography (SS-OCTA) scans of eyes with age-related macular degeneration (AMD) were used to replace color, autofluorescence, infrared reflectance, and dye-based fundus angiographic imaging for the diagnosis and staging of AMD. Through the use of different algorithms with the SS-OCTA scans, both structural and angiographic information can be viewed and assessed using both cross sectional and en face imaging strategies. Presented at the 2022 Charles L. Schepens, MD, Lecture at the American Academy of Ophthalmology Retina Subspecialty Day, Chicago, Illinois, on September 30,2022. Patients with AMD. Review of published literature and ongoing clinical research using SS-OCTA imaging in AMD. Swept-source OCT angiography imaging of AMD at different stages of disease progression. Volumetric SS-OCTA dense raster scans were used to diagnose and stage both exudative and nonexudative AMD. In eyes with nonexudative AMD, a single SS-OCTA scan was used to detect and measure structural features in the macula such as the area and volume of both typical soft drusen and calcified drusen, the presence and location of hyperreflective foci, the presence of reticular pseudodrusen, also known as subretinal drusenoid deposits, the thickness of the outer retinal layer, the presence and thickness of basal laminar deposits, the presence and area of persistent choroidal hypertransmission defects, and the presence of treatment-naïve nonexudative macular neovascularization. In eyes with exudative AMD, the same SS-OCTA scan pattern was used to detect and measure the presence of macular fluid, the presence and type of macular neovascularization, and the response of exudation to treatment with vascular endothelial growth factor inhibitors. In addition, the same scan pattern was used to quantitate choriocapillaris (CC) perfusion, CC thickness, choroidal thickness, and the vascularity of the choroid. Compared with using several different instruments to perform multimodal imaging, a single SS-OCTA scan provides a convenient, comfortable, and comprehensive approach for obtaining qualitative and quantitative anatomic and angiographic information to monitor the onset, progression, and response to therapies in both nonexudative and exudative AMD. Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Generating PET scan patterns in Alzheimer's by a mathematical model.

Alzheimer disease (AD) is the most common form of dementia. The cause of the disease is unknown, and it has no cure. Symptoms include cognitive decline, memory loss, and impairment of daily functioning. The pathological hallmarks of the disease are aggregation of plaques of amyloid-β (Aβ) and neurofibrillary tangles of tau proteins (τ), which can be detected in PET scans of the brain. The disease can remain asymptomatic for decades, while the densities of Aβ and τ continue to grow. Inflammation is considered an early event that drives the disease. In this paper, we develop a mathematical model that can produce simulated patterns of (Aβ,τ) seen in PET scans of AD patients. The model is based on the assumption that early inflammations, R and [Formula: see text], drive the growth of Aβ and τ, respectively. Recently approved drugs can slow the progression of AD in patients, provided treatment begins early, before significant damage to the brain has occurred. In line with current longitudinal studies, we used the model to demonstrate how to assess the efficacy of such drugs when given years before the disease becomes symptomatic.

Synergic effects of scanning pattern and heat treatment on the microstructure-hot tensile behavior relationships of powder bed fusion-laser beam manufactured Hastelloy X

Synergic effects of scanning pattern and heat treatment on the microstructure-hot tensile behavior relationships of powder bed fusion-laser beam manufactured Hastelloy X

HeLiPR: Heterogeneous LiDAR dataset for inter-LiDAR place recognition under spatiotemporal variations

Place recognition is crucial for robot localization and loop closure in simultaneous localization and mapping (SLAM). Light Detection and Ranging (LiDAR), known for its robust sensing capabilities and measurement consistency even in varying illumination conditions, has become pivotal in various fields, surpassing traditional imaging sensors in certain applications. Among various types of LiDAR, spinning LiDARs are widely used, while non-repetitive scanning patterns have recently been utilized in robotics applications. Some LiDARs provide additional measurements such as reflectivity, Near Infrared (NIR), and velocity from Frequency modulated continuous wave (FMCW) LiDARs. Despite these advances, there is a lack of comprehensive datasets reflecting the broad spectrum of LiDAR configurations for place recognition. To tackle this issue, our paper proposes the HeLiPR dataset, curated especially for place recognition with heterogeneous LiDARs, embodying spatiotemporal variations. To the best of our knowledge, the HeLiPR dataset is the first heterogeneous LiDAR dataset supporting inter-LiDAR place recognition with both non-repetitive and spinning LiDARs, accommodating different field of view (FOV)s and varying numbers of rays. The dataset covers diverse environments, from urban cityscapes to high-dynamic freeways, over a month, enhancing adaptability and robustness across scenarios. Notably, HeLiPR dataset includes trajectories parallel to MulRan sequences, making it valuable for research in heterogeneous LiDAR place recognition and long-term studies. The dataset is accessible at https://sites.google.com/view/heliprdataset.

Online Calibration of Extrinsic Parameters for Solid-State LIDAR Systems.

This work addresses the challenge of calibrating multiple solid-state LIDAR systems. The study focuses on three different solid-state LIDAR sensors that implement different hardware designs, leading to distinct scanning patterns for each system. Consequently, detecting corresponding points between the point clouds generated by these LIDAR systems-as required for calibration-is a complex task. To overcome this challenge, this paper proposes a method that involves several steps. First, the measurement data are preprocessed to enhance its quality. Next, features are extracted from the acquired point clouds using the Fast Point Feature Histogram method, which categorizes important characteristics of the data. Finally, the extrinsic parameters are computed using the Fast Global Registration technique. The best set of parameters for the pipeline and the calibration success are evaluated using the normalized root mean square error. In a static real-world indoor scenario, a minimum root mean square error of 7 cm was achieved. Importantly, the paper demonstrates that the presented approach is suitable for online use, indicating its potential for real-time applications. By effectively calibrating the solid-state LIDAR systems and establishing point correspondences, this research contributes to the advancement of multi-LIDAR fusion and facilitates accurate perception and mapping in various fields such as autonomous driving, robotics, and environmental monitoring.

Implementation of different scanning strategy to improve the mechanical and wear properties of 15-5 PH stainless steel fabricated by laser-directed energy deposition

In this work, the effect of three different scan patterns (island scan pattern, island alternative scan pattern and cross-hatch pattern) on mechanical and wear properties is investigated for 15–5 Precipitation-hardened (PH) stainless steel by using laser-directed energy deposition (L-DED). Six rectangular blocks are deposited with constant laser power and scanning speed. Currently, studies on the control of anisotropic behavior to improve the mechanical and wear properties by using different scan strategy, especially in the case of 15–5 PH grade stainless steel is relatively sparse. Therefore, it is necessary to learn how mechanical and wear properties are improving in these three different scan pattern samples. At first, the temperature data at a specific point is captured by using a non-contact type pyrometer during printing and the detailed microstructural characterization of that particular region is analyzed. All three different scan pattern samples mostly show δ-ferrite and lath martensite (α') structure. The X-ray diffraction shows, there is formation of secondary precipitation phases like NbCr2, NbC, and Cu rich precipitation in all three different scan pattern samples. The electron back scattered diffraction (EBSD) result exhibits the island alternative scan pattern samples have mostly mixed orientation direction, a high fraction of retained austenite (͠ 0.004) and high angle grain boundaries (HAGBs) compared to the island scan pattern and cross hatch pattern samples. The island alternative scan pattern samples show maximum improved mechanical properties i.e. average ultimate tensile strength (UTS) of 1316 MPa and up to 20% of elongation amongst all deposited samples. Similarly, the alternative island scan pattern samples exhibit a high coefficient of friction (COF) irrespective of the applied load. The SEM image of the worn surfaces clearly demonstrates abrasion and adhesion type wear in all the three different scan pattern samples.

Mitigation of bio-corrosion characteristics of coronary artery stent by optimising fs-laser micromachining parameters

Cardiovascular diseases, particularly coronary artery disease, pose big challenges to human life. Deployment of the stent is a preferable treatment for the above-mentioned disease. However, stents are usually made up of shape memory alloy called Nitinol. The poorer surface finish on the machined nitinol stents accelerates the migration of Nickel ions from the implanted nitinol stent, which is considered toxic and can lead to stenosis. The current study deals with controlling surface quality by minimising surface roughness and improving corrosion resistance. Femtosecond laser (fs-laser 10−15 s) micromachining was employed to machine the Nitinol surface to achieve sub-micron surface roughness. The Grey relational analysis (GRA)-coupled design of the experimental technique was implemented to determine optimal levels of four micromachining parameters (laser power, pulse frequency, scanning speed, and scanning pattern) varied at three levels to achieve minimum surface roughness and to maximise the volume ablation. The results show that to yield minimum surface roughness and maximum volume ablation, laser power and scanning speed are in a higher range. In contrast, the pulse frequency is lower, and the scanning pattern is in a zig-zag manner. ANOVA results manifest that scanning speed is the predominant factor in minimising surface roughness, followed by pulse frequency. Furthermore, the corrosion behaviour of the machined nitinol specimens was evaluated, and the results show that specimens with lower surface roughness had lower corrosion rates.

Retrieval of temperature profiles in tropical cyclone Nanmadol from resampled advanced technology microwave sounder observations

In the past, satellite microwave sounding data are often utilized for retrieving the thermal structure of tropical cyclones (TCs). However, the spatial resolutions and scan pattern of the instruments vary from one to another and can affect the retrievals of TC structures. In this study, Backus-Gilbert Inversion (BGI), ATOVS and AVHRR Pre-processing Package (AAPP) filter and N × N field-of-views (FOVs) average algorithms are utilized to resample the Advanced Technology Microwave Sounder (ATMS) data and the impacts from different resampling algorithms on brightness temperature structures and retrieved atmospheric temperature are analyzed. It is shown that for ATMS channel 1 to 2, the larger differences between original and resampled brightness temperatures are found in cloudy areas and at the land-ocean boundary. When ATMS channels 1–16 data are resampled to the AMSU-like resolution, the brightness temperatures become less noisy at channel 3 to 16 and can lead to a higher convergence rate in the one-dimension variational (1DVAR) algorithm. Furthermore, the warm core intensity of a tropical storm of Nanmadol is reduced by about 2.5 K and its size becomes smaller after the ATMS data is resampled to the AMSU resolution.

Live Iterative Ptychography.

We demonstrate live-updating ptychographic reconstruction with the extended ptychographical iterative engine, an iterative ptychography method, during ongoing data acquisition. The reconstruction starts with a small subset of the total data, and as the acquisition proceeds the data used for reconstruction are extended. This creates a live-updating view of object and illumination that allows monitoring the ongoing experiment and adjusting parameters with quick turn around. This is particularly advantageous for long-running acquisitions. We show that such a gradual reconstruction yields interpretable results already with a small subset of the data. We show simulated live processing with various scan patterns, parallelized reconstruction, and real-world live processing at the hard X-ray ptychographic nanoanalytical microscope PtyNAMi at the PETRA III beamline.

Characterization of Anisotropic Shape Memory Behavior of Thermoresponsive Components in 4D Printing.

Four-dimensional (4D) printing has emerged as a promising manufacturing technology in recent years and revolutionized products by adding shape-morphing capabilities when exposed to certain stimuli. Increasing research attention has been dedicated to studying the shape memory behaviors of the 4D fabricated structures. However, in-depth discussions on quantifying the influence of process parameters on shape fixity and recovery properties are limited, and the anisotropy induced by the layer-wise fabrication nature is significantly underreported. To further exploit the shape memory property of 4D printed structures, it is essential to investigate the process-induced anisotropic shape memory behaviors. In this study, the effects of critical process parameters on anisotropy in shape memory properties are mathematically quantified; meanwhile, the feasibility of tailoring the anisotropy of 4D printed parts is examined with joint consideration of total build time. Different scanning patterns are experimentally analyzed for their influence on anisotropic behaviors. It is found that the Triangle scanning pattern often leads to the best shape memory behaviors in different directions. The outcome of this study confirms the existence of anisotropy in both shape fixity and shape recovery ratios. In addition, the results also reveal that a smaller scanning angle tends to minimize the anisotropy and total fabrication time while ensuring satisfactory shape memory performance. Furthermore, layer thickness shows negligible effects on anisotropy, while the scanning angle and shape memory temperature suggest the opposite.

Glaucoma detection model by exploiting multi-region and multi-scan-pattern OCT images with dynamical region score.

Currently, deep learning-based methods have achieved success in glaucoma detection. However, most models focus on OCT images captured by a single scan pattern within a given region, holding the high risk of the omission of valuable features in the remaining regions or scan patterns. Therefore, we proposed a multi-region and multi-scan-pattern fusion model to address this issue. Our proposed model exploits comprehensive OCT images from three fundus anatomical regions (macular, middle, and optic nerve head regions) being captured by four scan patterns (radial, volume, single-line, and circular scan patterns). Moreover, to enhance the efficacy of integrating features across various scan patterns within a region and multiple regional features, we employed an attention multi-scan fusion module and an attention multi-region fusion module that auto-assign contribution to distinct scan-pattern features and region features adapting to characters of different samples, respectively. To alleviate the absence of available datasets, we have collected a specific dataset (MRMSG-OCT) comprising OCT images captured by four scan patterns from three regions. The experimental results and visualized feature maps both demonstrate that our proposed model achieves superior performance against the single scan-pattern models and single region-based models. Moreover, compared with the average fusion strategy, our proposed fusion modules yield superior performance, particularly reversing the performance degradation observed in some models relying on fixed weights, validating the efficacy of the proposed dynamic region scores adapted to different samples. Moreover, the derived region contribution scores enhance the interpretability of the model and offer an overview of the model's decision-making process, assisting ophthalmologists in prioritizing regions with heightened scores and increasing efficiency in clinical practice.