Article
Engineering, Biomedical
Poonam Rani Verma, Ashish Kumar Bhandari
Summary: This article proposes a fully unsupervised approach to brain extraction using a cascaded loss function and leaky ReLU activation function. The brain image is enhanced before extraction, resulting in better skull stripping.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2024)
Article
Computer Science, Interdisciplinary Applications
Iman Aganj, Bruce Fischl
Summary: A new approach for medical image segmentation is proposed, which calculates the probability of all possible atlas-to-image transformations and the expected label value (ELV), avoiding the issue of local optima. This method does not require actually performing deformable registration, hence saving computational costs.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2021)
Article
Neurosciences
Henry F. J. Tregidgo, Sonja Soskic, Juri Althonayan, Chiara Maffei, Koen Van Leemput, Polina Golland, Ricardo Insausti, Garikoitz Lerma-Usabiaga, Cesar Caballero-Gaudes, Pedro M. Paz-Alonso, Anastasia Yendiki, Daniel C. Alexander, Martina Bocchetta, Jonathan D. Rohrer, Juan Eugenio Iglesias
Summary: The human thalamus plays a crucial role in various brain functions and is linked to several neurological disorders. Current neuroimaging studies focus on the volume and connectivity of specific nuclei within the thalamus rather than studying the entire structure. However, accurately identifying these nuclei based on standard in vivo structural MRI is challenging due to limited image contrast. This study presents a Bayesian segmentation algorithm that combines prior information and likelihood models from both structural and diffusion MRI to improve thalamic segmentation.
Article
Computer Science, Information Systems
J. Jayanthi, M. Kavitha, T. Jayasankar, A. Sagai Francis Britto, N. B. Prakash, Mohamed Yacin Sikkandar, C. Bharathiraja
Summary: This study focuses on improving early diagnosis of glioma using a P-BTLBO algorithm, which automatically segments a brain tumor in a given MRI image. By preprocessing, segmenting, and extracting information from the MRI images, the results show that this method outperforms other existing algorithms.
CMC-COMPUTERS MATERIALS & CONTINUA
(2021)
Article
Computer Science, Information Systems
Wangbin Ding, Lei Li, Xiahai Zhuang, Liqin Huang
Summary: Multi-atlas segmentation (MAS) is a promising framework for medical image segmentation. Traditional MAS methods face limitations in terms of available atlases and computational burden. In this work, a novel cross-modality MAS framework using deep neural networks for image registration and label fusion is proposed, demonstrating improved efficiency and segmentation performance.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Article
Computer Science, Artificial Intelligence
Junjun Jiang, Jiayi Ma, Xianming Liu
Summary: In this paper, a label noise cleansing method based on spectral-spatial graphs (SSGs) is proposed to reduce label noise in hyperspectral image (HSI) analysis. By constructing an affinity graph based on spectral and spatial similarity and optimizing it, the level of label noise is effectively decreased. Moreover, multiscale segmentation-based multilayer SSGs (MSSGs) are developed to incorporate richer spatial information. Experimental results show that the proposed method significantly enhances the classification accuracy for training data with noisy labels.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Software Engineering
Zheng Liu, Xiaopeng Xin, Zheng Xu, Weijie Zhou, Chunxue Wang, Renjie Chen, Ying He
Summary: In this paper, a robust and reliable approach for geometric feature detection on surfaces within point clouds is presented. The approach accurately captures local surface variations at different feature sizes. By defining a bilateral weighted centroid projection-based metric, surface deviations are quantified. A structure-to-detail feature perception algorithm is proposed to accurately locate geometric features of varying sizes, and tensor analysis is used to extract boundary features. Experimental results demonstrate the effectiveness and versatility of the method in identifying a wide range of geometric characteristics within point clouds.
COMPUTER-AIDED DESIGN
(2023)
Article
Computer Science, Artificial Intelligence
R. Pitchai, Ch Madhu Babu, P. Supraja, Mahesh Kumar Challa
Summary: Automatic segmentation of the tumor region from MRI images is a challenging task in medical image analysis. The proposed SS-2D ConvNet technique achieves better performance with dice scores of 91%, accuracy of 89%, specificity of 98%, and sensitivity of 87%, compared to existing methods. Convoluted Neural Networks have shown improved effectiveness in recognition tasks.
NEURAL PROCESSING LETTERS
(2021)
Article
Computer Science, Interdisciplinary Applications
Wenna Wang, Xiuwei Zhang, Yu Ma, Hengfei Cui, Rui Xia, Yanning Zhang
Summary: The paper proposes a robust discriminative label fusion method under the multi-atlas framework, integrating metric learning and graph cuts. A novel distance metric learning is introduced to enhance discriminative observation, unlike current methods with fixed distance metrics.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Article
Plant Sciences
Weikuan Jia, Jinmeng Wei, Qi Zhang, Ningning Pan, Yi Niu, Xiang Yin, Yanhui Ding, Xinting Ge
Summary: This study presents a two-stage instance segmentation method based on the optimized mask RCNN for fruit and vegetable picking robots. By using a lightweight backbone network and a boundary patch refinement post-processing module, the model achieves higher accuracy and segmentation quality.
FRONTIERS IN PLANT SCIENCE
(2022)
Article
Engineering, Biomedical
Bruno Machado Pacheco, Guilherme de Souza e Cassia, Danilo Silva
Summary: State-of-the-art brain tumor segmentation is achieved using deep learning models on multi-modal MRIs. However, manual correction of images is time-consuming and may result in skull-stripping faults that negatively affect tumor segmentation quality. Training models on non-skull-stripped images may be the best option for achieving high performance in clinical practice.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Geochemistry & Geophysics
Ronghua Shang, Mengmeng Liu, Junkai Lin, Jie Feng, Yangyang Li, Rustam Stolkin, Licheng Jiao
Summary: Despite the challenges in processing SAR images, the proposed CSHLC algorithm shows higher accuracy and better preservation of corners and small targets, compared to seven other state-of-the-art algorithms.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Artificial Intelligence
Long Xie, Laura E. M. Wisse, Jiancong Wang, Sadhana Ravikumar, Pulkit Khandelwal, Trevor Glenn, Anica Luther, Sydney Lim, David A. Wolk, Paul A. Yushkevich
Summary: The study presents a 3D hybrid approach combining multi-atlas segmentation and deep convolutional neural networks, called Deep Label Fusion (DLF), for medical image segmentation. The experiments demonstrate that DLF achieves comparable accuracy to the state-of-the-art deep convolutional neural network pipeline when evaluated on similar datasets, while outperforming it on tasks that involve generalization to datasets with different characteristics. DLF also consistently improves upon conventional multi-atlas segmentation methods.
MEDICAL IMAGE ANALYSIS
(2023)
Article
Optics
Haoyu Lyu, Lingbao Kong, Shixiang Wang, Min Xu
Summary: The wavefront deformation in the non-null test of optical freeform surface measurement was corrected using a wavefront propagation model, resulting in improved measurement accuracy. A freeform surface wavefront correction (FSWC) measurement system was established to validate the proposed method. Simulation and experimental studies showed that the proposed method effectively reduces the influence of wavefront deformation in space propagation, and the freeform surface form accuracy measured by FSWC can reach 10 nm RMS.
Article
Engineering, Biomedical
Shengxiang Liang, Xiaolong Yin, Li Huang, Jiayang Huang, Junchao Yang, Xiuxiu Wang, Lixin Peng, Yusi Zhang, Zuanfang Li, Binbin Nie, Jing Tao
Summary: This study developed a new skull stripping method for rat brain MRI using the U-2-Net neural network model. The U-2-Net model showed better performance compared to traditional methods and software in quantitative evaluations.
PHYSICS IN MEDICINE AND BIOLOGY
(2023)
Article
Neurosciences
Katrin Parmar, Vladimir S. Fonov, Yvonne Naegelin, Michael Amann, Jens Wuerfel, D. Louis Collins, Laura Gaetano, Stefano Magon, Till Sprenger, Ludwig Kappos, Cristina Granziera, Charidimos Tsagkas
Summary: Through longitudinal imaging data of 163 MS patients, it was found that cerebellar volumes are associated with clinical scores, particularly motor and cognitive impairments. Regional cerebellar volume reduction is correlated with motor and cognitive disability in MS, serving as a predictor for future disease progression.
Article
Endocrinology & Metabolism
Min Su Kang, Monica Shin, Julie Ottoy, Arturo Aliaga Aliaga, Sulantha Mathotaarachchi, Kely Quispialaya, Tharick A. Pascoal, D. Louis Collins, M. Mallar Chakravarty, Axel Mathieu, Asa Sandelius, Kaj Blennow, Henrik Zetterberg, Gassan Massarweh, Jean-Paul Soucy, A. Claudio Cuello, Serge Gauthier, Michael Waterston, Nathan Yoganathan, Etienne Lessard, Arsalan Haqqani, Kerry Rennie, Danica Stanimirovic, Balu Chakravarthy, Pedro Rosa-Neto
Summary: This study utilized in vivo longitudinal study design and a variety of biomarkers to evaluate the efficacy of a novel brain-penetrating anti-amyloid fusion protein treatment in a transgenic rat model of Alzheimer's disease. The treatment significantly reduced brain amyloid-beta levels and improved related biomarkers.
JOURNAL OF CEREBRAL BLOOD FLOW AND METABOLISM
(2022)
Article
Clinical Neurology
Golia Shafiei, Vincent Bazinet, Mahsa Dadar, Ana L. Manera, D. Louis Collins, Alain Dagher, Barbara Borroni, Raquel Sanchez-Valle, Fermin Moreno, Robert Laforce, Caroline Graff, Matthis Synofzik, Daniela Galimberti, James B. Rowe, Mario Masellis, Maria Carmela Tartaglia, Elizabeth Finger, Rik Vandenberghe, Alexandre de Mendonca, Fabrizio Tagliavini, Isabel Santana, Chris Butler, Alex Gerhard, Adrian Danek, Johannes Levin, Markus Otto, Sandro Sorbi, Lize C. Jiskoot, Harro Seelaar, John C. van Swieten, Jonathan D. Rohrer, Bratislav Misic, Simon Ducharme
Summary: The patterns of neurodegeneration in frontotemporal dementia are influenced by the structural network organization. The atrophy patterns in both genetic and sporadic cases mainly affect areas associated with the limbic intrinsic network and insular cytoarchitectonic class. The study findings provide insights into how different pathological entities can lead to the same clinical syndrome.
Article
Clinical Neurology
Elisea De Somma, Julia O'Mahony, Robert A. Brown, Brian L. Brooks, E. Ann Yeh, Alonso Cardenas de La Parra, Douglas Arnold, D. Louis Collins, Josefina Maranzano, Sridar Narayanan, Ruth Ann Marrie, Amit Bar-Or, Brenda Banwell, Christine Till
Summary: This study found that cognitive functioning in adolescents with ADS remained stable over a two-year period, but declines were noted in certain cognitive tests such as auditory working memory, symbol digit modalities, and visual matching. Lower normalized brain volume at early stages may predict negative changes in cognitive abilities.
CHILD NEUROPSYCHOLOGY
(2022)
Article
Neurosciences
John D. Lewis, Henriette Acosta, Jetro J. Tuulari, Vladimir S. Fonov, D. Louis Collins, Noora M. Scheinin, Satu J. Lehtola, Aylin Rosberg, Kristian Lidauer, Elena Ukharova, Jani Saunavaara, Riitta Parkkola, Tuire Lahdesmaki, Linnea Karlsson, Hasse Karlsson
Summary: The corpus callosum is the largest fiber tract in the human brain, connecting homologous areas of the two cerebral hemispheres. Sex differences exist in the allometric relationships between the size of the corpus callosum and brain size, and these differences are already present in newborns, suggesting that sexual dimorphism in brain lateralization may have prenatal origins.
HUMAN BRAIN MAPPING
(2022)
Article
Neurosciences
Vladimir S. Fonov, Mahsa Dadar, D. Louis Collins
Summary: In this study, a fully automatic quality control method based on deep learning is proposed to replace human raters for quality control assessment in the stereotaxic registration of T1w brain scans.
Article
Computer Science, Interdisciplinary Applications
Etienne St-Onge, Eleftherios Garyfallidis, D. Louis Collins
Summary: A hierarchical search algorithm is proposed to efficiently compute the distance between similar tractography streamlines. This algorithm improves the speed and accuracy of tractogram clustering using a hierarchical framework and space-partitioning search trees.
Article
Computer Science, Artificial Intelligence
Joshua Bierbrier, Houssem-Eddine Gueziri, D. Louis Collins
Summary: Given the importance of accurate image registration in the medical field, this study provides a taxonomy and scoping review of methods for quantitatively estimating registration error. The review identifies trends, advantages, and potential sources of bias, and offers suggestions for best practices and future research directions.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Geriatrics & Gerontology
Cassandra Morrison, Mahsa Dadar, Ana L. Manera, D. Louis Collins
Summary: White matter hyperintensities (WMHs) may be an early pathological change in aging. The study examined the influence of race on WMHs and whether these differences are influenced by vascular risk factors. Results showed that vascular risk factors had higher prevalence in Blacks than Whites. After controlling for vascular factors, only the longitudinal parietal WMH group differences remained significant, suggesting that vascular factors contribute to racial group differences observed in WMHs.
NEUROBIOLOGY OF AGING
(2023)
Article
Geriatrics & Gerontology
Cassandra Morrison, Mahsa Dadar, Sylvia Villeneuve, Simon Ducharme, D. Louis Collins
Summary: Increased age and cognitive impairment are associated with an increase in cerebrovascular pathology. Different methods of classifying subjective cognitive decline (SCD) result in different white matter hyperintensities (WMHs) burden between those with SCD and those without SCD. The various methods used to define SCD may reflect different types of underlying pathologies.
Article
Multidisciplinary Sciences
Victoria Madge, Vladimir S. Fonov, Yiming Xiao, Lucy Zou, Courtney Jackson, Ronald B. Postuma, Alain Dagher, Edward A. Fon, D. Louis Collins
Summary: Parkinson's disease (PD) is a complex neurodegenerative disorder that affects various brain regions. This study presents MRI templates created from multiple contrast MRI modalities, including T1w, T2*w, T1-T2* fusion, R2*, T2w, PDw, FLAIR, susceptibility-weighted imaging, and neuromelanin-sensitive MRI. The templates were created from PD patients and healthy controls, and the dataset is publicly available on the NIST MNI Repository and NITRC.
Article
Neurosciences
John D. Lewis, Vladimir S. Fonov, D. Louis Collins
Summary: This study examines the impact of blood-flow artifacts on the registration of T1-weighted data to a template using data from the Autism Brain Imaging Data Exchange. By identifying and partially removing blood-flow artifacts, we found that registering deblooded data to a template resulted in significantly higher similarity values compared to registering original data. This highlights the potential serious impact of blood-flow artifacts on data processing that relies on template registration.
HUMAN BRAIN MAPPING
(2023)
Article
Clinical Neurology
Alfie Wearn, Lars Lau Raket, D. Louis Collins, R. Nathan Spreng
Summary: Early detection of Alzheimer's disease is crucial for preventive treatment strategies. Texture analysis of the hippocampus can detect microstructural changes before cognitive impairment occurs, providing additional information beyond hippocampal volume for predicting future cognitive decline.
BRAIN COMMUNICATIONS
(2023)
Meeting Abstract
Clinical Neurology
M. Dadar, S. Mahmoud, S. Narayanan, D. Collins, D. L. Arnold, J. Maranzano
MULTIPLE SCLEROSIS JOURNAL
(2022)
Article
Behavioral Sciences
Satu J. Lehtola, Jetro J. Tuulari, Linnea Karlsson, John D. Lewis, Vladimir S. Fonov, D. Louis Collins, Riitta Parkkola, Jani Saunavaara, Niloofar Hashempour, Juho Pelto, Tuire Lahdesmaki, Noora M. Scheinin, Hasse Karlsson
Summary: Previous studies have shown that maternal pregnancy-specific anxiety (PSA) is associated with later difficulties in child emotional and social cognition, as well as memory, which are closely related to the amygdala and the hippocampus. This study aims to investigate the associations between PSA and newborn amygdalar and hippocampal volumes, as well as whether these associations are sex-specific. The results suggest a sexually dimorphic association between mid-pregnancy PSA and newborn amygdalar volumes.
STRESS-THE INTERNATIONAL JOURNAL ON THE BIOLOGY OF STRESS
(2022)