Depth Axis Research Articles

Light-field tomographic fluorescence lifetime imaging microscopy

Fluorescence lifetime imaging microscopy (FLIM) is a powerful imaging technique that enables the visualization of biological samples at the molecular level by measuring the fluorescence decay rate of fluorescent probes. This provides critical information about molecular interactions, environmental changes, and localization within biological systems. However, creating high-resolution lifetime maps using conventional FLIM systems can be challenging, as it often requires extensive scanning that can significantly lengthen acquisition times. This issue is further compounded in three-dimensional (3D) imaging because it demands additional scanning along the depth axis. To tackle this challenge, we developed a computational imaging technique called light-field tomographic FLIM (LIFT-FLIM). Our approach allows for the acquisition of volumetric fluorescence lifetime images in a highly data-efficient manner, significantly reducing the number of scanning steps required compared to conventional point-scanning or line-scanning FLIM imagers. Moreover, LIFT-FLIM enables the measurement of high-dimensional data using low-dimensional detectors, which are typically low cost and feature a higher temporal bandwidth. We demonstrated LIFT-FLIM using a linear single-photon avalanche diode array on various biological systems, showcasing unparalleled single-photon detection sensitivity. Additionally, we expanded the functionality of our method to spectral FLIM and demonstrated its application in high-content multiplexed imaging of lung organoids. LIFT-FLIM has the potential to open up broad avenues in both basic and translational biomedical research.

Cascaded spectral light field tomography.

We present cascaded spectral light field tomography for multicolor imaging in three dimensions (3D). Building upon light field tomography, our method uses a Dove prism array and a cylindrical lens array to transform a 3D scene into one-dimensional (1D) projections. To further enhance the reconstructed image quality, we incorporate a rotating Dove prism to increase the number of projection angles and a scanning light sheet to sparsify the sample along the depth axis. The resulting 1D projections are then spectrally dispersed for parallel spectral measurements. We demonstrate the effectiveness of our system in both fluorescence and scattering microscopy applications.

The effect of osteotomy depth and hinge axis orientation on biplanar surgical accuracy in medial opening-wedge high tibial osteotomy-a deeper understanding by 3D simulations.

Not all surgical osteotomy steps have been properly investigated for their potential impact on surgical accuracy. The main study objective was to investigate the osteotomy parameters that have respectively major and minor impact on coronal and sagittal bony accuracy in medial opening-wedge high tibial osteotomy (MOWHTO). Three tibias from an existing 3D MOWHTO osteotomy database were chronologically selected based on segmentation quality, tibial plateau size and the presence of tibial varus. The study consisted of three parts: (I) translating the hinge axis in the coronal plane and switching the osteotomy starting point (30-40 mm) and depth, (II) the hinge axis was rotated stepwise by 10° to perform five simulations, (III) the hinge axis was rotated in the axial plane stepwise by 10° towards anterolateral to perform four simulations (0°, +10°, +20°, +30°). The medial proximal tibial angle (MPTA) and lateral tibial slope were the primary outcomes. Simulations were performed with 5, 10 and 15 mm gap distraction. In the coronal plane, maximum difference in osteotomy depth was 10 mm which represented an MPTA difference of 0.8°-1.1° in 10 mm gap distraction and 1.2°-2.0° in 15 mm gap distraction. Tibial slope remained unchanged. Rotating the hinge axis in the sagittal plane delivered minor changes on both MPTA (<0.5°) and tibial slope (<1.5°) at 10 mm gap distraction. Per 10° of axial rotation of the hinge axis towards anterolateral, the tibial slope increased by 1.0°-1.3° in 10 mm gap distraction while the MPTA remains nearly unchanged. The study showed that the medio-lateral osteotomy length is the main parameter for obtaining bony accuracy in the coronal plane and maintaining a strict perpendicular axial hinge axis position is crucial in preserving the native tibial slope. Correct axial alignment of the hinge axis can be obtained by creating an equal osteotomy depth of the anterior and posterior tibial cortices in the lateral metaphyseal area.

Comparative Evaluation of Animated Scatter Plot Transitions.

Scatter plots are popular for displaying 2D data, but in practice, many data sets have more than two dimensions. For the analysis of such multivariate data, it is often necessary to switch between scatter plots of different dimension pairs, e.g., in a scatter plot matrix (SPLOM). Alternative approaches include a "grand tour" for an overview of the entire data set or creating artificial axes from dimensionality reduction (DR). A cross-cutting concern in all techniques is the ability of viewers to find correspondence between data points in different views. Previous work proposed animations to preserve the mental map between view changes and to trace points as well as clusters between scatter plots of the same underlying data set. In this article, we evaluate a variety of spline- and rotation-based view transitions in a crowdsourced user study focusing on ecological validity. Using the study results, we assess each animation's suitability for tracing points and clusters across view changes. We evaluate whether the order of horizontal and vertical rotation is relevant for task accuracy. The results show that rotations with an orthographic camera or staged expansion of a depth axis significantly outperform all other animation techniques for the traceability of individual points. Further, we provide a ranking of the animated transition techniques for traceability of individual points. However, we could not find any significant differences for the traceability of clusters. Furthermore, we identified differences by animation direction that could guide further studies to determine potential confounds for these differences. We publish the study data for reuse and provide the animation framework as a D3.js plug-in.

Tree species occurring in Amazonian wetland forests consistently show broader range sizes and niche breadths than trees in upland forests.

Generally, species with broad niches also show large range sizes. We investigated the relationship between hydrological niche breadth and geographic range size for Amazonian tree species seeking to understand the role of habitat specialization to Amazonian wetlands and upland forests on the current distribution of tree species. We obtained 571,092 valid occurrence points from GBIF and SpeciesLink to estimate the range size and the niche breadth of 76% of all known Amazonian tree species (5150 tree species). Hydrological niche breadth was measured on different unidimensional axes defined by (1) total annual precipitation; (2) precipitation seasonality; (3) actual evapotranspiration; and (4) water table depth. Geographic range sizes were estimated using alpha-hull adjustments. General linear models were used to relate niche breadth to range size while contrasting tree species occurring and not occurring in wetlands. The hydrological niche breadth of Amazonian tree species varied mostly along the water table depth axis. The average range size for an Amazonian tree species was 751,000 km2 (median of 154,000 km2 and standard deviation of 1,550,000 km2). Niche breadth-range size relationships for Amazonian tree species were positive for all models, and the explanatory power of the models improved when including whether a species occurred in wetlands or in terrestrial uplands. Wetland species had steeper positive slopes for the niche breadth-range size relationship, and consistently larger range sizes for a given niche breadth. Amazonian tree species varied strongly in hydrological niche breadth and range size, but most species had narrow niche breadths and range sizes. Our results suggest that the South American riverscape may have been acting as a corridor for species dispersal in the Neotropical lowlands.

Role of gut microbiota in cardiovascular diseases - a comprehensive review.

The connection between cardiovascular illnesses and the gut microbiota has drawn more and more attention in recent years. According to research, there are intricate relationships between dietary elements, gut bacteria, and their metabolites that affect cardiovascular health. In this study, the role of gut microbiota in cardiovascular disorders is examined, with an emphasis on the cardiac consequences brought on by changes in gut microbiota. This essay discusses the gut-heart axis in depth and in detail. It talks about clinical research looking at how soy consumption, probiotic supplements, and dietary changes affected gut microbiota and cardiovascular risk variables. Our goal is to clarify the possible pathways that connect gut microbiota to cardiovascular health and the implications for upcoming treatment approaches. The authors examine the composition, roles, and effects of the gut microbiota on cardiovascular health, including their contributions to hypertension, atherosclerosis, lipid metabolism, and heart failure. Endotoxemia, inflammation, immunological dysfunction, and host lipid metabolism are some of the potential processes investigated for how the gut microbiota affects cardiac outcomes. The research emphasizes the need for larger interventional studies and personalized medicine strategies to completely understand the complexity of the gut-heart axis and its implications for the management of cardiovascular disease. The development of novel treatment strategies and cutting-edge diagnostic technologies in cardiovascular medicine may be facilitated by a better understanding of this axis.

3D Optical Coherence Tomography image processing in BISCAP: characterization of biofilm structure and properties.

BISCAP is a state-of-the-art tool for automatically characterizing biofilm images obtained from Optical Coherence Tomography. Limited availability of other software tools is reported in the field. BISCAP's first version processes 2D images only. Processing 3D images is a problem of greater scientific relevance since it deals with the entire structure of biofilms instead of their 2D slices. Building on the image-processing principles and algorithms proposed earlier for 2D images, these were adapted to the 3D case, and a more general implementation of BISCAP was developed. The primary goal concerns the extension of the initial methodology to incorporate the depth axis in 3D images; multiple improvements were also made to boost computational performance. The calculation of structural properties and visual outputs was extended to offer new insights into the 3D structure of biofilms. BISCAP was tested using 3D images of biofilms with different morphologies, consistently delivering accurate characterizations of 3D structures in a few minutes using standard laptop machines. Low user dependency is required for image analysis. BISCAP is available from https://github.com/diogonarciso/BISCAP. All images used in the tutorials and the validation examples are available from https://web.fe.up.pt/∼fgm/biscap3d.

Z-Splat: Z-Axis Gaussian Splatting for Camera-Sonar Fusion.

Differentiable 3D-Gaussian splatting (GS) is emerging as a prominent technique in computer vision and graphics for reconstructing 3D scenes. GS represents a scene as a set of 3D Gaussians with varying opacities and employs a computationally efficient splatting operation along with analytical derivatives to compute the 3D Gaussian parameters given scene images captured from various viewpoints. Unfortunately, capturing surround view (360° viewpoint) images is impossible or impractical in many real-world imaging scenarios, including underwater imaging, rooms inside a building, and autonomous navigation. In these restricted baseline imaging scenarios, the GS algorithm suffers from a well-known 'missing cone' problem, which results in poor reconstruction along the depth axis. In this paper, we demonstrate that using transient data (from sonars) allows us to address the missing cone problem by sampling high-frequency data along the depth axis. We extend the Gaussian splatting algorithms for two commonly used sonars and propose fusion algorithms that simultaneously utilize RGB camera data and sonar data. Through simulations, emulations, and hardware experiments across various imaging scenarios, we show that the proposed fusion algorithms lead to significantly better novel view synthesis (5dB improvement in PSNR) and 3D geometry reconstruction (60% lower Chamfer distance).

A Comparative Evaluation of Optical See-Through Augmented Reality in Surgical Guidance.

During traditional surgeries, planning and instrument guidance is displayed on an external screen. Recent developments of augmented reality (AR) techniques can overcome obstacles including hand-eye discoordination and heavy mental load. Among these AR technologies, optical see-through (OST) schemes with stereoscopic displays can provide depth perception and retain the physical scene for safety considerations. However, limitations still exist in certain AR systems and the influence of these factors on surgical performance is yet to explore. To this end, experiments of multi-scale surgical tasks were carried out to compare head-mounted display (HMD) AR and autostereoscopic image overlay (AIO) AR, concerning objective performance and subjective evaluation. To solely analyze effects brought by display techniques, the tracking system in each included display system was identical and similar tracking accuracy was proved by a preliminary experiment. Focus and context rendering was utilized to enhance in-situ visualization for surgical guidance. Latency values of all display systems were assessed and a delay experiment proved the latency differences had no significant impact on user performance. Results of multi-scale surgical tasks showed that HMD outperformed in detailed operations probably due to stable resolution along the depth axis, while AIO had better performance in larger-scale operations for better depth perception. This article helps point out the critical limitations of current OST AR techniques and potentially promotes the progress of AR applications in surgical guidance.

The effect of different depth planes during a manual tracking task in three-dimensional virtual reality space

Unlike ballistic arm movements such as reaching, the contribution of depth information to the performance of manual tracking movements is unclear. Thus, to understand how the brain handles information, we investigated how a required movement along the depth axis would affect behavioral tracking performance, postulating that it would be affected by the amount of depth movement. We designed a visually guided planar tracking task that requires movement on three planes with different depths: a fronto-parallel plane called ROT (0), a sagittal plane called ROT (90), and a plane rotated by 45° with respect to the sagittal plane called ROT (45). Fifteen participants performed a circular manual tracking task under binocular and monocular visions in a three-dimensional (3D) virtual reality space. As a result, under binocular vision, ROT (90), which required the largest depth movement among the tasks, showed the greatest error in 3D. Similarly, the errors (deviation from the target path) on the depth axis revealed significant differences among the tasks. Under monocular vision, significant differences in errors were observed only on the lateral axis. Moreover, we observed that the errors in the lateral and depth axes were proportional to the required movement on these axes under binocular vision and confirmed that the required depth movement under binocular vision determined depth error independent of the other axes. This finding implies that the brain may independently process binocular vision information on each axis. Meanwhile, the required depth movement under monocular vision was independent of performance along the depth axis, indicating an intractable behavior. Our findings highlight the importance of handling depth movement, especially when a virtual reality situation, involving tracking tasks, is generated.

Transoral Ultrasound in the Outpatient Clinic for the Diagnostic Work-Up of Oropharyngeal Cancer: A Feasibility Study.

Magnetic resonance imaging (MRI) is the preferred imaging modality for oropharyngeal cancers (OPCs), but it has difficulties distinguishing between small OPCs and unilateral tonsil hypertrophy. We hypothesized that surgeon-performed transoral ultrasound (US) could be used to accurately detect T-stage OPCs. We performed a single-center prospective diagnostic accuracy study including patients with suspected or biopsy-verified OPCs during outpatient appointments. All patients were offered transoral US and MRI. If transoral US could not be tolerated by the patient, transcervical US was performed. The primary outcome was the diagnostic accuracy of detecting OPCs with US compared to MRI, using histopathology as the reference standard. The secondary outcome was comparing the primary tumor diameters between US and MRI blinded to each other. Out of the 26 patients included in the study, 21 (81%) had OPCs. Transoral US could be performed in 21/21 and 1/5 patients with suspected palatine and lingual tonsil OPCs, respectively. Overall, US diagnostic accuracy was 92%, compared to 81% with MRI (p = 0.37). US and MRI had a high correlation between tumor diameters in the anteroposterior diameter (R = 0.80, p < 0.001), corresponding to the depth axis on US. In conclusion, this small study showed the promise and feasibility of transoral US to improve the initial clinical evaluations of patients with suspected OPCs.

MmPose-NLP: A Natural Language Processing Approach to Precise Skeletal Pose Estimation Using mmWave Radars.

In this article, we presented mmPose-NLP, a novel natural language processing (NLP) inspired sequence-to-sequence (Seq2Seq) skeletal key-point estimator using millimeter-wave (mmWave) radar data. To the best of our knowledge, this is the first method to precisely estimate up to 25 skeletal key points using mmWave radar data alone. Skeletal pose estimation is critical in several applications ranging from autonomous vehicles, traffic monitoring, patient monitoring, and gait analysis, to defense security forensics, and aid both preventative and actionable decision making. The use of mmWave radars for this task, over traditionally employed optical sensors, provides several advantages, primarily its operational robustness to scene lighting and adverse weather conditions, where optical sensor performance degrade significantly. The mmWave radar point-cloud (PCL) data are first voxelized (analogous to tokenization in NLP) and N frames of the voxelized radar data (analogous to a text paragraph in NLP) is subjected to the proposed mmPose-NLP architecture, where the voxel indices of the 25 skeletal key points (analogous to keyword extraction in NLP) are predicted. The voxel indices are converted back to real-world 3-D coordinates using the voxel dictionary used during the tokenization process. Mean absolute error (MAE) metrics were used to measure the accuracy of the proposed system against the ground truth, with the proposed mmPose-NLP offering < 3 cm localization errors in the depth, horizontal, and vertical axes. The effect of the number of input frames versus performance/accuracy was also studied for N = {1,2, ... ,10} . A comprehensive methodology, results, discussions, and limitations are presented in this article. All the source codes and results are made available on GitHub for further research and development in this critical yet emerging domain of skeletal key-point estimation using mmWave radars.

Three-dimensional sonification as a surgical guidance tool

AbstractInteractive Sonification is a well-known guidance method in navigation tasks. Researchers have repeatedly suggested the use of interactive sonification in neuronavigation and image-guided surgery. The hope is to reduce clinicians’ cognitive load through a relief of the visual channel, while preserving the precision provided through image guidance. In this paper, we present a surgical use case, simulating a craniotomy preparation with a skull phantom. Through auditory, visual, and audiovisual guidance, non-clinicians successfully find targets on a skull that provides hardly any visual or haptic landmarks. The results show that interactive sonification enables novice users to navigate through three-dimensional space with a high precision. The precision along the depth axis is highest in the audiovisual guidance mode, but adding audio leads to higher durations and longer motion trajectories.

Radiological evaluation of inferior alveolar nerve displacement after removal of impacted mandibular third molars prior to sagittal split osteotomy

IntroductionWe hypothesize that the removal of mandibular third molars (M3) 6 months prior to a bilateral sagittal split osteotomy (BSSO) could allow the displacement of the inferior alveolar nerve (IAN) in a favorable lingual position. This study aimed to radiographically compare the position of IAN before and after M3 removal in patients with Class II malocclusion. Materials and methodThe CBCT images of 30 randomly selected patients (mean age 15.5 years, 19 females and 11 males) were segmented regarding the mandibular bone and the IAN canal. Mandibles were then superimposed and compared using 3D slicer (www.slicer.org). An orthonormal system was constructed, and the coordinates of IAN were assessed in the x- (horizontal axis), y- (depth axis), and z- (vertical axis) directions. ResultsThe mean changes in x- and z-values were 0.37 %, -0.09 % for the right IAN, 0.07 %, and -0.10 % for the left IAN, respectively. Y-axis was the dimension the most impacted by the M3 removal with a mean variation of -11.96 % for the right IAN, and 0.45 % for the left nerve (p1=0.74 and p2=0.04, respectively). Three patients presented a change in the IAN position superior to 1 mm on at least one coordinate axis. We observed a more important change in x-values of the right IAN in male than in female (p = 0.04), and no significant modifications regarding the other dimensions. Finally, there was no correlation between the age of the patients and the changes in IAN position. ConclusionThis study confirms the absence of influence of mandibular third molar removal on the inferior alveolar nerve route prior to BSSO.

Learning the Deep and the Shallow: Deep-Learning-Based Depth Phase Picking and Earthquake Depth Estimation

Abstract Automated teleseismic earthquake monitoring is an essential part of global seismicity analysis. Although constraining epicenters in an automated fashion is an established technique, constraining event depths is substantially more difficult. One solution to this challenge is teleseismic depth phases, but these can currently not be identified precisely by automatic detection methods. Here, we propose two deep-learning models, DepthPhaseTEAM and DepthPhaseNet, to detect and pick depth phases. For training the models, we create a dataset based on the ISC-EHB bulletin—a high-quality catalog with detailed phase annotations. We show how backprojecting the predicted phase arrival probability curves onto the depth axis yields accurate estimates of earthquake depth. Furthermore, we show how a multistation model, DepthPhaseTEAM, leads to better and more consistent predictions than the single-station model, DepthPhaseNet. To allow direct application of our models, we integrate them within the SeisBench library.

TOF-assisted binocular vision accuracy improvement method for underwater fish size inspection.

To address problems such as the lack of accuracy in acquiring depth maps for dynamic fish 3D measurements by usual binocular vision or a time-of-flight (TOF) depth camera, a TOF-assisted binocular vision depth acquisition algorithm is used to obtain high-quality depth maps. The TOF depth energy function is designed to guide the binocular stereo matching process, which improves the correct matching rate of binocular matching in low-texture regions; the TOF and binocular stereo matching confidence weighting functions are designed to achieve the fusion of the two at pixel level to improve the matching quality of fish in the occluded overlapping regions. The experimental results show that the TOF-assisted binocular vision system improves the accuracy of fish size measurement compared to single binocular vision while reducing the measurement error when the fish body has a significant inclination along the depth axis.

NEIM-03 A GRADUAL INTEGRATION MODEL BASED ON MRI TO PREDICT MOLECULAR PROFILES IN PATIENTS WITH GLIOMA

Abstract Glioma is the most common primary brain tumor. Molecular profiles, including IDH mutation, 1p/19q co-deletion, and MGMT promoter methylation, are highly correlated with the prognosis and clinical decision-making of glioma. To predict the molecular profiles from rough segmentation MRI, we developed and validated a gradual integration model. We proposed a multi-task pseudo-3D model based on rough segmented multiparametric MRI to predict molecular profiles. In this model, convolution along the depth axis was performed between convolution blocks. A total of 750 patients with glioma were retrospectively enrolled from Sun Yatsen University Cancer Center(n=576) and The Cancer Imaging Archive(n=167) to validate the performance of the framework. The model was developed and validated on the local dataset and tested on an independent external (TCIA) test set. For IDH, 1p/19q, and MGMT predictions on the TCIA set, the gradual fusion model achieves AUCs of 0.87, 0.80, and 0.71, respectively, outperforming both input or output integration models and 2D models. To conclude, our gradual fusion model demonstrated the potential to predict molecular profiles based on pre-treatment MRI. Gradual integration may be a reliable method to design 3D medical imaging models.

Automated Visual Inspection of Defects in Transparent Display Layers Using Light-Field 3-D Imaging

Since a display panel comprises multiple layered components, defects may occur within different layers through manufacturing processes. Traditional visual inspection systems with a 2D camera cannot identify the occurrence location among layers. Several 3D imaging technologies, such as CT, TSOM, and MRI, suffer from slow performance and a large form factor. In this work, we propose a novel visual inspection method to detect defects on a display panel using light-field 3D imaging. Without powering the target display panel, we first acquire the high-resolution depth information of defects located inside the transparent layers. We then convert the depth information to the object coordinate system to estimate the physical locations of defects. We automatically classify the types of defects and their layer locations along the depth axis in multiple transparent layers of the display panel. Lastly, our experimental results validate that our method can successfully detect and classify various display defects.

The Cosmos as the Hierarchy of Self-Similarity of the One: Plato’s Doctrine of the Beautiful in the Light of the Theory of Fractals

The introduction of the intensiveness (depth) axis, orthogonal to the extensiveness (length–width–height) axis makes it possible to consider the beautiful Cosmos as the hierarchy of resemblances of the One, on the basis of the diaeresis of line proposed by Plato in the dialogue The Republic. Each line segment obtained as a result of diaeresis is designated by one (1), zero (0) or a numerical combination consisting of ones and zeros. One (1) is assigned the value “intensiveness”, and zero (0) is assigned the value “extensiveness”. Through the application of the intensiveness and extensiveness axes, the interpretation of the segments on each of the levels of the diaeresis of line along the intensiveness axis is given. The One (1=⟂) on zero level of diaeresis is beyond the dialectical opposites, but nevertheless beautiful Νοῦς (Nous, the Cosmic Mind) is Its likeness. On the first level of the diaeresis, both the intensive (1, or point), as the representation of the mental (Nous), and the extensive (0, or line), as the representation of the material, are derived from the One (1=⟂). Nous is purely intensive. The intensive is the same (τὸ ταὐτόν). The mixture of the same and the other (τὸ ἕτερον) is the similar (τὸ ὅμοιον). The similar is “…infinitely fairer than the dissimilar” [Plat. Tim. 33b]. The structuring of the other (as the dissimilar) by Νοῦς (as the same) leads to the creation of true likeness (εἰκών), the beautiful Cosmos, which can be represented by the scheme of the diaeresis of line. The result of the diaeresis of line is a fractal – a self-similar object, each part of which repeats the whole on a reduced scale. Thus, the scheme of the diaeresis of line represents the fractal hierarchy of the Cosmos. Since fractal is self-similar, the hierarchy of the Cosmos is the hierarchy of self-similarity of the One. The diaeresis of line represents the similarity to the One, i.e., self-similarity of the One on each of the levels of the diaeresis of line along the intensiveness axis. Any similarity is self-similarity of the One, and for any existing object, the similarity to the One-Good is true similarity to itself/oneself. Since the One is the source of the Beauty of the Cosmos and of every object in the Cosmos, the condition for preserving or returning beauty to the resemblances is to remain similar to or become similar to the One.

Automatic Labeling of Natural Landmarks for Wheelchair Motion Planning

Labeling landmarks for the mobile plan of the automatic electric wheelchair is essential, because it can assist disabled people. In particular, labeled landmark images will help the wheelchairs to locate landmarks and move more accurately and safely. Here, we propose an automatic detection of natural landmarks in RGBD images for navigation of mobile platforms in an indoor environment. This method can reduce the time for manually collecting and creating a dataset of landmarks. The wheelchair, equipped with a camera system, is allowed to move along corridors to detect and label natural landmarks automatically. These landmarks contain the camera and wheelchair positions with the 3D coordinates when storing the labeled landmark. The feature density method is comprised of Oriented FAST and Rotated BRIEF (ORB) feature extractors. Moreover, the central coordinates of the marked points in the obtained RGB images will be mapped to the images with the depth axis for determining the position of the RGB-D camera system in the spatial domain. An encoder and kinematics equations are applied to determine the position during movement. As expected, the system shows good results, such as a high IoU value of over 0.8 at a distance of less than 2 m and a fast time of 41.66 ms for object detection. This means that our technique is very effective for the automatic movement of the wheelchair.