Advanced searches left 3/3

Brain Tumor - Astrophysics Data System

Summarized by Plex Scholar
Last Updated: 10 June 2022

* If you want to update the article please login/register

Unsupervised Deformable Image Registration with Absent Correspondences in Pre-operative and Post-Recurrence Brain Tumor MRI Scans

"Pre-operative and post-recurrence brain photographs are often needed to determine the effectiveness of brain gliomas therapy. Our results on 3D clinical data from the BraTS-Reg challenge show that our technique can increase image alignment when compared to traditional and deep learning-based registration strategies with or without cost function masking. ".

Source link: https://ui.adsabs.harvard.edu/abs/2022arXiv220603900M/abstract


mmFormer: Multimodal Medical Transformer for Incomplete Multimodal Learning of Brain Tumor Segmentation

"In this work, we present the first attempt to exploit the Transformer for multimodal brain tumor segmentation that is robust to any combinatorial subset of available modalities. " A novel multimodal learning Transformer with three main features; the hybrid encoders that link a convolutional encoder and an intramodal Transformer for both local and global context modeling within each modality; and an intermodal Transformer that converts the long-range correlations between modalities for modality-invariant features with global semantics corresponding to tumor region; and a novel multimodal Transformer for incomplete multimodal learning; and a decoder that performs multimodal transformation; and a The latest mmFormer outperforms the state-of-the-art techniques for incomplete multimodal brain tumor segmentation on nearly all subsets of incomplete modalities, with an average 19. 0 percent increase in tumor segmentation with only one available modality.

Source link: https://ui.adsabs.harvard.edu/abs/2022arXiv220602425Z/abstract


Examining the behaviour of state-of-the-art convolutional neural networks for brain tumor detection with and without transfer learning

"Distinguishing normal from malignant and determining the tumor type are two key elements of brain tumor diagnosis. " The validation set for all the models is the same, with train results being 60% for validation while the rest is 40% for validation. With the accuracy of 99. 75% and 98. 61% for the multi-class dataset, the EfficientNet-B5 architecture outperforms all the state-of-art models in this study, with the accuracy reaching 99. 75% and 98. 61% for the multi-class dataset. "This research also shows the behavior of validation loss convergence in various weight initialization techniques. ".

Source link: https://ui.adsabs.harvard.edu/abs/2022arXiv220601735A/abstract


Lipid and metabolite profiles of human brain tumors by desorption electrospray ionization-MS

"Brain tumors can be a major source of morbidity and mortality. " Mass spectra can be used to distinguish between healthy and diseased tissues by comparing peak intensities using multivariate results. These findings will provide surgeons with near-real-time pathologic results and help guide the intraoperative resection of the tumor at the difficult to detect peritumoral boundaries.

Source link: https://ui.adsabs.harvard.edu/abs/2016PNAS..113.1486J/abstract


How tissue fluidity influences brain tumor progression

"Glioblastoma is the most common and aggressive malignant brain tumor in adults," has one of the most dismal prognoses of all cancers. Neurotumor fluidity is peculiar in that it decreases as water content rises, allowing GBM to infiltratively thumb into normal surrounding brain tissue, enabling GBM to infiltrating brain tissue. ".

Source link: https://ui.adsabs.harvard.edu/abs/2020PNAS..117..128S/abstract


Ambient mass spectrometry for the intraoperative molecular diagnosis of human brain tumors

"The primary aim of brain tumor surgery is to increase tumor removal while still preserving brain function. " We have developed a method to quickly analyze and classify brain tumors based on lipid data obtained by desorption electrospray ionization mass spectrometry. Based on 36 glioma and 19 meningiomas samples, a classification system was developed to distinguish gliomas and meningiomas in this research. The classifier was tested and findings were confirmed for intraoperative use by analyzing and diagnosing tissue sections from 32 surgical specimens obtained from five research subjects who underwent brain tumor resection. Our research shows that DESI-MS technology can help to determine the histology type of brain tumors. Our results show that ambient mass spectrometry can be used to guide brain tumor surgery by providing fast diagnosis and tumor margin estimation in near-real time.

Source link: https://ui.adsabs.harvard.edu/abs/2013PNAS..110.1611E/abstract


Is this good enough? On expert perception of brain tumor segmentation quality

"Using quantitative measures such as the Dice score and Hausdorff distance, Deep Learning segmentation algorithms' performance is routinely established. " A thorough review of segmentation quality by experts would aid in the successful collaboration of health care professionals with DL segmentation algorithms. Here, we present the results of a research on expert quality perception of brain tumor segmentation of brain MR images generated by a DL segmentation scheme. On a scale from 1 to 4, eight expert medical professionals were asked to rate segmentations' quality on a scale from 1 to 4. We conclude that segmentation quality ratings for manual brain tumor segmentation are vulnerable to variability, similar to inter-rater variability observed for manual brain tumor segmentation, due to the confusion of tumor boundaries and individual perplexity. A better understanding of expert quality perception is expected to help with the development of more personalised DL schemes for integration into the clinical workflow. ".

Source link: https://ui.adsabs.harvard.edu/abs/2022SPIE12035E..0PH/abstract


Closed-loop control of targeted ultrasound drug delivery across the blood-brain/tumor barriers in a rat glioma model

"Focused ultrasound is now the only method of reversible blood-brain barrier disruption for targeted drug delivery without incision or radiation. " Here, a closed-loop, real-time control system is shown to be capable of retaining stable microbubble oscillations at a preset level while minimizing microbubble behavior that can result in vascular damage. ".

Source link: https://ui.adsabs.harvard.edu/abs/2017PNAS..11410281S/abstract


Weakly supervised brain tumor segmentation via semantic affinity deep neural network

"Image segmentation tasks are considered costly. " Medical image segmentation jobs are prohibitively costly, considering both the staff and the error margin. Researchers were unlikely to develop deep learning and machine learning solutions on new datasets that were not annotated by professional staff due to a lack of supervision. Example segmentation is used in this paper by weak supervision to develop a deep neural network. The Multimodal Brain Tumor Segmentation Challenge 3D MRI scans were used for this study. Each pixel is converted to a 32-dimensional vector expressing a semantic identity, according to this network. Lastly, we determine the semantic distance between suspected lesion points and the entire map in order to categorize input data. Over three test sets of 38 patients each, we have an average Dice score of 0. 73.

Source link: https://ui.adsabs.harvard.edu/abs/2022SPIE12032E..3BY/abstract


Intraoperative mass spectrometry mapping of an onco-metabolite to guide brain tumor surgery

"We report that a validated molecular marker [u20142-hydroxyglutarate] obtained from isocitrate dehydrogenase 1 mutant gliomasu2014 can be quickly detected from tumors using a simple form of ambient MS that does not require sample preparation. " To show that desorption electrospray ionization MS could be used to find residual tumors in the patient that would have been left behind," we use the Advanced Multimodality Image Guided Operating Suite at Brigham and Women's Hospital.

Source link: https://ui.adsabs.harvard.edu/abs/2014PNAS..11111121S/abstract

* Please keep in mind that all text is summarized by machine, we do not bear any responsibility, and you should always check original source before taking any actions

* Please keep in mind that all text is summarized by machine, we do not bear any responsibility, and you should always check original source before taking any actions