Mixed-reality head-mounted display in cranial neurosurgery: A proof-of-concept study DOI Creative Commons
Lukas Andereggen, Gwendoline Boillat,

Joshua Haegler

et al.

Brain Hemorrhages, Journal Year: 2024, Volume and Issue: unknown

Published: July 1, 2024

Mixed-reality (MR) head-mounted displays (HMD) offer virtual augmentations registered with real objects, allowing for direct patient-centered lesion visualization. In contrast to other surgical subspecialties, however, the application of MR in neurosurgery remains poor. this proof-of-concept study, we aimed at evaluating applicability, educational value, and accuracy HMD as compared standard neuronavigation (SN) planning treatment patients undergoing neurovascular tumor surgeries. A 3D hologram patient's anatomy was generated from conventional CT scan, MRI, and/or rotational angiography (3D-RA), integrated into HMD. The participating surgeons completed a standardized questionnaire, which evaluated SN, detail visualization benefits limitations hologram. Eight consecutive (n = 4) or pathologies were selected MR. mean (±SD) setup time significantly longer than SN (8.3 ± 1.5 min vs. 4.8 1.3 min; p < 0.001), independent pathology applied (i.e., tumor: 8.0 2.0 4.3 1.3, 0.02, vascular: 8.7 0.9 5.4 1.1; 0.001). Surgeons wearing succeeded moving respective operators' angles identifying shape configuration lesion. device superior regard on account its improved spatial awareness. current method is however limited representation small perforators bony involvement tumors. may become valuable tool preoperative planning, education guidance complex procedures patients, yet further development necessary improve clinical applicability

Language: Английский

Modified U-Net with attention gate for enhanced automated brain tumor segmentation DOI
Shoffan Saifullah, Rafał Dreżewski, Anton Yudhana

et al.

Neural Computing and Applications, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 2, 2025

Language: Английский

Citations

1
RenalSegNet: automated segmentation of renal tumor, veins, and arteries in contrast-enhanced CT scans DOI Creative Commons
Rashid Khan, Chao Chen, Asim Zaman

et al.

Complex & Intelligent Systems, Journal Year: 2025, Volume and Issue: 11(2)

Published: Jan. 7, 2025

Renal carcinoma is a frequently seen cancer globally, with laparoscopic partial nephrectomy (LPN) being the primary form of treatment. Accurately identifying renal structures such as kidneys, tumors, veins, and arteries on CT scans crucial for optimal surgical preparation However, automatic segmentation these remains challenging due to kidney's complex anatomy variability imaging data. This study presents RenalSegNet, novel deep-learning framework automatically segmenting structure in contrast-enhanced images. RenalSegNet has an innovative encoder-decoder architecture, including FlexEncoder Block efficient multivariate feature extraction MedSegPath mechanism advanced distribution fusion. Evaluated KiPA dataset, achieved remarkable performance, average dice score 86.25%, IOU 76.75%, Recall 86.69%, Precision 86.48%, HD 15.78 mm, AVD 0.79 mm. Ablation studies confirm critical roles MedFuse components achieving results. RenalSegNet's robust performance highlights its potential clinical applications offers significant advances treatment by contributing accurate preoperative planning postoperative evaluation. Future improvements model accuracy applicability will involve integrating techniques, unsupervised transformer-based approaches.

Language: Английский

Citations

0
Three-stage registration pipeline for dynamic lung field of chest X-ray images based on convolutional neural networks DOI Creative Commons

Yingjian Yang,

Jie Zheng, Peng Guo

et al.

Frontiers in Artificial Intelligence, Journal Year: 2025, Volume and Issue: 8

Published: March 12, 2025

Background The anatomically constrained registration network (AC-RegNet), which yields plausible results, has emerged as the state-of-the-art architecture for chest X-ray (CXR) images. Nevertheless, accurate lung field results may be more favored and exciting than of entire CXR images hold promise dynamic analysis in clinical practice. Objective Based on above, a model based AC-RegNet static is urgently developed to register these fields quantitative analysis. Methods This paper proposes fully automatic three-stage pipeline First, mask are generated from pre-trained standard segmentation with Then, abstraction designed generate their corresponding Finally, we propose three-step training method train AC-RegNet, obtaining (AC-RegNet_V3). Results proposed AC-RegNet_V3 four basic networks achieve mean dice similarity coefficient (DSC) 0.991, 0.993, Hausdorff distance (HD) 12.512, 12.813, 12.449, 13.661, average symmetric surface (ASSD) 0.654, 0.550, 0.572, 0.564, squared (MSD) 559.098, 577.797, 548.189, 559.652, respectively. Besides, compared images, DSC been significantly improved by 7.2, 7.4, 7.4% ( p -value &lt; 0.0001). Meanwhile, HD 8.994, 8.693, 9.057, 7.845 Similarly, ASSD 4.576, 4.680, 4.658, 4.658 Last, MSD 508.936, 519.776, 517.904, 520.626 Conclusion Our demonstrated its effectiveness registration. Therefore, it could become powerful tool practice, such pulmonary airflow detection air trapping location.

Language: Английский

Citations

0
Challenges, optimization strategies, and future horizons of advanced deep learning approaches for brain lesion segmentation DOI
Asim Zaman, Mazen M. Yassin,

Irfan Mehmud

et al.

Methods, Journal Year: 2025, Volume and Issue: unknown

Published: April 1, 2025

Language: Английский

Citations

0
Hemi-diaphragm detection of chest X-ray images based on convolutional neural network and graphics DOI

Yingjian Yang,

Jie Zheng, Peng Guo

et al.

Journal of X-Ray Science and Technology, Journal Year: 2024, Volume and Issue: 32(5), P. 1273 - 1295

Published: July 12, 2024

Chest X-rays (CXR) are widely used to facilitate the diagnosis and treatment of critically ill emergency patients in clinical practice. Accurate hemi-diaphragm detection based on postero-anterior (P-A) CXR images is crucial for diaphragm function assessment provide precision healthcare these vulnerable populations.

Language: Английский

Citations

1
The Neural Frontier of Future Medical Imaging: A Review of Deep Learning for Brain Tumor Detection DOI Creative Commons
Tarek Berghout

Journal of Imaging, Journal Year: 2024, Volume and Issue: 11(1), P. 2 - 2

Published: Dec. 24, 2024

Brain tumor detection is crucial in medical research due to high mortality rates and treatment challenges. Early accurate diagnosis vital for improving patient outcomes, however, traditional methods, such as manual Magnetic Resonance Imaging (MRI) analysis, are often time-consuming error-prone. The rise of deep learning has led advanced models automated brain feature extraction, segmentation, classification. Despite these advancements, comprehensive reviews synthesizing recent findings remain scarce. By analyzing over 100 papers past half-decade (2019-2024), this review fills that gap, exploring the latest methods paradigms, summarizing key concepts, challenges, datasets, offering insights into future directions using learning. This also incorporates an analysis previous targets three main aspects: results revealed primarily focuses on Convolutional Neural Networks (CNNs) their variants, with a strong emphasis transfer pre-trained models. Other Generative Adversarial (GANs) Autoencoders, used while Recurrent (RNNs) employed time-sequence modeling. Some integrate Internet Things (IoT) frameworks or federated real-time diagnostics privacy, paired optimization algorithms. However, adoption eXplainable AI (XAI) remains limited, despite its importance building trust diagnostics. Finally, outlines opportunities, focusing image quality, underexplored techniques, expanding deeper representations model behavior recurrent expansion advance imaging

Language: Английский

Citations

1
Mixed-reality head-mounted display in cranial neurosurgery: A proof-of-concept study DOI Creative Commons
Lukas Andereggen, Gwendoline Boillat,

Joshua Haegler

et al.

Brain Hemorrhages, Journal Year: 2024, Volume and Issue: unknown

Published: July 1, 2024

Mixed-reality (MR) head-mounted displays (HMD) offer virtual augmentations registered with real objects, allowing for direct patient-centered lesion visualization. In contrast to other surgical subspecialties, however, the application of MR in neurosurgery remains poor. this proof-of-concept study, we aimed at evaluating applicability, educational value, and accuracy HMD as compared standard neuronavigation (SN) planning treatment patients undergoing neurovascular tumor surgeries. A 3D hologram patient's anatomy was generated from conventional CT scan, MRI, and/or rotational angiography (3D-RA), integrated into HMD. The participating surgeons completed a standardized questionnaire, which evaluated SN, detail visualization benefits limitations hologram. Eight consecutive (n = 4) or pathologies were selected MR. mean (±SD) setup time significantly longer than SN (8.3 ± 1.5 min vs. 4.8 1.3 min; p < 0.001), independent pathology applied (i.e., tumor: 8.0 2.0 4.3 1.3, 0.02, vascular: 8.7 0.9 5.4 1.1; 0.001). Surgeons wearing succeeded moving respective operators' angles identifying shape configuration lesion. device superior regard on account its improved spatial awareness. current method is however limited representation small perforators bony involvement tumors. may become valuable tool preoperative planning, education guidance complex procedures patients, yet further development necessary improve clinical applicability

Language: Английский

Citations

0