Next Issue
Volume 12, April
Previous Issue
Volume 12, February
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
6.7
3.3
Journals
J. Imaging
Volume 12
Issue 3
share announcement

J. Imaging, Volume 12, Issue 3 (March 2026) – 50 articles

Cover Story (view full-size image): Resolving individual neurons across whole intact organs, from transgenic mouse brains to centimeter-thick human tissue slabs, is a defining challenge in modern neuroscience. This cover image demonstrates ClearScope, a light-sheet theta microscope from MBF Bioscience designed to meet that challenge. Hundreds of neuronal cell bodies, dendrites, and axons are visible across cleared Thy1-eGFP mouse cerebral cortex, captured here with a 10× objective and 488 nm laser excitation. The vivid color palette is a depth encoding: hue maps position along the optical axis, compressing the full 3D tissue volume into one striking 2D projection. By using dual oblique illumination with perpendicular detection, ClearScope eliminates the lateral size constraints inherent to conventional light-sheet designs. View this paper
  • Issues are regarded as officially published after their release is announced to the table of contents alert mailing list.
  • You may sign up for e-mail alerts to receive table of contents of newly released issues.
  • PDF is the official format for papers published in both, html and pdf forms. To view the papers in pdf format, click on the "PDF Full-text" link, and use the free Adobe Reader to open them.
Order results
Result details
Section
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
18 pages, 6071 KB  
Article
DFENet: A Novel Dual-Path Feature Extraction Network for Semantic Segmentation of Remote Sensing Images
by Li Cao, Zishang Liu, Yan Wang and Run Gao
J. Imaging 2026, 12(3), 141; https://doi.org/10.3390/jimaging12030141 - 23 Mar 2026
Viewed by 295
Abstract
Semantic segmentation of remote sensing images (RSIs) is a fundamental task in geoscience research. However, designing efficient feature fusion modules remains challenging for existing dual-branch or multi-branch architectures. Furthermore, existing deep learning-based architectures predominantly concentrate on spatial feature modeling and context capturing while [...] Read more.
Semantic segmentation of remote sensing images (RSIs) is a fundamental task in geoscience research. However, designing efficient feature fusion modules remains challenging for existing dual-branch or multi-branch architectures. Furthermore, existing deep learning-based architectures predominantly concentrate on spatial feature modeling and context capturing while inherently neglecting the exploration and utilization of critical frequency-domain features, which is crucial for addressing issues of semantic confusion and blurred boundaries in complex remote sensing scenes. To address the challenges of feature fusion and the lack of frequency-domain information, we propose a novel dual-path feature extraction network (DFENet) in this paper. Specifically, a dual-path module (DPM) is developed in DFENet to extract global and local features, respectively. In the global path, after applying the channel splitting strategy, four feature extraction strategies are innovatively integrated to extract global features from different granularities. According to the strategy of supplementing frequency-domain information, a frequency-domain feature extraction block (FFEB) dominated by discrete Wavelet transform (DWT) is designed to effectively captures both high- and low-frequency components. Experimental results show that our method outperforms existing state-of-the-art methods in terms of segmentation performance, achieving a mean intersection over union (mIoU) of 83.09% on the ISPRS Vaihingen dataset and 86.05% on the ISPRS Potsdam dataset. Full article
(This article belongs to the Section Image and Video Processing)
►▼ Show Figures

Figure 1

21 pages, 4335 KB  
Article
Real-Time Small UAV Detection in Complex Airspace Using YOLOv11 with Residual Attention and High-Resolution Feature Enhancement
by Chuang Han, Md Redwan Ullah, Amrul Kayes, Khalid Hasan, Md Abdur Rouf, Md Rakib Hasan, Shen Tao, Guo Gengli and Mohammad Masum Billah
J. Imaging 2026, 12(3), 140; https://doi.org/10.3390/jimaging12030140 - 20 Mar 2026
Viewed by 370
Abstract
Detecting small unmanned aerial vehicles (UAVs) in complex airspace presents significant challenges due to their minimal pixel footprint, resemblance to birds, and frequent occlusion. To address these issues, we propose YOLOv11-ResCBAM, a novel real-time detection framework that integrates a Residual Convolutional Block Attention [...] Read more.
Detecting small unmanned aerial vehicles (UAVs) in complex airspace presents significant challenges due to their minimal pixel footprint, resemblance to birds, and frequent occlusion. To address these issues, we propose YOLOv11-ResCBAM, a novel real-time detection framework that integrates a Residual Convolutional Block Attention Module (ResCBAM) and a high-resolution P2 detection head into the YOLOv11 architecture. ResCBAM enhances channel and spatial feature refinement while preserving original feature contexts through residual connections, and the P2 head maintains fine spatial details crucial for small-object localization. Evaluated on a custom dataset of 4917 images (11,733 after augmentation) across three classes (drone, bird, airplane), our model achieves a mean average precision at the 0.5–0.95 IoU threshold (mAP@0.5–0.95) of 0.845, representing a 7.9% improvement over the baseline YOLOv11n, while maintaining real-time inference at 50.51 FPS. Cross-dataset validation on VisDrone2019-DET and UAVDT benchmarks demonstrates promising generalization trends. This work demonstrates the effectiveness of the proposed approach for UAV surveillance systems, balancing detection accuracy with computational efficiency for deployment in security-critical environments. Full article
(This article belongs to the Topic Computer Vision and Image Processing, 3rd Edition)
►▼ Show Figures

Figure 1

29 pages, 5347 KB  
Article
Optimized Reinforcement Learning-Driven Model for Remote Sensing Change Detection
by Yan Zhao, Zhiyun Xiao, Tengfei Bao and Yulong Zhou
J. Imaging 2026, 12(3), 139; https://doi.org/10.3390/jimaging12030139 - 19 Mar 2026
Viewed by 233
Abstract
In recent years, deep learning has driven remarkable progress in remote sensing change detection (CD); however, practical deployment is still hindered by two limitations. First, CD results are easily degraded by imaging-induced uncertainties—mixed pixels and blurred boundaries, radiometric inconsistencies (e.g., shadows and seasonal [...] Read more.
In recent years, deep learning has driven remarkable progress in remote sensing change detection (CD); however, practical deployment is still hindered by two limitations. First, CD results are easily degraded by imaging-induced uncertainties—mixed pixels and blurred boundaries, radiometric inconsistencies (e.g., shadows and seasonal illumination changes), and slight residual misregistration—leading to pseudo-changes and fragmented boundaries. Second, prevailing methods follow a static one-pass inference paradigm and lack an explicit feedback mechanism for adaptive error correction, which weakens generalization in complex or unseen scenes. To address these issues, we propose a feedback-driven CD framework that integrates a dual-branch U-Net with deep reinforcement learning (RL) for pixel-level probabilistic iterative refinement of an initial change probability map. The backbone produces a preliminary posterior estimate of change likelihood from multi-scale bi-temporal features, while a PPO-based RL agent formulates refinement as a Markov decision process. The agent leverages a state representation that fuses multi-scale features, prediction confidence/uncertainty, and spatial consistency cues (e.g., neighborhood coherence and edge responses) to apply multi-step corrective actions. From an imaging and interpretation perspective, the RL module can be viewed as a learnable, self-adaptive imaging optimization mechanism: for high-risk regions affected by blurred boundaries, radiometric inconsistencies, and local misalignment, the agent performs feedback-driven multi-step corrections to improve boundary fidelity and spatial coherence while suppressing pseudo-changes caused by shadows and illumination variations. Experiments on four datasets (CDD, SYSU-CD, PVCD, and BRIGHT) verify consistent improvements. Using SiamU-Net as an example, the proposed RL refinement increases mIoU by 3.07, 2.54, 6.13, and 3.1 points on CDD, SYSU-CD, PVCD, and BRIGHT, respectively, with similarly consistent gains observed when the same RL module is integrated into other representative CD backbones. Full article
(This article belongs to the Section AI in Imaging)
►▼ Show Figures

Figure 1

15 pages, 2099 KB  
Review
Current Trends and Future Prospects of Radiomics and Machine Learning (ML) Models in Spinal Tumors—A Narrative Review
by Vivek Sanker, Suhrud Panchawgh, Anmol Kaur, Vinay Suresh, Dhanya Mahesh, Eeman Ahmad, Srinath Hariharan, Dhiraj Pangal, Maria Jose Cavgnaro, Mirabela Rusu, John Ratliff and Atman Desai
J. Imaging 2026, 12(3), 138; https://doi.org/10.3390/jimaging12030138 - 19 Mar 2026
Viewed by 322
Abstract
The intersection between radiomics, the computational analysis of imaging data, and machine learning (ML) may lead to new developments in the diagnosis, prognosis, and management of diseases. For spinal tumors specifically, applications of these fields appear promising. In this educational narrative review, we [...] Read more.
The intersection between radiomics, the computational analysis of imaging data, and machine learning (ML) may lead to new developments in the diagnosis, prognosis, and management of diseases. For spinal tumors specifically, applications of these fields appear promising. In this educational narrative review, we provide a summary of the current advancements in radiomics and artificial intelligence (AI), as well as applications of both fields in the diagnosis and management of spinal tumors. We also provide a suggested workflow of radiomics and machine learning analysis of spinal tumors for researchers, including a list and description of commonly used radiomic features. Future directions in the field of radiomics and machine learning applications to spinal tumors may involve validating already proposed algorithms with larger datasets, ensuring that all computational applications to patient care maintain high ethical standards, and continuing work in developing novel and highly accurate computational techniques to enhance patient outcomes. Full article
(This article belongs to the Topic Machine Learning and Deep Learning in Medical Imaging)
►▼ Show Figures

Figure 1

18 pages, 2314 KB  
Article
Efficient Two-Stage Autofocus for Micro-Assembly Based on Joint Spatial-Frequency Image Quality Assessment
by Jianpeng Zhang, Tianbo Kang, Xin Zhao, Mingzhu Sun and Yi Yang
J. Imaging 2026, 12(3), 137; https://doi.org/10.3390/jimaging12030137 - 19 Mar 2026
Viewed by 270
Abstract
Reliable autofocus is a fundamental prerequisite for precise positioning in micro-assembly systems, where complex reflections, scale variations, and narrow depth-of-field often degrade the robustness of traditional sharpness metrics. To address these challenges, we propose an efficient two-stage autofocus method for a dual-camera micro-vision [...] Read more.
Reliable autofocus is a fundamental prerequisite for precise positioning in micro-assembly systems, where complex reflections, scale variations, and narrow depth-of-field often degrade the robustness of traditional sharpness metrics. To address these challenges, we propose an efficient two-stage autofocus method for a dual-camera micro-vision system based on a spatial-frequency image quality assessment (IQA) model. First, we design WaveMamba-IQA for image sharpness estimation, synergistically combining the Discrete Wavelet Transform with Vision Transformers to capture high-frequency details and semantic features, further enhanced by Multi-Linear Transposed Attention and Vision Mamba for global context modeling. Moreover, we implement a coarse-to-fine autofocus workflow, employing the Covariance Matrix Adaptation Evolution Strategy for global optimization on the horizontal camera, followed by geometric prior-based precise adjustment for the oblique camera. Experimental results on a custom microsphere dataset demonstrate that WaveMamba-IQA achieves a Spearman correlation coefficient of 0.9786. Furthermore, the integrated system achieves a 98.33% autofocus success rate across varying lighting conditions. This method significantly improves the robustness and automation level of micro-assembly systems, effectively overcoming the limitations of manual and traditional focusing techniques. Full article
(This article belongs to the Section Image and Video Processing)
►▼ Show Figures

Figure 1

16 pages, 1459 KB  
Article
Machine Learning-Assisted Classification of Pathogenic Yeasts Using Laser Light Scattering and Conventional Microscopy
by Xiaoxuan Liu, Shamanth Shankarnarayan, Zexi Cheng, Manisha Gupta, Wojciech Rozmus, Mrinal Mandal, Daniel A. Charlebois and Ying Yin Tsui
J. Imaging 2026, 12(3), 136; https://doi.org/10.3390/jimaging12030136 - 19 Mar 2026
Viewed by 298
Abstract
Yeast infections are a major concern in clinical settings, and several known species are recognized for their antifungal drug resistance, especially the multidrug-resistant pathogen Candidozyma auris. It is of increasing importance to identify pathogenic yeasts to improve treatment outcomes. We present a [...] Read more.
Yeast infections are a major concern in clinical settings, and several known species are recognized for their antifungal drug resistance, especially the multidrug-resistant pathogen Candidozyma auris. It is of increasing importance to identify pathogenic yeasts to improve treatment outcomes. We present a technique to identify these yeast pathogens using machine learning with a neural network (DenseNet-201) on images obtained from laser light scattering and conventional microscopy. We performed the binary classification of seven species of pathogenic yeast based on their light scattering patterns and their microscopy images. We achieved an average classification accuracy of 95.3% for light scattering patterns and 96.6% for microscopy images of the yeast cells. We also demonstrate high classification accuracy when isolating Candidozyma auris images from all other species combined, at an average of 95.1% for light scattering patterns and 96.7% for microscopy images. The high average classification accuracies suggest that both light scattering and microscopy image data can be combined with machine learning models to classify pathogenic yeasts. Full article
(This article belongs to the Section AI in Imaging)
►▼ Show Figures

Figure 1

13 pages, 2003 KB  
Article
External Validation of an Open-Source Model for Automated Muscle Segmentation in CT Imaging of Cancer Patients
by Hendrik Erenstein, Jona Van den Broeck, Annemieke van der Heij-Meijer, Wim P. Krijnen, Aldo Scafoglieri, Harriët Jager-Wittenaar, Martine Sealy and Peter van Ooijen
J. Imaging 2026, 12(3), 135; https://doi.org/10.3390/jimaging12030135 - 18 Mar 2026
Viewed by 282
Abstract
Computed tomography (CT) at the third lumbar vertebra (L3) is widely used for muscle quantification, but manual segmentation is labor intensive. This study externally validates an AI model, trained on a public dataset, for automated L3 muscle segmentation using an independent cohort, including [...] Read more.
Computed tomography (CT) at the third lumbar vertebra (L3) is widely used for muscle quantification, but manual segmentation is labor intensive. This study externally validates an AI model, trained on a public dataset, for automated L3 muscle segmentation using an independent cohort, including a subgroup analysis of subject characteristics (e.g., age and a history of cancer). The AI model was trained on 900 CT scans with expert annotations from a publicly available repository. Validation was performed on 232 PET CT scans from the University Hospital Brussels, each manually segmented by an expert. Segmentation post-processing employed a density-based clustering algorithm to discard arm muscles and Hounsfield unit (HU) thresholding to refine the muscle segmentation. Performance was assessed using the Dice Similarity Coefficient (DSC) and Segmentation Surface Error (SSE). The model achieved a median DSC of 0.978 and a median SSE of 3.863 cm2 across the validation set. At lower BMI values, the model was more prone to overestimation of muscle surface area. Most segmentation errors occurred in the abdominal wall muscles. Analysis showed no significant difference between arm positioning above the head and alongside the body, indicating robustness to minor artifacts from arm positioning. The AI model delivers accurate, automated L3 muscle segmentation, supporting larger-scale body composition studies. However, diminished accuracy at low BMI values and limited demographic diversity of the data highlight the need for broader validation. Full article
(This article belongs to the Topic Machine Learning and Deep Learning in Medical Imaging)
►▼ Show Figures

Figure 1

25 pages, 9628 KB  
Article
Real-Time Endoscopic Video Enhancement via Degradation Representation Estimation and Propagation
by Handing Xu, Zhenguo Nie, Tairan Peng and Xin-Jun Liu
J. Imaging 2026, 12(3), 134; https://doi.org/10.3390/jimaging12030134 - 16 Mar 2026
Viewed by 349
Abstract
Endoscopic images are often degraded by uneven illumination, motion blur, and tissue occlusion, which obscure critical anatomical details and complicate surgical manipulation. This issue is particularly pronounced in single-port endoscopic surgery, where the imaging capability of the camera is further constrained by limited [...] Read more.
Endoscopic images are often degraded by uneven illumination, motion blur, and tissue occlusion, which obscure critical anatomical details and complicate surgical manipulation. This issue is particularly pronounced in single-port endoscopic surgery, where the imaging capability of the camera is further constrained by limited working space. While deep learning-based enhancement methods have demonstrated impressive performance, most existing approaches remain too computationally demanding for real-time surgical use. To address this challenge, we propose an efficient stepwise endoscopic image enhancement framework that introduces an implicit degradation representation as an intermediate feature to guide the enhancement module toward high-quality results. The framework further exploits the temporal continuity of endoscopic videos, based on the assumption that image degradation evolves smoothly over short time intervals. Accordingly, high-quality degradation representations are estimated only on key frames at fixed intervals, while the representations for the remaining frames are obtained through fast inter-frame propagation, thereby significantly improving computational efficiency while maintaining enhancement quality. Experimental results demonstrate that our method achieves an excellent balance between enhancement quality and computational efficiency. Further evaluation on the downstream segmentation task suggests that our method substantially enhances the understanding of the surgical scene, validating that implicitly learning and degradation representation propagation offer a practical pathway for real-time clinical application. Full article
(This article belongs to the Section Medical Imaging)
►▼ Show Figures

Figure 1

18 pages, 23505 KB  
Article
ArtUnmasked: A Multimodal Classifier for Real, AI, and Imitated Artworks
by Akshad Chidrawar and Garima Bajwa
J. Imaging 2026, 12(3), 133; https://doi.org/10.3390/jimaging12030133 - 16 Mar 2026
Viewed by 341
Abstract
Differentiating AI-generated, real, or imitated artworks is becoming a tedious and computationally challenging problem in digital art analysis. AI-generated art has become nearly indistinguishable from human-made works, posing a significant threat to copyrighted content. This content is appearing on online platforms, at exhibitions, [...] Read more.
Differentiating AI-generated, real, or imitated artworks is becoming a tedious and computationally challenging problem in digital art analysis. AI-generated art has become nearly indistinguishable from human-made works, posing a significant threat to copyrighted content. This content is appearing on online platforms, at exhibitions, and in commercial galleries, thereby escalating the risk of copyright infringement. This sudden increase in generative images raises concerns like authenticity, intellectual property, and the preservation of cultural heritage. Without an automated, comprehensible system to determine whether an artwork has been AI-generated, authentic (real), or imitated, artists are prone to the reduction of their unique works. Institutions also struggle to curate and safeguard authentic pieces. As the variety of generative models continues to grow, it becomes a cultural necessity to build a robust, efficient, and transparent framework for determining whether a piece of art or an artist is involved in potential copyright infringement. To address these challenges, we introduce ArtUnmasked, a practical and interpretable framework capable of (i) efficiently distinguishing AI-generated artworks from real ones using a lightweight Spectral Artifact Identification (SPAI), (ii) a TagMatch-based artist filtering module for stylistic attribution, and (iii) a DINOv3–CLIP similarity module with patch-level correspondence that leverages the one-shot generalization ability of modern vision transformers to determine whether an artwork is authentic or imitated. We also created a custom dataset of ∼24K imitated artworks to complement our evaluation and support future research. The complete implementation is available in our GitHub repository. Full article
(This article belongs to the Section AI in Imaging)
►▼ Show Figures

Figure 1

18 pages, 11393 KB  
Article
Bounding Box-Guided Diffusion for Synthesizing Industrial Images and Segmentation Maps
by Emanuele Caruso, Francesco Pelosin, Alessandro Simoni and Oswald Lanz
J. Imaging 2026, 12(3), 132; https://doi.org/10.3390/jimaging12030132 - 16 Mar 2026
Viewed by 266
Abstract
Synthetic dataset generation in Computer Vision, particularly for industrial applications, is still underexplored. Industrial defect segmentation, for instance, requires highly accurate labels, yet acquiring such data is costly and time-consuming. To address this challenge, we propose a novel diffusion-based pipeline for generating high-fidelity [...] Read more.
Synthetic dataset generation in Computer Vision, particularly for industrial applications, is still underexplored. Industrial defect segmentation, for instance, requires highly accurate labels, yet acquiring such data is costly and time-consuming. To address this challenge, we propose a novel diffusion-based pipeline for generating high-fidelity industrial datasets with minimal supervision. Our approach conditions the diffusion model on enriched bounding-box representations to produce precise segmentation masks, ensuring realistic and accurately localized defect synthesis. Compared to existing layout-conditioned generative methods, our approach improves defect consistency and spatial accuracy. We introduce two quantitative metrics to evaluate the effectiveness of our method and assess its impact on a downstream segmentation task trained on real and synthetic data. Our results demonstrate that diffusion-based synthesis can bridge the gap between artificial and real-world industrial data, fostering more reliable and cost-efficient segmentation models. Full article
(This article belongs to the Topic Visual Computing and Understanding: New Developments and Trends)
►▼ Show Figures

Figure 1

29 pages, 15263 KB  
Article
Advanced Sensitive Feature Machine Learning for Aesthetic Evaluation Prediction of Industrial Products
by Jinyan Ouyang, Ziyuan Xi, Jianning Su, Shutao Zhang, Ying Hu and Aimin Zhou
J. Imaging 2026, 12(3), 131; https://doi.org/10.3390/jimaging12030131 - 16 Mar 2026
Viewed by 273
Abstract
As product aesthetics increasingly drive consumer preference, quantitative evaluation remains hindered by subjective evaluation biases and the black-box nature of modern artificial intelligence. This study proposes an advanced machine learning framework incorporating sensitivity-aware morphological features for the aesthetic evaluation of industrial products, with [...] Read more.
As product aesthetics increasingly drive consumer preference, quantitative evaluation remains hindered by subjective evaluation biases and the black-box nature of modern artificial intelligence. This study proposes an advanced machine learning framework incorporating sensitivity-aware morphological features for the aesthetic evaluation of industrial products, with automotive design as a representative case. An aesthetic index system and its quantitative formulations are first developed to capture the morphological characteristics of product form. Subjective weights are determined via grey relational analysis (GRA), while objective weights are calculated using the coefficient of variation method (CVM) integrated with the technique for order preference by similarity to an ideal solution (TOPSIS). A game-theoretic weighting approach is then employed to fuse subjective and objective weights, thereby establishing a multi-scale aesthetic evaluation system. Sensitivity analysis is applied to identify six key indicators, forming a high-quality dataset. To enhance prediction performance, a novel model—improved lung performance-based optimization with backpropagation neural network (ILPOBP)—is proposed, where the optimization process leverages a maximin latin hypercube design (MLHD) to enhance exploration efficiency. The ILPOBP model effectively predicts aesthetic ratings based on limited morphological input data. Experimental results demonstrate that the ILPOBP model outperforms baseline models in terms of accuracy and robustness when handling complex aesthetic information, achieving a significantly lower test set mean absolute relative error (MARE = 4.106%). To further enhance model interpretability, Shapley additive explanations (SHAP) are employed to elucidate the internal decision-making mechanisms, offering reverse design insights for product optimization. The proposed framework offers a novel and effective approach for integrating machine learning into the aesthetic assessment of industrial product design. Full article
(This article belongs to the Section AI in Imaging)
►▼ Show Figures

Figure 1

21 pages, 4219 KB  
Article
3D-StyleGAN2-ADA: Volumetric Synthesis of Realistic Prostate T2W MRI
by Claudia Giardina and Verónica Vilaplana
J. Imaging 2026, 12(3), 130; https://doi.org/10.3390/jimaging12030130 - 14 Mar 2026
Viewed by 305
Abstract
This work investigates the extension of StyleGAN2-ADA to three-dimensional prostate T2-weighted (T2W) MRI generation. The architecture is adapted to operate on 3D anisotropic volumes, enabling stable training at a clinically relevant resolution of 256×256×24, where a baseline 3D-StyleGAN [...] Read more.
This work investigates the extension of StyleGAN2-ADA to three-dimensional prostate T2-weighted (T2W) MRI generation. The architecture is adapted to operate on 3D anisotropic volumes, enabling stable training at a clinically relevant resolution of 256×256×24, where a baseline 3D-StyleGAN fails to converge. Quantitative evaluation using Fréchet Inception Distance (FID), Kernel Inception Distance (KID), and generative Precision–Recall metrics demonstrates substantial improvements over a 3D-StyleGAN baseline. Specifically, FID decreased from 114.2 to 27.3, while generative Precision increased from 0.22 to 0.82, indicating markedly improved fidelity and alignment with the real data distribution. Beyond generative metrics, the synthetic volumes were evaluated through radiomic feature analysis and downstream prostate segmentation. Synthetic data augmentation resulted in segmentation performance comparable to real-data training, supporting that volumetric generation preserves anatomically relevant structures, while multivariate radiomic analyses showed strong global feature alignment between real and synthetic volumes. These findings indicate that a 3D extension of StyleGAN2-ADA enables stable high-resolution volumetric prostate MRI synthesis while preserving anatomically coherent structure and global radiomic characteristics. Full article
(This article belongs to the Topic Machine Learning and Deep Learning in Medical Imaging)
►▼ Show Figures

Figure 1

14 pages, 3098 KB  
Article
Quantitative Ultrasound Texture Analysis of Breast Tumor Responses to Chemotherapy: Comparison of a Cart-Based and a Wireless Ultrasound Scanner
by David Alberico, Maria Lourdes Anzola Pena, Laurentius O. Osapoetra, Lakshmanan Sannachi, Joyce Yip, Sonal Gandhi, Frances Wright, Michael Oelze and Gregory J. Czarnota
J. Imaging 2026, 12(3), 129; https://doi.org/10.3390/jimaging12030129 - 13 Mar 2026
Viewed by 299
Abstract
This study assessed the level of agreement between quantitative ultrasound (QUS) feature estimates derived from ultrasound images of breast tumors in women with locally advanced breast cancer (LABC) produced using a cart-based and a handheld ultrasound system. Thirty LABC patients receiving neoadjuvant chemotherapy [...] Read more.
This study assessed the level of agreement between quantitative ultrasound (QUS) feature estimates derived from ultrasound images of breast tumors in women with locally advanced breast cancer (LABC) produced using a cart-based and a handheld ultrasound system. Thirty LABC patients receiving neoadjuvant chemotherapy were imaged at two separate times: a pre-treatment ‘baseline’ time point, and four weeks after the start of chemotherapy. Three sets of QUS features were produced using the reference phantom technique, one for each imaging time and a third set calculated by taking the differences in feature estimates between times. Cross-system statistical testing using the Wilcoxon signed-rank test was performed for each feature set to assess the level of feature estimate agreement between ultrasound systems. The Bland–Altman method was employed to graphically assess feature sets for systematic skew. The range of p-values was 4.50 × 10−11 to 0.277 for the baseline features, 2.77 × 10−5 to 0.865 for the week 4 features, and 2.03 × 10−9 to 1 for the feature differences. For the feature differences, all five of the primary QUS features (MBF, SS, SI, ASD, AAC) were found to be in agreement between the two scanner types at the 5% confidence level. For the baseline feature set and week 4 feature set, 0 out of 5 and 3 out of 5 of the primary features were found to be in agreement, respectively. Of the 20 QUS texture features examined, the number and proportion of the total for each feature set which were found to have statistically significant similarity in their sample medians at the 5% confidence level were as follows: 2 out of 20 (10%) for the baseline features; 17 out of 20 (85%) for the week 4 features; and 12 out of 20 (60%) for the feature differences. The specific texture features found to be in agreement varied between QUS-specific feature sets. Overall, a moderate level of agreement between sets of feature differences produced using the two systems was demonstrated. Full article
(This article belongs to the Section Medical Imaging)
►▼ Show Figures

Figure 1

30 pages, 3812 KB  
Review
Video-Based 3D Reconstruction: A Review of Photogrammetry and Visual SLAM Approaches
by Ali Javadi Moghadam, Abbas Kiani, Reza Naeimaei, Shirin Malihi and Ioannis Brilakis
J. Imaging 2026, 12(3), 128; https://doi.org/10.3390/jimaging12030128 - 13 Mar 2026
Viewed by 788
Abstract
Three-dimensional (3D) reconstruction using images is one of the most significant topics in computer vision and photogrammetry, with wide-ranging applications in robotics, augmented reality, and mapping. This study investigates methods of 3D reconstruction using video (especially monocular video) data and focuses on techniques [...] Read more.
Three-dimensional (3D) reconstruction using images is one of the most significant topics in computer vision and photogrammetry, with wide-ranging applications in robotics, augmented reality, and mapping. This study investigates methods of 3D reconstruction using video (especially monocular video) data and focuses on techniques such as Structure from Motion (SfM), Multi-View Stereo (MVS), Visual Simultaneous Localization and Mapping (V-SLAM), and videogrammetry. Based on a statistical analysis of SCOPUS records, these methods collectively account for approximately 6863 journal publications up to the end of 2024. Among these, about 80 studies are analyzed in greater detail to identify trends and advancements in the field. The study also shows that the use of video data for real-time 3D reconstruction is commonly addressed through two main approaches: photogrammetry-based methods, which rely on precise geometric principles and offer high accuracy at the cost of greater computational demand; and V-SLAM methods, which emphasize real-time processing and provide higher speed. Furthermore, the application of IMU data and other indicators, such as color quality and keypoint detection, for selecting suitable frames for 3D reconstruction is investigated. Overall, this study compiles and categorizes video-based reconstruction methods, emphasizing the critical step of keyframe extraction. By summarizing and illustrating the general approaches, the study aims to clarify and facilitate the entry path for researchers interested in this area. Finally, the paper offers targeted recommendations for improving keyframe extraction methods to enhance the accuracy and efficiency of real-time video-based 3D reconstruction, while also outlining future research directions in addressing challenges like dynamic scenes, reducing computational costs, and integrating advanced learning-based techniques. Full article
(This article belongs to the Section Computer Vision and Pattern Recognition)
►▼ Show Figures

Figure 1

26 pages, 2632 KB  
Article
Automated Malaria Ring Form Classification in Blood Smear Images Using Ensemble Parallel Neural Networks
by Pongphan Pongpanitanont, Naparat Suttidate, Manit Nuinoon, Natthida Khampeeramao, Sakhone Laymanivong and Penchom Janwan
J. Imaging 2026, 12(3), 127; https://doi.org/10.3390/jimaging12030127 - 12 Mar 2026
Viewed by 266
Abstract
Manual microscopy for malaria diagnosis is labor-intensive and prone to inter-observer variability. This study presents an automated binary classification approach for detecting malaria ring-form infections in thin blood smear single-cell images using a parallel neural network framework. Utilizing a balanced Kaggle dataset of [...] Read more.
Manual microscopy for malaria diagnosis is labor-intensive and prone to inter-observer variability. This study presents an automated binary classification approach for detecting malaria ring-form infections in thin blood smear single-cell images using a parallel neural network framework. Utilizing a balanced Kaggle dataset of 27,558 erythrocyte crops, images were standardized to 128 × 128 pixels and subjected to on-the-fly augmentation. The proposed architecture employs a dual-branch fusion strategy, integrating a convolutional neural network for local morphological feature extraction with a multi-head self-attention branch to capture global spatial relationships. Performance was rigorously evaluated using 10-fold stratified cross-validation and an independent 10% hold-out test set. Results demonstrated high-level discrimination, with all models achieving an ROC–AUC of approximately 0.99. The primary model (Model#1) attained a peak mean accuracy of 0.9567 during cross-validation and 0.97 accuracy (macro F1-score: 0.97) on the independent test set. In contrast, increasing architectural complexity in Model#3 led to a performance decline (0.95 accuracy) due to higher false-positive rates. These findings suggest that moderate-capacity feature fusion, combining convolutional descriptors with attention-based aggregation, provides a robust and generalizable solution for automated malaria screening without the risks associated with over-parameterization. Despite a strong performance, immediate clinical use remains limited because the model was developed on pre-segmented single-cell images, and external validation is still required before routine implementation. Full article
(This article belongs to the Section AI in Imaging)
►▼ Show Figures

Figure 1

20 pages, 3983 KB  
Article
Parameter Selection in Coupled Dynamical Systems for Tomographic Image Reconstruction
by Ryosuke Kasai, Omar M. Abou Al-Ola and Tetsuya Yoshinaga
J. Imaging 2026, 12(3), 126; https://doi.org/10.3390/jimaging12030126 - 12 Mar 2026
Viewed by 224
Abstract
This study investigates the performance of image-reconstruction methods derived from coupled dynamical systems for solving linear inverse problems, focusing on how appropriate parameter selection enhances noise-suppression capability in tomographic image reconstruction. Our previous work has established the stability of linear and nonlinear variants [...] Read more.
This study investigates the performance of image-reconstruction methods derived from coupled dynamical systems for solving linear inverse problems, focusing on how appropriate parameter selection enhances noise-suppression capability in tomographic image reconstruction. Our previous work has established the stability of linear and nonlinear variants of such systems on the basis of Lyapunov’s theorem. However, the influence of parameter choice on reconstruction quality has not been fully clarified. To address this issue, we introduce a parameter adjustment strategy based on an optimization principle. Two complementary optimization strategies are considered. The first employs ground-truth images to determine optimal parameter values that serve as a numerical benchmark for evaluating reconstruction performance. The second relies solely on measured projection data, enabling practical application without prior knowledge of the true image. Numerical experiments using phantoms with relatively high noise levels demonstrate that appropriate parameter selection markedly improves reconstruction accuracy and robustness. These results clarify how properly tuned reconstruction methods derived from coupled dynamical systems can effectively exploit their inherent dynamics to achieve noise suppression in tomographic inverse problems. Full article
(This article belongs to the Special Issue Advances in Photoacoustic Imaging: Tomography and Applications)
►▼ Show Figures

Figure 1

24 pages, 8525 KB  
Article
Consistency-Driven Dual-Teacher Framework for Semi-Supervised Zooplankton Microscopic Image Segmentation
by Zhongwei Li, Yinglin Wang, Dekun Yuan, Yanping Qi and Xiaoli Song
J. Imaging 2026, 12(3), 125; https://doi.org/10.3390/jimaging12030125 - 12 Mar 2026
Viewed by 244
Abstract
In-depth research on marine biodiversity is essential for understanding and protecting marine ecosystems, where semantic segmentation of marine species plays a crucial role. However, segmenting microscopic zooplankton images remains challenging due to highly variable morphologies, complex boundaries, and the scarcity of high-quality pixel-level [...] Read more.
In-depth research on marine biodiversity is essential for understanding and protecting marine ecosystems, where semantic segmentation of marine species plays a crucial role. However, segmenting microscopic zooplankton images remains challenging due to highly variable morphologies, complex boundaries, and the scarcity of high-quality pixel-level annotations that require expert knowledge. Existing semi-supervised methods often rely on single-model perspectives, producing unreliable pseudo-labels and limiting performance in such complex scenarios. To address these challenges, this paper proposes a consistency-driven dual-teacher framework tailored for zooplankton segmentation. Two heterogeneous teacher networks are employed: one captures global morphological features, while the other focuses on local fine-grained details, providing complementary and diverse supervision and alleviating overfitting under limited annotations. In addition, a dynamic fusion-based pseudo-label filtering strategy is introduced to adaptively integrate hard and soft labels by jointly considering prediction consistency and confidence scores, thereby enhancing supervision flexibility. Extensive experiments on the Zooplankton-21 Microscopic Segmentation Dataset (ZMS-21), a self-constructed microscopic zooplankton dataset demonstrate that the proposed method consistently outperforms existing semi-supervised segmentation approaches under various annotation ratios, achieving mIoU scores of 64.80%, 69.58%, 70.32%, and 73.92% with 1/16, 1/8, 1/4, and 1/2 labeled data, respectively. Full article
(This article belongs to the Topic Visual Computing and Understanding: New Developments and Trends)
►▼ Show Figures

Figure 1

20 pages, 14718 KB  
Article
Selective Trace Mix: A New Processing Tool to Enhance Seismic Imaging of Complex Subsurface Structures
by Mohamed Rashed, Nassir Al-Amri, Riyadh Halawani, Ali Atef and Hussein Harbi
J. Imaging 2026, 12(3), 124; https://doi.org/10.3390/jimaging12030124 - 12 Mar 2026
Viewed by 276
Abstract
In seismic imaging, the trace mixing process involves merging neighboring traces in seismic data to enhance the signal-to-noise ratio and improve the continuity and spatial coherence of seismic data. In regions with complex subsurface structures, current trace mix filters are often ineffective as [...] Read more.
In seismic imaging, the trace mixing process involves merging neighboring traces in seismic data to enhance the signal-to-noise ratio and improve the continuity and spatial coherence of seismic data. In regions with complex subsurface structures, current trace mix filters are often ineffective as they introduce artifacts that reduce interpretability and obscure the signatures of important structures, such as faults and folds. We introduce the selective trace mix as a novel, data-dependent filter. This filter enhances amplitude consistency, spatial coherence, and the definition of reflections, while it preserves complex structures and maintains their clarity. Selective trace mix uses sequential steps of evaluation, referencing, exclusion, weighting, and normalization of all samples within the filter operator. As a result, selective trace mix is a temporally and spatially variable, data-dependent filter. The filter’s effectiveness is validated using both synthetic and real field seismic data. Synthetic data is a portion of the Marmousi seismic model, while real data include land and marine seismic datasets imaging complex subsurface fault/fold structures. When compared to three of the commonly used conventional filters, the selective trace mix yields far better results in terms of horizon integrity and fault clarity. Full article
(This article belongs to the Topic Image Processing, Signal Processing and Their Applications)
►▼ Show Figures

Figure 1

29 pages, 2885 KB  
Article
Influence of Off-Centre Positioning, Scan Direction, and Localiser Projection Angle on Organ-Specific Radiation Doses in Low-Dose Chest CT: A Simulation Study Across Four Scanner Models
by Louise D’hondt, Claudia Haentjens, Pieter-Jan Kellens, Annemiek Snoeckx and Klaus Bacher
J. Imaging 2026, 12(3), 123; https://doi.org/10.3390/jimaging12030123 - 11 Mar 2026
Viewed by 415
Abstract
With the considerable number of low-dose CT examinations performed in lung cancer screening, variations in participant positioning, scan direction, or localiser angle are likely to occur in practice. These variations are known to affect automatic tube current modulation (ATCM) operation, yet organ-specific dose [...] Read more.
With the considerable number of low-dose CT examinations performed in lung cancer screening, variations in participant positioning, scan direction, or localiser angle are likely to occur in practice. These variations are known to affect automatic tube current modulation (ATCM) operation, yet organ-specific dose implications across CT models remain unknown. Therefore, this simulation study systematically characterised the effect of the aforementioned variations. Using the Alderson RANDO phantom, ATCM profiles were established on CT scanners from four major vendors (GE, Siemens, Canon, Philips) after introducing vertical and lateral mispositioning, craniocaudal and caudocranial scan directions, and varying localiser projection angles. Additionally, off-centre positioning and scan direction changes preceded by either a single posteroanterior (PA) or dual (PA+lateral) localiser were evaluated. Doses to the lungs, heart, thyroid, liver, and breasts were calculated from Monte Carlo simulations of each setup for 32 patient-specific voxel models. The results demonstrate statistically significant and scanner-dependent dose variations. PA localisers generally produced the highest organ doses. However, on the Philips system, organ dose increases of at least 50% were observed after the lateral projection angle. GE and Siemens scanners showed pronounced dose increases following downward mispositioning with a single PA localiser (18–50% and 5–25%, respectively), an effect largely mitigated by adding a lateral localiser. Canon and Philips scanners exhibited generally stable ATCM behaviour after vertical off-centring, although Canon showed notable dose increases upon lateral mispositioning, with dose increases up to 37.5% and 34% after a single PA or dual localiser, respectively. Variations in scan direction displayed highly model- and organ-dependent effects. Dose deviations were largely mitigated after dual localisers for the GE, Canon, and Philips scanner types. Here, organ dose differences were within an absolute range of 10%, indicating that a change in scan direction preceded by a dual localiser can reduce extreme dose deviations. Remarkably, no significant difference was observed solely for the Siemens scanner when combined with a dual localiser, as lung, heart, breast, and liver doses remained significantly (between 20 and 35%) lower when scanning craniocaudally, whereas the thyroid dose in this setup remained considerably higher (up to 20% mean increase). Ultimately, findings indicate that seemingly minor protocol deviations can lead to significant underestimation of anticipated organ-specific doses associated with lung cancer screening. Scanner-specific optimisation, supported by medical physics expertise, is therefore essential. Full article
(This article belongs to the Section Medical Imaging)
►▼ Show Figures

Figure 1

26 pages, 8878 KB  
Article
A Spectrally Compatible Pseudo-Panchromatic Intensity Reconstruction for PCA-Based UAS RGB–Multispectral Image Fusion
by Dimitris Kaimaris
J. Imaging 2026, 12(3), 122; https://doi.org/10.3390/jimaging12030122 - 11 Mar 2026
Viewed by 223
Abstract
The paper presents a method for generating a pseudo-panchromatic (PPAN) orthophotomosaic that is spectrally compatible with the multispectral (MS) orthophotomosaic, and it targets the fusion of unmanned aircraft system (UAS) RGB–MS orthophotomosaics when no true panchromatic band is available. In typical UAS imaging [...] Read more.
The paper presents a method for generating a pseudo-panchromatic (PPAN) orthophotomosaic that is spectrally compatible with the multispectral (MS) orthophotomosaic, and it targets the fusion of unmanned aircraft system (UAS) RGB–MS orthophotomosaics when no true panchromatic band is available. In typical UAS imaging systems, RGB and multispectral sensors operate independently and exhibit different spectral responses and spatial resolutions, making the construction of a spectrally compatible substitution intensity a critical challenge for component substitution fusion. The conventional RGB-derived PPAN preserves spatial detail but is constrained by RGB–MS spectral incompatibility, expressed as reduced corresponding-band similarity. The proposed hybrid intensity (PPANE) increases the mean corresponding-band correlation from 0.842 (PPANA) to 0.928 (PPANE) and reduces the across-site mean SAM from 5.782° to 4.264°, while maintaining spatial sharpness comparable to the RGB-derived intensity. It is proposed that the PPANE orthophotomosaic be produced as a hybrid intensity (single band) image. Specifically, a multispectral-visible-derived intensity is resampled onto the RGB grid and statistically integrated with RGB spatial detail, followed by mild high-frequency enhancement to produce the final PPANE orthophotomosaic. Principal Component Analysis (PCA) fusion is applied to seven archaeological sites in Northern Greece. Spectral quality is evaluated on the MS grid using band-wise (corresponding-band) correlation and the Spectral Angle Mapper (SAM), while the spatial sharpness of the fused NIR orthophotomosaic is assessed using Tenengrad and Laplacian variance. The PPANE orthophotomosaic consistently increases correlations relative to PPANA (especially in Red Edge/NIR) and reduces the mean site-mean SAM. PPANC yields the lowest SAM but also the lowest spatial sharpness/clarity, whereas PPANE maintains spatial sharpness/clarity comparable to PPANA, supporting a balance between spectral consistency and spatial detail, as also confirmed through comparative evaluation against established component substitution fusion methods. The approach is reproducible and avoids full histogram matching; instead, it relies on explicitly defined linear standardization steps (mean–std normalization) and controlled spatial sharpening, and performs consistently across different scenes. Full article
(This article belongs to the Section Color, Multi-spectral, and Hyperspectral Imaging)
►▼ Show Figures

Figure 1

30 pages, 2375 KB  
Article
Deep Learning Based Computer-Aided Detection of Prostate Cancer Metastases in Bone Scintigraphy: An Experimental Analysis
by Eslam Jabali, Omar Almomani, Louai Qatawneh, Sinan Badwan, Yazan Almomani, Mohammad Al-soreeky, Alia Ibrahim and Natalie Khalil
J. Imaging 2026, 12(3), 121; https://doi.org/10.3390/jimaging12030121 - 11 Mar 2026
Viewed by 821
Abstract
Bone scintigraphy is a widely available and cost-effective modality for detecting skeletal metastases in prostate cancer, yet visual interpretation can be challenging due to heterogeneous uptake patterns, benign mimickers, and a high reporting workload, motivating robust computer-aided decision support. In this study, we [...] Read more.
Bone scintigraphy is a widely available and cost-effective modality for detecting skeletal metastases in prostate cancer, yet visual interpretation can be challenging due to heterogeneous uptake patterns, benign mimickers, and a high reporting workload, motivating robust computer-aided decision support. In this study, we present an experimental evaluation of fourteen convolutional neural network (CNN) architectures for binary metastasis classification in planar bone scintigraphy using a unified protocol. Fourteen models, CNN (baseline), AlexNet, VGG16, VGG19, ResNet18, ResNet34, ResNet50, ResNet50-attention, DenseNet121, DenseNet169, DenseNet121-attention, WideResNet50_2, EfficientNet-B0, and ConvNeXt-Tiny, were trained and tested on 600 scan images (300 normal, 300 metastatic) from the Jordanian Royal Medical Services under identical preprocessing and augmentation with stratified five-fold cross-validation. We report mean ± SD for AUC-ROC, accuracy, precision, sensitivity (recall), F1-score, specificity, and Cohen’s κ, alongside calibration via the Brier score and deployment indicators (parameters, FLOPs, model size, and inference time). DenseNet121 achieved the best overall balance of diagnostic performance and reliability, reaching AUC-ROC 96.0 ± 1.2, accuracy 89.2 ± 2.2, sensitivity 83.7 ± 3.4, specificity 94.7 ± 2.2, F1-score 88.5 ± 2.5, κ = 0.783 ± 0.045, and the strongest calibration (Brier 0.080 ± 0.013), with stable fold-to-fold behaviour. DenseNet121-attention produced the highest AUC-ROC (96.3 ± 1.1) but exhibited greater variability in specificity, indicating less consistent false-alarm control. Complexity analysis supported DenseNet121 as deployable (~7.0 M parameters, ~26.9 MB, ~92 ms/image), whereas heavier models yielded only limited additional clinical value. These results support DenseNet121 as a reliable backbone for automated metastasis detection in planar scintigraphy, with future work focusing on external validation, threshold optimisation, interpretability, and model compression for clinical adoption. Full article
(This article belongs to the Section AI in Imaging)
►▼ Show Figures

Figure 1

11 pages, 2216 KB  
Article
Decoding the Heart Through Computed Tomography: Early Cardiomyopathy Detection Using Ensemble-Based Segmentation and Radiomics
by Theodoros Tsampras, Alexios Antonopoulos, Theodora Karamanidou, Georgios Kalykakis, Konstantinos Tsioufis and Charalambos Vlachopoulos
J. Imaging 2026, 12(3), 120; https://doi.org/10.3390/jimaging12030120 - 10 Mar 2026
Viewed by 264
Abstract
Diagnosis of cardiomyopathies often depends on overt phenotypic manifestations, delaying patient management. This underscores the need for population-level opportunistic screening tools using clinically indicated CT scans to detect subclinical myocardial disease. This study developed an Ensemble Machine Learning (ML) model to automatically segment [...] Read more.
Diagnosis of cardiomyopathies often depends on overt phenotypic manifestations, delaying patient management. This underscores the need for population-level opportunistic screening tools using clinically indicated CT scans to detect subclinical myocardial disease. This study developed an Ensemble Machine Learning (ML) model to automatically segment the left ventricular myocardium from CT data and estimate the probability of underlying myocardial disease using radiomic feature analysis. A total of 60 CT scans (~12,000 images) were used to train ML models for left ventricular myocardium segmentation, including scans from both healthy individuals and patients with myocardial disease. A novel Ensemble model was developed and externally validated on 10 independent CT scans. Subsequently, 100 unseen CT scans were segmented manually and automatically for radiomic feature analysis. After removing highly correlated features through intra-class variation and correlation filtering, the refined dataset was used for model training and testing. Key predictive features were identified, and model performance was evaluated. The four best-performing models (Unet++, ED w/ASC, FPN, and TresUNET) were combined to form an Ensemble model, achieving a final DICE score of 0.882 after hyperparameter optimization. External validation yielded a DICE score of 0.907. Radiomic feature analysis identified 15 key predictors of myocardial disease in both manual and automatic segmentation datasets. The model demonstrated strong performance in detecting underlying myocardial disease, with AUCs of 0.85 and 0.8, respectively. This study presents a fully automated CT-based framework for LV myocardial segmentation and radiomic phenotyping that accurately estimates the probability of underlying myocardial disease. The model demonstrates strong generalizability across different CT protocols and highlights the potential role of AI-driven CT analysis for early, non-invasive cardiomyopathy screening at a population level. Full article
(This article belongs to the Special Issue Advances and Challenges in Cardiovascular Imaging)
►▼ Show Figures

Figure 1

22 pages, 1747 KB  
Review
Talking Head Generation Through Generative Models and Cross-Modal Synthesis Techniques
by Hira Nisar, Salman Masood, Zaki Malik and Adnan Abid
J. Imaging 2026, 12(3), 119; https://doi.org/10.3390/jimaging12030119 - 10 Mar 2026
Viewed by 574
Abstract
Talking Head Generation (THG) is a rapidly advancing field at the intersection of computer vision, deep learning, and speech synthesis, enabling the creation of animated human-like heads that can produce speech and express emotions with high visual realism. The core objective of THG [...] Read more.
Talking Head Generation (THG) is a rapidly advancing field at the intersection of computer vision, deep learning, and speech synthesis, enabling the creation of animated human-like heads that can produce speech and express emotions with high visual realism. The core objective of THG systems is to synthesize coherent and natural audio–visual outputs by modeling the intricate relationship between speech signals, facial dynamics, and emotional cues. These systems find widespread applications in virtual assistants, interactive avatars, video dubbing for multilingual content, educational technologies, and immersive virtual and augmented reality environments. Moreover, the development of THG has significant implications for accessibility technologies, cultural preservation, and remote healthcare interfaces. This survey paper presents a comprehensive and systematic overview of the technological landscape of Talking Head Generation. We begin by outlining the foundational methodologies that underpin the synthesis process, including generative adversarial networks (GANs), motion-aware recurrent architectures, and attention-based models. A taxonomy is introduced to organize the diverse approaches based on the nature of input modalities and generation goals. We further examine the contributions of various domains such as computer vision, speech processing, and human–robot interaction, each of which plays a critical role in advancing the capabilities of THG systems. The paper also provides a detailed review of datasets used for training and evaluating THG models, highlighting their coverage, structure, and relevance. In parallel, we analyze widely adopted evaluation metrics, categorized by their focus on image quality, motion accuracy, synchronization, and semantic fidelity. Operating parameters such as latency, frame rate, resolution, and real-time capability are also discussed to assess deployment feasibility. Special emphasis is placed on the integration of generative artificial intelligence (GenAI), which has significantly enhanced the adaptability and realism of talking head systems through more powerful and generalizable learning frameworks. Full article
(This article belongs to the Special Issue AI-Driven Multimodal Image and Video Processing: Advances and Applications)
►▼ Show Figures

Figure 1

26 pages, 23966 KB  
Article
ClearScope: A Fully Integrated Light-Sheet Theta Microscope for Sub-Micron-Resolution Imaging Without Lateral Size Constraints
by Matthew G. Fay, Peter J. Lang, David S. Denu, Nathan J. O’Connor, Benjamin Haydock, Jeffrey Blaisdell, Nicolas Roussel, Alissa Wilson, Sage R. Aronson, Veronica Pessino, Paul J. Angstman, Cheng Gong, Tanvi Butola, Orrin Devinsky, Jayeeta Basu, Raju Tomer and Jacob R. Glaser
J. Imaging 2026, 12(3), 118; https://doi.org/10.3390/jimaging12030118 - 10 Mar 2026
Cited by 1 | Viewed by 777
Abstract
Three-dimensional (3D) ex vivo imaging of cleared tissue from intact brains from animal models, human brain surgical specimens, and large postmortem human and non-human primate brain specimens is essential for understanding physiological neural connectivity and pathological alterations underlying neurological and neuropsychiatric disorders. Contemporary [...] Read more.
Three-dimensional (3D) ex vivo imaging of cleared tissue from intact brains from animal models, human brain surgical specimens, and large postmortem human and non-human primate brain specimens is essential for understanding physiological neural connectivity and pathological alterations underlying neurological and neuropsychiatric disorders. Contemporary light-sheet microscopy enables rapid, high-resolution imaging of large, cleared samples but is limited by the orthogonal arrangement of illumination and detection optics, which constrains specimen size. Light-sheet theta microscopy (LSTM) overcomes this limitation by employing two oblique illumination paths while maintaining a perpendicular detection geometry. Here, we report the development of a next-generation, fully integrated and user-friendly LSTM system that enables uniform subcellular-resolution imaging (with subcellular resolution determined by the lateral performance of the system) throughout large specimens without constraining lateral (XY) dimensions. The system provides a seamless workflow encompassing image acquisition, data storage, pre- and post-processing, enhancement and quantitative analysis. Performance is demonstrated by high-resolution 3D imaging of intact mouse brains and human brain samples, including complete downstream analyses such as digital neuron tracing, vascular reconstruction and design-based stereological analysis. This enhanced and accessible LSTM implementation enables rapid quantitative mapping of molecular and cellular features in very large biological specimens. Full article
(This article belongs to the Section Neuroimaging and Neuroinformatics)
►▼ Show Figures

Figure 1

17 pages, 314 KB  
Review
The Evolving Role of Radiation Therapy Technologists in Head and Neck Cancer: A Narrative Review and Operational Framework
by Andrea Lastrucci, Ilaria Morelli, Nicola Iosca, Isacco Desideri, Eva Serventi, Yannick Wandael, Carlotta Becherini, Viola Salvestrini, Vittorio Miele, Renzo Ricci, Lorenzo Livi, Pierluigi Bonomo and Daniele Giansanti
J. Imaging 2026, 12(3), 117; https://doi.org/10.3390/jimaging12030117 - 10 Mar 2026
Viewed by 329
Abstract
Head and neck cancer (HNC) management requires highly coordinated multidisciplinary care. Radiation Therapy Technologists (RTTs) have increasingly expanded their role beyond technical execution, contributing to patient positioning, treatment delivery, monitoring, and supportive care. This narrative review integrates evidence from published literature with structured [...] Read more.
Head and neck cancer (HNC) management requires highly coordinated multidisciplinary care. Radiation Therapy Technologists (RTTs) have increasingly expanded their role beyond technical execution, contributing to patient positioning, treatment delivery, monitoring, and supportive care. This narrative review integrates evidence from published literature with structured experiential insights collected through focus group discussions with RTTs and other multidisciplinary team (MDT) members. The resulting conceptual and operational framework highlights RTT contributions across the radiotherapy pathway, including adaptive planning, workflow coordination, and patient-centered interventions, supported by imaging and artificial intelligence (AI) tools for predictive modeling and treatment optimization. By facilitating communication, monitoring anatomical and functional changes, and integrating AI-informed insights, RTTs support timely interventions, reduce treatment interruptions, and enhance treatment safety and precision. Structured training, formal recognition of advanced practice roles, and interprofessional collaboration are key to maximizing the impact of RTTs in HNC care. This review provides a practical roadmap for institutions, professional societies, and research initiatives to support the evolution of RTT roles in complex radiotherapy settings. Full article
(This article belongs to the Topic Artificial Intelligence in Public Health: Current Trends and Future Possibilities, 2nd Edition)
18 pages, 2888 KB  
Article
Assessing RGB Color Reliability via Simultaneous Comparison with Hyperspectral Data on Pantone® Fabrics
by Cindy Lorena Gómez-Heredia, Jose David Ardila-Useda, Andrés Felipe Cerón-Molina, Jhonny Osorio-Gallego and Jorge Andrés Ramírez-Rincón
J. Imaging 2026, 12(3), 116; https://doi.org/10.3390/jimaging12030116 - 10 Mar 2026
Viewed by 509
Abstract
Accurate color property measurements are critical for advancing artificial vision in real-time industrial applications. RGB imaging remains highly applicable and widely used due to its practicality, accessibility, and high spatial resolution. However, significant uncertainties in extracting chromatic information highlight the need to define [...] Read more.
Accurate color property measurements are critical for advancing artificial vision in real-time industrial applications. RGB imaging remains highly applicable and widely used due to its practicality, accessibility, and high spatial resolution. However, significant uncertainties in extracting chromatic information highlight the need to define when conventional digital images can reliably provide accurate color data. This work simultaneously compares six chromatic properties across 700 Pantone® TCX fabric samples, using optical data acquired simultaneously from both hyperspectral (HSI) and digital (RGB) cameras. The results indicate that the accurate interpretation of optical information from RGB (sRGB and REC2020) images is significantly influenced by lightness (L*) values. Samples with bright and unsaturated colors (L*> 50) reach ratio-to-performance-deviation (RPD) values above 2.5 for four properties (L*, a*, b* hab), indicating a good correlation between HSI and RGB information. Absolute color difference comparisons (∆Ea) between HSI and RGB images yield values exceeding 5.5 units for red-yellow-green samples and up to 9.0 units for blue and purple tones. In contrast, relative color differences (∆Er) comparisons show a significant decrease, with values falling below 3.0 for all lightness values, indicating the practical equivalence of both methodologies according to the Two One-Sided Test (TOST) statistical analysis. These results confirm that RGB imagery achieves reliable color consistency when evaluated against a practical reference. Full article
(This article belongs to the Special Issue Advanced Hyperspectral Imaging: Techniques, Data Analysis, and Applications)
►▼ Show Figures

Graphical abstract

19 pages, 3692 KB  
Article
Automated Processing and Deviation Analysis of 3D Pipeline Point Clouds Based on Geometric Features
by Shaofeng Jin, Kangrui Fu, Chengzhen Yang and Huanhuan Rui
J. Imaging 2026, 12(3), 115; https://doi.org/10.3390/jimaging12030115 - 9 Mar 2026
Viewed by 487
Abstract
To meet the strict non-contact measurement requirements for the assembly of aircraft engine pipelines and to overcome the limitations of the traditional three-dimensional laser scanning workflow, this study proposes an automated pipeline point cloud processing and deviation analysis framework. Through a standardized three-dimensional [...] Read more.
To meet the strict non-contact measurement requirements for the assembly of aircraft engine pipelines and to overcome the limitations of the traditional three-dimensional laser scanning workflow, this study proposes an automated pipeline point cloud processing and deviation analysis framework. Through a standardized three-dimensional laser scanning procedure, high-resolution pipeline point clouds are obtained and preprocessed. Based on the geometric characteristics of the pipeline, automated algorithms for point cloud feature segmentation, axis extraction, and model registration are developed. Particularly, the three-dimensional extended Douglas–Peucker (DP) algorithm is introduced to achieve efficient point cloud downsampling while retaining necessary geometric and structural features. These algorithms are fully integrated into a unified software platform, supporting one-click operation, and can automatically analyze and obtain five key types of pipeline deviations: angular deviation, radial deviation, axial deviation, roundness error, and diameter error. The platform also provides intuitive visualization effects and comprehensive report generation functions to facilitate quantitative inspection and analysis. Test results show that the proposed method significantly improves the processing efficiency and measurement reliability of complex pipeline systems. The developed framework provides a powerful practical solution for the automated geometric inspection of aircraft engine pipelines and lays a solid foundation for subsequent quality assessment tasks. Full article
(This article belongs to the Special Issue Intelligent 3D Vision: Reconstruction, Understanding, Generative Modeling, and Applications)
►▼ Show Figures

Figure 1

24 pages, 8735 KB  
Article
Evaluation of High Dynamic Range Imaging Methods for Luminance Measurements
by Lou Gevaux, Alejandro Ferrero, Alice Dupiau, Ángela Sáez, Markos Antonopoulos and Constantinos Bouroussis
J. Imaging 2026, 12(3), 114; https://doi.org/10.3390/jimaging12030114 - 9 Mar 2026
Viewed by 446
Abstract
Imaging luminance measurement is increasingly used in lighting applications, but the limited dynamic range of camera sensors requires using high dynamic range (HDR) imaging methods for evaluating scenes with large luminance contrasts. This work aims at investigating how parameters of HDR imaging techniques [...] Read more.
Imaging luminance measurement is increasingly used in lighting applications, but the limited dynamic range of camera sensors requires using high dynamic range (HDR) imaging methods for evaluating scenes with large luminance contrasts. This work aims at investigating how parameters of HDR imaging techniques may impact luminance measurement accuracy, using a numerical evaluation. A numerical simulation framework based on a digital twin of an imaging system and synthetic high-contrast luminance scenes is used to introduce controlled systematic error sources and quantify their impact on HDR luminance accuracy. The results support the identification of HDR approaches most suitable for producing luminance measurements traceable to the International System of Units (SI). Full article
(This article belongs to the Section Computational Imaging and Computational Photography)
►▼ Show Figures

Figure 1

15 pages, 1398 KB  
Review
A Taxonomy of Six Perceptual Cues Underlying Photorealism in 3D-Rendered Architectural Scenes: A Cue-Based Narrative Review
by Matija Grašić, Andrija Bernik and Vladimir Cviljušac
J. Imaging 2026, 12(3), 113; https://doi.org/10.3390/jimaging12030113 - 8 Mar 2026
Viewed by 313
Abstract
Perceived photorealism in architectural 3D rendering is not determined solely by physical accuracy or rendering complexity but also by a limited set of visual cues that observers rely on when judging realism. This literature review synthesizes findings from 41 peer-reviewed studies spanning perception [...] Read more.
Perceived photorealism in architectural 3D rendering is not determined solely by physical accuracy or rendering complexity but also by a limited set of visual cues that observers rely on when judging realism. This literature review synthesizes findings from 41 peer-reviewed studies spanning perception science, computer graphics, and immersive visualization, with the aim of identifying the cues that most strongly contribute to perceived photorealism in rendered scenes. Convergent evidence from psychophysical experiments, user studies in virtual and augmented reality, and rendering-oriented analyses indicate that six cue categories consistently dominate realism judgments. Across the reviewed literature, realism judgments depend less on scene complexity or the number of visual elements and more on the consistency and plausibility of these cues for supporting inferences about shape, material, and spatial layout. These findings suggest that photorealism emerges from the alignment of the rendered image structure with perceptual expectations learned from real-world visual experience. The implications for architectural visualization workflows and directions for future research on cue interactions and perceptual thresholds are discussed. Full article
(This article belongs to the Section Computational Imaging and Computational Photography)
►▼ Show Figures

Figure 1

28 pages, 56643 KB  
Article
Endo-DET: A Domain-Specific Detection Framework for Multi-Class Endoscopic Disease Detection
by Yijie Lu, Yixiang Zhao, Qiang Yu, Wei Shao and Renbin Shen
J. Imaging 2026, 12(3), 112; https://doi.org/10.3390/jimaging12030112 - 6 Mar 2026
Viewed by 409
Abstract
Gastrointestinal cancers account for roughly a quarter of global cancer incidence, and early detection through endoscopy has proven effective in reducing mortality. Multi-class endoscopic disease detection, however, faces three persistent challenges: feature redundancy from non-pathological content, severe illumination inconsistency across imaging modalities, and [...] Read more.
Gastrointestinal cancers account for roughly a quarter of global cancer incidence, and early detection through endoscopy has proven effective in reducing mortality. Multi-class endoscopic disease detection, however, faces three persistent challenges: feature redundancy from non-pathological content, severe illumination inconsistency across imaging modalities, and extreme scale variability with blurry boundaries. This paper introduces Endo-DET, a domain-specific detection framework addressing these challenges through three synergistic components. The Adaptive Lesion-Discriminative Filtering (ALDF) module achieves lesion-focused attention via sparse simplex projection, reducing complexity from O(N2) to O(αN2). The Global–Local Illumination Modulation Neck (GLIM-Neck) enables illumination-aware multi-scale fusion through four cooperative mechanisms, maintaining stable performance across white-light endoscopy, narrow-band imaging, and chromoendoscopy. The Lesion-aware Unified Calibration and Illumination-robust Discrimination (LUCID) module uses dual-stream reciprocal modulation to integrate boundary-sensitive textures with global semantics while suppressing instrument artifacts. Experiments on EDD2020, Kvasir-SEG, PolypGen2021, and CVC-ClinicDB show that Endo-DET improves mAP50-95 over the DEIM baseline by 5.8, 10.8, 4.1, and 10.1 percentage points respectively, with mAP75 gains of 6.1, 10.3, 6.8, and 9.3 points, and Recall50-95 improvements of 10.9, 12.1, 11.1, and 11.5 points. Running at 330 FPS with TensorRT FP16 optimization, Endo-DET achieves consistent cross-dataset improvements while maintaining real-time capability, providing a methodological foundation for clinical computer-aided diagnosis. Full article
(This article belongs to the Topic Machine Learning and Deep Learning in Medical Imaging)
►▼ Show Figures

Figure 1

Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-50
Previous Issue
Next Issue
Back to TopTop