MDFU-Net: Multiscale dilated features up-sampling network for accurate segmentation of tumor from heterogeneous brain data DOI Creative Commons
Haseeb Sultan, Muhammad Owais, Se Hyun Nam

et al.

Journal of King Saud University - Computer and Information Sciences, Journal Year: 2023, Volume and Issue: 35(5), P. 101560 - 101560

Published: April 20, 2023

The existing methods for accurate brain tumor (BT) segmentation based on homogeneous datasets show significant performance degradation in actual clinical applications and lacked heterogeneous data analysis. To address these issues, we designed a deep learning-based multiscale dilated features up-sampling network (MDFU-Net) BT from data. Our method primarily uses the strength of (MDF) inside encoder module to improve performance. For final segmentation, simple yet effective decoder is process dense spatial MDF. experiments, our MDFU-Net trained one dataset tested with another environment, showing quantitative results Dice similarity coefficient (DC) 62.66%, intersection over union (IoU) 56.96%, specificity (Spe) 99.29%, sensitivity (Sen) 51.98%, which were higher than those state-of-the-art methods. There are several reasons lower values evaluation metrics dataset, including change characteristics different MRI modalities, presence minor lesions, highly imbalanced dataset. Moreover, experimental showed that achieved DC 82.96%, IoU 74.94%, Spe 99.89%, Sen 68.05%, also system, as well data, can be advantageous radiologists medical experts.

Language: Английский

Novel research and future prospects of artificial intelligence in cancer diagnosis and treatment DOI Creative Commons
Chaoyi Zhang, Jin Xu,

Rong Tang

et al.

Journal of Hematology & Oncology, Journal Year: 2023, Volume and Issue: 16(1)

Published: Nov. 27, 2023

Research into the potential benefits of artificial intelligence for comprehending intricate biology cancer has grown as a result widespread use deep learning and machine in healthcare sector availability highly specialized datasets. Here, we review new approaches how they are being used oncology. We describe might be detection, prognosis, administration treatments introduce latest large language models such ChatGPT oncology clinics. highlight applications omics data types, offer perspectives on various types combined to create decision-support tools. also evaluate present constraints challenges applying precision Finally, discuss current may surmounted make useful clinical settings future.

Language: Английский

Citations

55
Magnetic Resonance Imaging of Primary Adult Brain Tumors: State of the Art and Future Perspectives DOI Creative Commons
Matia Martucci, Rosellina Russo,

Francesco Schimperna

et al.

Biomedicines, Journal Year: 2023, Volume and Issue: 11(2), P. 364 - 364

Published: Jan. 26, 2023

MRI is undoubtedly the cornerstone of brain tumor imaging, playing a key role in all phases patient management, starting from diagnosis, through therapy planning, to treatment response and/or recurrence assessment. Currently, neuroimaging can describe morphologic and non-morphologic (functional, hemodynamic, metabolic, cellular, microstructural, sometimes even genetic) characteristics tumors, greatly contributing diagnosis follow-up. Knowing technical aspects, strength limits each MR technique crucial correctly interpret studies address clinicians best strategy. This article aimed provide an overview assessment adult primary tumors. We started basilar conventional/morphological sequences, then analyzed, one by one, non-morphological techniques, finally highlighted future perspectives, such as radiomics artificial intelligence.

Language: Английский

Citations

36
AdvancedMRTechniques for Preoperative Glioma Characterization: Part 2 DOI Creative Commons
Gilbert Hangel, Bárbara Schmitz‐Abecassis, Nico Sollmann

et al.

Journal of Magnetic Resonance Imaging, Journal Year: 2023, Volume and Issue: 57(6), P. 1676 - 1695

Published: March 13, 2023

Preoperative clinical MRI protocols for gliomas, brain tumors with dismal outcomes due to their infiltrative properties, still rely on conventional structural MRI, which does not deliver information tumor genotype and is limited in the delineation of diffuse gliomas. The GliMR COST action wants raise awareness about state art advanced techniques gliomas possible translation. This review describes current methods, limits, applications preoperative assessment glioma, summarizing level validation different techniques. In this second part, we magnetic resonance spectroscopy (MRS), chemical exchange saturation transfer (CEST), susceptibility-weighted imaging (SWI), MRI-PET, MR elastography (MRE), MR-based radiomics applications. first part addresses dynamic susceptibility contrast (DSC) contrast-enhanced (DCE) arterial spin labeling (ASL), diffusion-weighted vessel imaging, fingerprinting (MRF). EVIDENCE LEVEL: 3. TECHNICAL EFFICACY: Stage 2.

Language: Английский

Citations

27
Artificial intelligence-based radiomics in bone tumors: Technical advances and clinical application DOI
Yichen Meng, Yue Yang, Miao Hu

et al.

Seminars in Cancer Biology, Journal Year: 2023, Volume and Issue: 95, P. 75 - 87

Published: July 26, 2023

Language: Английский

Citations

23
Advances in Neuro-Oncological Imaging: An Update on Diagnostic Approach to Brain Tumors DOI Open Access
Paniz Sabeghi, P Zaránd, Sina Zargham

et al.

Cancers, Journal Year: 2024, Volume and Issue: 16(3), P. 576 - 576

Published: Jan. 30, 2024

This study delineates the pivotal role of imaging within field neurology, emphasizing its significance in diagnosis, prognostication, and evaluation treatment responses for central nervous system (CNS) tumors. A comprehensive understanding both capabilities limitations inherent emerging technologies is imperative delivering a heightened level personalized care to individuals with neuro-oncological conditions. Ongoing research endeavors rectify some radiological modalities, aiming augment accuracy efficacy management brain review dedicated comparison critical examination latest advancements diverse modalities employed neuro-oncology. The objective investigate their respective impacts on cancer staging, prognosis, post-treatment monitoring. By providing analysis these this aims contribute collective knowledge field, fostering an informed approach care. In conclusion, outlook appears promising, sustained exploration domain anticipated yield further breakthroughs, ultimately enhancing outcomes grappling CNS

Language: Английский

Citations

12
Multitask deep learning for prediction of microvascular invasion and recurrence‐free survival in hepatocellular carcinoma based on MRI images DOI
Fang Wang, Gan Zhan, Qingqing Chen

et al.

Liver International, Journal Year: 2024, Volume and Issue: 44(6), P. 1351 - 1362

Published: March 4, 2024

Accurate preoperative prediction of microvascular invasion (MVI) and recurrence-free survival (RFS) is vital for personalised hepatocellular carcinoma (HCC) management. We developed a multitask deep learning model to predict MVI RFS using MRI scans.

Language: Английский

Citations

9
TransResUNet: Revolutionizing Glioma Brain Tumor Segmentation Through Transformer-Enhanced Residual UNet DOI Creative Commons
Novsheena Rasool, Javaid Iqbal Bhat, Niyaz Ahmad Wani

et al.

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 72105 - 72116

Published: Jan. 1, 2024

Accurate segmentation of brain tumors from MRI sequences is essential across diverse clinical scenarios, facilitating precise delineation anatomical structures and disease-affected areas. This study presents an innovative deep-learning method for segmenting glioma tumors, utilizing a hybrid architecture that combines ResNet U-Net with Transformer blocks. The proposed model adeptly encompasses both the local global contextual details present in scans. includes encoder based on extracting hierarchical features, followed by residual blocks to enhance feature representation while maintaining spatial information. Additionally, central transformer block, incorporating Multi-Head Attention mechanisms, enables modeling long-range dependencies comprehension, progressively refining interactions. To handle structural scale variations within images, skip connections are utilized during decoding phase. Transposed convolutional layers decoder upsample maps, retaining information earlier layers. A rigorous assessment model's functionality was carried out BraTS2019 dataset, employing comprehensive set evaluation metrics including accuracy, IOU score, specificity, sensitivity, dice precision. focused individual tumor classes, namely whole, core, enhancing regions. During validation, suggested demonstrated remarkable scores 0.91, 0.89, 0.84 whole tumor, core respectively, yielding impressive overall accuracy rate 98%.

Language: Английский

Citations

9
Advances of Artificial Intelligence in Clinical Application and Scientific Research of Neuro-oncology: Current Knowledge and Future Perspectives DOI Creative Commons
Yihong Zhan, Yuanyue Hao, Xiang Wang

et al.

Critical Reviews in Oncology/Hematology, Journal Year: 2025, Volume and Issue: unknown, P. 104682 - 104682

Published: March 1, 2025

Brain tumors refer to the abnormal growths that occur within brain's tissue, comprising both primary neoplasms and metastatic lesions. Timely detection, precise staging, suitable treatment, standardized management are of significant clinical importance for extending survival rates brain tumor patients. Artificial intelligence (AI), a discipline computer science, is leveraging its robust capacity information identification combination revolutionize traditional paradigms oncology care, offering substantial potential precision medicine. This article provides an overview current applications AI in tumors, encompassing technologies, their working mechanisms workflow, contributions diagnosis as well role scientific research, particularly drug innovation revealing microenvironment. Finally, paper addresses existing challenges, solutions, future application prospects. review aims enhance our understanding provide valuable insights forthcoming inquiries.

Language: Английский

Citations

1
An MRI Radiogenomic Signature to Characterize the Transcriptional Heterogeneity Associated with Prognosis and Biological Functions in Glioblastoma DOI Creative Commons
Xiaoqing Zhang, Xiaoyu Zhang, Jie Zhu

et al.

Frontiers in Bioscience-Landmark, Journal Year: 2025, Volume and Issue: 30(3)

Published: March 19, 2025

Background: The study sought to establish a radiogenomic signature evaluate the transcriptional heterogeneity that reflects prognosis and tumour-related biological functions of patients with glioblastoma. Methods: Transcriptional subclones were identified via fully unsupervised deconvolution RNA sequencing. A genomic prognostic risk score was developed from subclone proportions in development dataset (n = 532) independently verified testing 225). Multimodal magnetic resonance imaging (MRI) analysis involved feature extraction three distinct anatomical regions across four sequences. Key features selected construct predictive 99), subsequent survival conducted image 233). Results: total 8 identified, which metabolic pathway spinocerebellar ataxia independent factors for overall survival. effectively differentiated patient subgroups divergent outcomes both (p < 0.001) datasets 0.0003). Nineteen radiomic signature, these being linked hallmark cancer pathways malignant behaviours cells. predicted (hazard ratios (HR) 1.67, p 0.011). Conclusions: established characterize underlying

Language: Английский

Citations

1
MRI-based classification of IDH mutation and 1p/19q codeletion status of gliomas using a 2.5D hybrid multi-task convolutional neural network DOI Creative Commons
Satrajit Chakrabarty, Pamela LaMontagne, Joshua S. Shimony

et al.

Neuro-Oncology Advances, Journal Year: 2023, Volume and Issue: 5(1)

Published: Jan. 1, 2023

Abstract Background IDH mutation and 1p/19q codeletion status are important prognostic markers for glioma that currently determined using invasive procedures. Our goal was to develop artificial intelligence-based methods noninvasively determine molecular alterations from MRI. Methods Pre-operative MRI scans of 2648 patients were collected Washington University School Medicine (WUSM; n = 835) publicly available Brain Tumor Segmentation (BraTS; 378), LGG (n 159), Ivy Glioblastoma Atlas Project (Ivy GAP; 41), The Cancer Genome (TCGA; 461), the Erasmus Glioma Database (EGD; 774) datasets. A 2.5D hybrid convolutional neural network proposed simultaneously localize classify its by leveraging imaging features prior knowledge clinical records tumor location. models trained on 223 348 cases tasks, respectively, tested one internal (TCGA) two external (WUSM EGD) test sets. Results For IDH, best-performing model achieved areas under receiver operating characteristic (AUROC) 0.925, 0.874, 0.933 precision-recall curves (AUPRC) 0.899, 0.702, 0.853 internal, WUSM, EGD sets, respectively. 1p/19q, best AUROCs 0.782, 0.754, 0.842, AUPRCs 0.588, 0.713, those three data-splits, Conclusions high accuracy unseen data showcases generalization capabilities suggests potential perform “virtual biopsy” tailoring treatment planning overall management gliomas.

Language: Английский

Citations

21