Article
Biology
Papangkorn Inkeaw, Salita Angkurawaranon, Piyapong Khumrin, Nakarin Inmutto, Patrinee Traisathit, Jeerayut Chaijaruwanich, Chaisiri Angkurawaranon, Imjai Chitapanarux
Summary: This paper introduces a new method for automatically segmenting hemorrhage subtypes in head CT scans based on a deep learning model. The experimental results show that the proposed method outperforms previous studies in terms of segmentation performance for each hemorrhage subtype.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Multidisciplinary Sciences
Zhegao Piao, Yeong Hyeon Gu, Hailin Jin, Seong Joon Yoo
Summary: This study proposes a TransHarDNet image segmentation model for diagnosing intracerebral hemorrhage in brain CT scan images. By applying a HarDNet block to the U-Net architecture and connecting the encoder and decoder with a transformer block, the network complexity was reduced, the inference speed improved, and the high performance compared to conventional models was maintained.
SCIENTIFIC REPORTS
(2023)
Article
Computer Science, Information Systems
Lu Li, Meng Wei, Bo Liu, Kunakorn Atchaneeyasakul, Fugen Zhou, Zehao Pan, Shimran A. Kumar, Jason Y. Zhang, Yuehua Pu, David S. Liebeskind, Fabien Scalzo
Summary: This paper proposes a U-net based deep learning framework for automatic detection and segmentation of hemorrhage strokes in CT brain images. By comparing different Deep Learning topologies, adopting adversarial training, and training and evaluating the model on two different datasets, the effectiveness, robustness, and advantages of the proposed deep learning model in hemorrhage lesion diagnosis have been demonstrated, making it possible to be a clinical decision support tool in stroke diagnosis.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2021)
Article
Computer Science, Information Systems
Abhishek Bal, Minakshi Banerjee, Amlan Chakrabarti, Punit Sharma
Summary: Automated brain tumor segmentation of MR image is a challenging task. The proposed method using rough-fuzzy C-means and shape based topological properties achieved better performance in terms of statistical volume metrics compared to previous algorithms.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Neurosciences
Jun Xu, Rongguo Zhang, Zijian Zhou, Chunxue Wu, Qiang Gong, Huiling Zhang, Shuang Wu, Gang Wu, Yufeng Deng, Chen Xia, Jun Ma
Summary: This study evaluated a deep framework optimized for the segmentation and quantification of ICH, EDH, and SDH, showing high accuracy and reliability in manual segmentations. The Dense U-Net model outperformed the ABC/2 method in both internal and external tests, indicating its potential for efficiently developing treatment strategies for intracranial hemorrhage in clinics.
FRONTIERS IN NEUROSCIENCE
(2021)
Article
Multidisciplinary Sciences
Lei Wu, Haishuai Wang, Yining Chen, Xiang Zhang, Tianyun Zhang, Ning Shen, Guangyu Tao, Zhongquan Sun, Yuan Ding, Weilin Wang, Jiajun Bu
Summary: In this study, an AI system named MULLET was proposed for precise and fully automatic segmentation of liver lesions in real-patient CECT images. MULLET effectively embeds the important ROIs and explores multi-phase contexts using a transformer-based attention mechanism. Evaluated on a large dataset, MULLET demonstrated significant performance gains compared to state of the art methods.
Article
Biology
Bohao Xu, Yingwei Fan, Jingming Liu, Guobin Zhang, Zhiping Wang, Zhili Li, Wei Guo, Xiaoying Tang
Summary: In this study, an automatic segmentation network (CHSNet) was proposed to segment lesions in cranial CT images based on the characteristics of acute cerebral hemorrhage images. The network achieved 3D visualization and localization of the cranial lesions after segmentation. Experimental results demonstrated the effectiveness of the model on a dataset of 203 patients, achieving high performance in segmenting hemorrhage in CT images of stroke patients.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Clinical Neurology
Kevin J. Chung, Hulin Kuang, Alyssa Federico, Hyun Seok Choi, Linda Kasickova, Abdulaziz Sulaiman Al Sultan, MacKenzie Horn, Mark Crowther, Stuart J. Connolly, Patrick Yue, John T. Curnutte, Andrew M. Demchuk, Bijoy K. Menon, Wu Qiu
Summary: In this study, a novel interactive segmentation method based on convex optimization was spatially and volumetrically validated to accurately and reliably measure intracranial hemorrhage growth. The method showed good spatial overlap, excellent volume correlation, and good repeatability, indicating its utility for measuring intracranial hemorrhage volume and volume change on non-contrast CT images.
INTERNATIONAL JOURNAL OF STROKE
(2021)
Article
Computer Science, Hardware & Architecture
Zhe Chen, Nan Qiu, Hui Feng, Dongfang Dai
Summary: Accurate lung tumor segmentation is crucial for radiotherapy and targeted therapy, and PET and CT imaging provide complementary evidence. This paper proposes a novel joint level set model that integrates PET and CT evidence in a unified energy form for co-segmentation of lung tumors.
COMPUTERS & ELECTRICAL ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Xiaoliang Lei, Xiaosheng Yu, Jianning Chi, Ying Wang, Jingsi Zhang, Chengdong Wu
Summary: This study introduces an automatic sparse constrained level set method for brain tumor segmentation in MR images, achieving high accuracy and stability through the construction of a sparse representation model and an energy function based on the level set method.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Engineering, Biomedical
Cristobal Arrieta, Carlos A. Sing-Long, Joaquin Mura, Pablo Irarrazaval, Marcelo E. Andia, Sergio Uribe, Cristian Tejos
Summary: This paper presents a new method for incorporating shape prior knowledge based on intrinsic alignment approach, but extending it for scaling, translation and rotation invariance. The approach uses a regularization term based on eigenvalues and eigenvectors of the covariance matrix, leading to a new set of evolution equations. Testing on 2D and 3D synthetic and medical images shows the effectiveness of using shape priors with intrinsic scaling, translation and rotation alignment in different segmentation problems.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2021)
Article
Multidisciplinary Sciences
Mohammed A. Al-masni, Dong-Hyun Kim
Summary: An end-to-end deep learning segmentation method called CMM-Net, incorporating global contextual features and dilated convolution module, achieved superior performance on multiple medical imaging datasets. The method showed excellent segmentation results for skin lesions, retinal blood vessels, and brain tumors, demonstrating high generalizability.
SCIENTIFIC REPORTS
(2021)
Article
Clinical Neurology
Meera Srikrishna, Nicholas J. Ashton, Alexis Moscoso, Joana B. Pereira, Rolf A. Heckemann, Danielle van Westen, Giovanni Volpe, Joel Simren, Anna Zettergren, Silke Kern, Lars-Olof Wahlund, Bibek Gyanwali, Saima Hilal, Joyce Chong Ruifen, Henrik Zetterberg, Kaj Blennow, Eric Westman, Christopher Chen, Ingmar Skoog, Michael Scholl
Summary: This study found that CT-based volumetric measures can accurately distinguish patients with neurodegenerative diseases from healthy individuals, as well as patients with prodromal dementia from controls. These measures are significantly associated with cognitive functioning, biochemical markers, and neuroimaging markers of neurodegenerative diseases. After further validation, CT-based volumetric measures have the potential to become a preferred examination tool for the diagnosis of neurodegenerative diseases.
ALZHEIMERS & DEMENTIA
(2023)
Article
Computer Science, Information Systems
Lifang Zhou, Lu Wang, Weisheng Li, Bangjun Lei, Jianxun Mi, Weibin Yang
Summary: This study presents a multi-stage framework for liver location and segmentation, using Faster RCNN for liver region localization and a Gaussian mixture model-based signed distance function to enhance shape prior flexibility. Experimental results demonstrate the efficacy of the proposed method on 40 CT scan images.
Article
Medicine, General & Internal
Andreas Sakkas, Christel Weiss, Marcel Ebeling, Frank Wilde, Sebastian Pietzka, Qasim Mohammad, Oliver Christian Thiele, Robert Andreas Mischkowski
Summary: The aim of the study was to identify clinical indicators for primary cranial CT imaging in patients with mild traumatic brain injury (mTBI). The study also aimed to evaluate the need for short-term hospitalization based on clinical and CT findings. A retrospective study was conducted on mTBI patients over a five-year period, analyzing demographic data, clinical and radiological findings, and outcomes. The study found that a Glasgow Coma Scale (GCS) score of <15, loss of consciousness, amnesia, seizures, cephalgia, somnolence, dizziness, nausea, and clinical signs of fracture were significantly associated with acute intracranial hemorrhage (ICH).
JOURNAL OF CLINICAL MEDICINE
(2023)
Article
Cell & Tissue Engineering
Smarajit Chakraborty, Wee Kiat Ong, Winifred W. Y. Yau, Zhihong Zhou, K. N. Bhanu Prakash, Sue-Anne Toh, Weiping Han, Paul M. Yen, Shigeki Sugii
Summary: The study establishes CD10 as a functionally relevant ASC biomarker, positively determining adipocyte maturation and browning potential of ASCs. CD10 regulates ASC's adipogenic maturation non-canonically by modulating endogenous lipolysis, indicating CD10 as a useful biomarker for pro-adipogenic drug screening, with dexamethasone and retinoic acid identified as stimulator and inhibitor of adipogenesis, respectively.
STEM CELL RESEARCH & THERAPY
(2021)
Article
Neurosciences
Bo Xu Ren, Isaac Huen, Zi Jun Wu, Hong Wang, Meng Yun Duan, Ilonka Guenther, K. N. Bhanu Prakash, Feng Ru Tang
Summary: Early life radiation exposure during different developmental stages induces varied brain pathophysiological changes, which may be related to the development of neurological and neuropsychological disorders later in life.
Article
Radiology, Nuclear Medicine & Medical Imaging
Prakash K. N. Bhanu, Channarayapatna Srinivas Arvind, Ling Yun Yeow, Wen Xiang Chen, Wee Shiong Lim, Cher Heng Tan
Summary: The study developed a deep learning-based MRI fat analysis system for automated quantification of subcutaneous and visceral fat compartments, demonstrating high accuracy and reproducibility, providing comprehensive fat compartment composition analysis and visualization in less than 10 seconds.
MAGNETIC RESONANCE MATERIALS IN PHYSICS BIOLOGY AND MEDICINE
(2022)
Article
Clinical Neurology
Ralf A. Kockro, Eike Schwandt, Florian Ringel, Christian Valentin Eisenring, Wieslaw Lucjan Nowinski
Summary: The study evaluated the use of a 3D interactive atlas for teaching surgical skull base anatomy in a clinical setting, showing that students and residents significantly improved their anatomical knowledge after interacting with the software. The interactive 3D computer graphical environments are highly suitable for conveying complex anatomy and surgical concepts, yet remain underutilized in clinical practice.
JOURNAL OF NEUROLOGICAL SURGERY PART B-SKULL BASE
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Wieslaw L. Nowinski
Summary: Despite the development of various brain-related resources, there is currently no large, systematic, comprehensive, extendable, and beautiful 3D reconstructed image repository of a living human brain that extends to the head and neck. In this study, the author created such a repository and populated it with images derived from a 3D atlas constructed from MRI and CT scans. The repository features multiple standard views, modes of presentation, and spatially co-registered image sequences, and contains galleries constructed from different tissue classes.
JOURNAL OF DIGITAL IMAGING
(2022)
Article
Computer Science, Interdisciplinary Applications
Wieslaw L. Nowinski
Summary: The development of human brain atlas is research-oriented and has limited applications in clinical practice. This article introduces a new definition and architecture of a user-extendable reference human brain atlas for education, research, and clinical use. The proposed architecture supports knowledge gathering, presentation, use, sharing, and discovery through four functional units and a user interface, contributing to neuroeducation, research, and clinical decision-making.
Article
Multidisciplinary Sciences
Geetha Soujanya Chilla, Ling Yun Yeow, Qian Hui Chew, Kang Sim, K. N. Bhanu Prakash
Summary: Schizophrenia is a major psychiatric disorder that poses a burden on patients and caregivers. This study successfully classified schizophrenia and healthy control groups using diverse neuroanatomical measures and ensemble methods, and also found correlations between specific neuroanatomical measures and quality of life assessment scores.
SCIENTIFIC REPORTS
(2022)
Article
Clinical Neurology
Eelin Tan, Khurshid Merchant, Bhanu Prakash Kn, Arvind Cs, Joseph J. Zhao, Seyed Ehsan Saffari, Poh Hwa Tan, Phua Hwee Tang
Summary: This study developed a machine learning model based on CT images to predict the amplification status of MYCN in pediatric neuroblastoma. The model showed good predictive accuracy in clinical data.
CHILDS NERVOUS SYSTEM
(2022)
Article
Anatomy & Morphology
Wieslaw L. Nowinski
Summary: This paper proposes a new definition and construction method for sulci, as well as a novel approach for their presentation. The proposed definition and method are morphology-based, quantitative, and simple, and have important implications for neuroeducation.
JOURNAL OF ANATOMY
(2022)
Article
Neurosciences
Mazira Mohammad Ghazali, Che Mohd Nasril Che Mohd Nassir, Nur Suhaila Idris, Geetha Chilla, Bhanu K. N. Prakash, Muzaimi Mustapha
Summary: This study highlights the subtle impacts of occult cerebral small vessel disease (CSVD) manifestations (WMHs and ePVS) on neurocognition in asymptomatic working-aged adults with low-to-moderate cardiocerebrovascular risk scores, providing valuable information from a population-based cohort in a suburban area of Malaysia.
JOURNAL OF INTEGRATIVE NEUROSCIENCE
(2022)
Article
Neuroimaging
Wieslaw L. Nowinski
Summary: This study integrates neuroanatomy with neuroradiology by creating labeled images to bridge 2D radiology and 3D anatomy. The method utilizes a 3D brain atlas to create planar radiologic and surface neuroanatomic images, enabling precise spatial correspondence. The labeled dual 2D-2D/3D neuroimage sequences created for various structures and systems can be used for education, research, and clinical practice.
NEURORADIOLOGY JOURNAL
(2023)
Article
Psychiatry
Qian Hui Chew, K. N. Bhanu Prakash, Li Yang Koh, Geetha Chilla, Ling Yun Yeow, Kang Sim
Summary: This study aimed to replicate earlier findings by Chand et al. (2020) in identifying neuroanatomical subtypes of schizophrenia patients using brain structural measures. Two subtypes, SG-1 and SG-2, were found, with SG-1 associated with brain ventricle enlargements, basal ganglia volume increase, longer illness duration, and deficit status, while SG-2 was associated with reductions in cortical and subcortical structures. These replicated findings have clinical implications for early intervention, response monitoring, and prognosis of schizophrenia. Future studies may consider using a multi-modal neuroimaging approach to further understand the neurobiological composition of different subtypes.
SCHIZOPHRENIA RESEARCH
(2022)
Article
Computer Science, Interdisciplinary Applications
K. N. Bhanu Prakash, C. S. Arvind, Abdalla Mohammed, Krishna Kanth Chitta, Xuan Vinh To, Hussein Srour, Fatima Nasrallah
Summary: This study demonstrates the effectiveness of the GA-UNet model for automatic segmentation and quantification of traumatic brain injury (TBI) lesions using multi-parametric MR data. The study also reveals the patterns of changes in TBI lesions over time, providing a promising approach for large cohort studies.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Ling Yun Yeow, Yu Xuan Teh, Xinyu Lu, Arvind Channarayapatna Srinivasa, Eelin Tan, Timothy Shao Ern Tan, Phua Hwee Tang, Bhanu Prakash Kn
Summary: This study proposes an end-to-end deep-learning framework for automatic tumor segmentation and radiomics features-based classification of MYCN gene amplification, achieving high accuracy.
JOURNAL OF COMPUTER ASSISTED TOMOGRAPHY
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Manish Gawali, C. S. Arvind, Shriya Suryavanshi, Harshit Madaan, Ashrika Gaikwad, K. N. Bhanu Prakash, Viraj Kulkarni, Aniruddha Pant
Summary: This study compares three privacy-preserving distributed learning techniques in developing binary classification models for detecting tuberculosis, analyzing their performance in terms of classification, communication and computational costs, and training time. A novel distributed learning architecture called SplitFedv3 and alternate mini-batch training for split learning are proposed, showing better performance than existing methods in experiments.
MEDICAL IMAGE UNDERSTANDING AND ANALYSIS (MIUA 2021)
(2021)