Article
Neurosciences
Hang Wu, Dong Wang, Yueyao Liu, Musi Xie, Liwei Zhou, Yiwen Wang, Jin Cao, Yujuan Huang, Mincong Qiu, Pengmin Qin
Summary: Using multivariate pattern analysis and dynamic causal modelling analysis, this study explored the regional activation pattern and inter-region effective connection during the perception of the subject's own name (SON). The findings showed that SON and other names could be distinguished based on the activation pattern in the primary auditory cortex. The study also revealed an excitatory effect of SON on connections from the anterior insula/inferior frontal gyrus to the primary auditory cortex and temporoparietal junction, highlighting the importance of the influence of the insula on the primary auditory cortex during self-processing.
HUMAN BRAIN MAPPING
(2023)
Article
Engineering, Biomedical
Charles Guan, Tyson Aflalo, Kelly Kadlec, Jorge Gamez de Leon, Emily R. Rosario, Ausaf Bari, Nader Pouratian, Richard A. Andersen
Summary: This study aims to enable neural control of individual prosthetic fingers for paralyzed participants. By implanting neural arrays in the posterior parietal cortex (PPC), researchers were able to accurately decode attempted finger movements and achieve control of individual fingers through a brain-machine interface (BMI). The study also identified neural signals in the brain that are linked to finger movements. These findings have significant implications for hand restoration strategies in individuals with tetraplegia.
JOURNAL OF NEURAL ENGINEERING
(2023)
Article
Neurosciences
Stefan Czoschke, Cora Fischer, Tara Bahador, Christoph Bledowski, Jochen Kaiser
Summary: The study showed concurrent representations of pitch and location of a single object in multiple brain regions, supporting feature integration in working memory.
JOURNAL OF NEUROSCIENCE
(2021)
Article
Neurosciences
Ke Bo, Lihan Cui, Siyang Yin, Zhenhong Hu, Xiangfei Hong, Sungkean Kim, Andreas Keil, Mingzhou Ding
Summary: This study investigates the temporal dynamics of affective scene processing in the brain using simultaneous EEG-fMRI recordings. The results show that perceptual processing of complex scenes begins in early visual cortex within 80 ms, followed by the ventral visual cortex at 100 ms. Affect-specific neural representations start to form between 200-300 ms, supported mainly by occipital and temporal cortices. These representations are stable and last up to 2 seconds, indicating the involvement of distributed brain areas in sustaining affective scene processing.
Article
Computer Science, Interdisciplinary Applications
David Lopez-Garcia, Jose M. G. Penalver, Juan M. Gorriz, Maria Ruz
Summary: This study introduces MVPAlab, a MATLAB-based flexible decoding toolbox for multidimensional electroencephalography and magnetoencephalography data. The toolbox implements several machine learning algorithms to perform multivariate pattern analyses, cross-classification, temporal generalization matrices, and feature and frequency contribution analyses. The results show that MVPAlab is capable of discriminating between different experimental conditions.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2022)
Article
Chemistry, Analytical
Haroon Khan, Farzan M. Noori, Anis Yazidi, Md Zia Uddin, M. N. Afzal Khan, Peyman Mirtaheri
Summary: The study conducted experiments on individual finger-tapping using fNIRS technology, achieving promising classification accuracy through sophisticated data processing and machine learning algorithms. This could lead to further research directions in fNIRS-based brain-computer interface applications.
Article
Neurosciences
Ke Bo, Siyang Yin, Yuelu Liu, Zhenhong Hu, Sreenivasan Meyyappan, Sungkean Kim, Andreas Keil, Mingzhou Ding
Summary: The study found that decoding accuracy for unpleasant versus neutral and pleasant versus neutral images in retinotopic visual areas was significantly above chance level, indicating that affective scenes trigger valence-specific neural representations in these areas, influenced by reentry signals from anterior brain regions.
Article
Engineering, Biomedical
Brian M. Dekleva, Jeffrey M. Weiss, Michael L. Boninger, Jennifer L. Collinger
Summary: This study aims to improve cursor click decoding for point-and-click and click-and-drag control using iBCI technology. By identifying prominent neural responses related to hand grasp, a new approach based on transient responses was developed, which outperformed the standard binary state classification method. This transient-based approach provides high degree of cursor click control, marking an important step towards high-performance cursor control and clinical translation of iBCI technology.
JOURNAL OF NEURAL ENGINEERING
(2021)
Article
Chemistry, Multidisciplinary
Arnau Dillen, Fakhreddine Ghaffari, Olivier Romain, Bram Vanderborght, Uros Marusic, Sidney Grospretre, Ann Nowe, Romain Meeusen, Kevin De Pauw
Summary: Brain-computer interfaces (BCIs) can enable individuals to interact with devices based on their brain activity. However, the high costs associated with research-grade electroencephalogram (EEG) acquisition devices make them impractical for everyday use. This study demonstrates that decoding movement intention from a limited number of sensors is feasible, opening up the possibility of using commercial sensor devices for BCI control.
APPLIED SCIENCES-BASEL
(2023)
Article
Neurosciences
Sumner L. Norman, David Maresca, Vassilios N. Christopoulos, Whitney S. Griggs, Charlie Demene, Mickael Tanter, Mikhail G. Shapiro, Richard A. Andersen
Summary: This study demonstrates the use of functional ultrasound neuroimaging to detect the neural correlates of movement planning in the brain, providing a critical advancement towards the development of less invasive, high resolution, and scalable neuro-recording and brain interface tools.
Article
Engineering, Biomedical
Minji Lee, Ji-Hoon Jeong, Yun-Hee Kim, Seong-Whan Lee
Summary: The classification performance of motor imagery in stroke patients is higher than in healthy controls, indicating better neural control during motor imagery for stroke patients. However, there is no significant difference between the accuracies of motor execution and motor imagery, suggesting the need for further research into this area.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2021)
Article
Neurosciences
Jiafan Lin, Dongrong Lai, Zijun Wan, Linqing Feng, Junming Zhu, Jianmin Zhang, Yueming Wang, Kedi Xu
Summary: In this study, the representation and decoding of different laterality and regions arm motor imagery in unilateral motor cortex were examined using local field potentials (LFPs). The results showed that different tasks had significant differences in average energy and could be decoded using LFP signals. Moreover, the 135-300 Hz band signal had the highest decoding accuracy and the contralateral and bilateral signals had more similar power activation patterns and larger signal correlation than contralateral and ipsilateral signals, bilateral and ipsilateral signals.
FRONTIERS IN HUMAN NEUROSCIENCE
(2023)
Article
Neurosciences
Huixiang Yang, Kenji Ogawa
Summary: This study investigated whether different types of motor imageries can be classified based on the location of the activation peaks or the multivariate pattern analysis (MVPA) of functional magnetic resonance imaging (fMRI), and compared the difference between visual motor imagery (VI) and kinesthetic motor imagery (KI).
Article
Psychology, Biological
Fariba Sharifian, Daniel Schneider, Stefan Arnau, Edmund Wascher
Summary: Decoding electroencephalogram brain representations is essential in understanding cognitive information processing. Utilizing multivariate pattern analysis of task variations provides a detailed spatio-temporal outline of information flow. Results suggest the presence of a higher order feedback control system and its effects on behavior.
INTERNATIONAL JOURNAL OF PSYCHOPHYSIOLOGY
(2021)
Article
Multidisciplinary Sciences
Omair Ali, Muhammad Saif-ur-Rehman, Susanne Dyck, Tobias Glasmachers, Ioannis Iossifidis, Christian Klaes
Summary: Brain-computer interfaces (BCIs) enable communication between humans and machines by translating brain activity into control commands. However, the EEG signals commonly used in BCIs are often contaminated with noise, making it difficult to find meaningful patterns for classification. This study proposes a novel feature extraction method, a data augmentation method, and a CNN architecture to improve the classification accuracy of EEG signals.
SCIENTIFIC REPORTS
(2022)
Article
Oncology
Teng Zhang, Chengxiu Zhang, Yan Zhong, Yingli Sun, Haijie Wang, Hai Li, Guang Yang, Quan Zhu, Mei Yuan
Summary: This study investigates the potential of radiomics as a diagnostic tool to predict invasiveness of part-solid nodules with solid components. Results show that radiomics model outperforms clinical-radiographic model in predicting invasive lesions.
FRONTIERS IN ONCOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Peipei Wang, Eryuan Gao, Jinbo Qi, Xiaoyue Ma, Kai Zhao, Jie Bai, Yong Zhang, Huiting Zhang, Guang Yang, Jingliang Cheng, Guohua Zhao
Summary: Quantitative analysis of mean apparent propagator (MAP)-MRI is useful for distinguishing glioblastoma (GBM) from solitary brain metastasis (SBM). Multivariate analysis combined with multiple regions of interest (ROIs) can improve diagnostic performance.
EUROPEAN JOURNAL OF RADIOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jinbo Qi, Peipei Wang, Guohua Zhao, Eryuan Gao, Kai Zhao, Ankang Gao, Jie Bai, Huiting Zhang, Guang Yang, Yong Zhang, Xiaoyue Ma, Jingliang Cheng
Summary: This study explored the value of histogram analysis based on NODDI in differentiating between GBM and SBM, and compared the diagnostic performance of two ROI placements. The results showed that the multivariate logistic regression model based on NODDI histogram analysis had better performance than the optimal single parameter in distinguishing GBM from SBM, and the ROI placed on the whole tumor area exhibited better diagnostic performance.
JOURNAL OF MAGNETIC RESONANCE IMAGING
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Ke-Wen Jiang, Yang Song, Ying Hou, Rui Zhi, Jing Zhang, Mei-Ling Bao, Hai Li, Xu Yan, Wei Xi, Cheng-Xiu Zhang, Ye-Feng Yao, Guang Yang, Yu-Dong Zhang
Summary: A study developed and tested an artificial intelligence (AI) system for diagnosing clinically significant prostate cancer (CsPC) using MRI. The AI system performed comparably or better than radiologists in both internal and external tests. Factors such as Gleason score, lesion location, PI-RADS score, and lesion size significantly impacted the accuracy of the AI system.
JOURNAL OF MAGNETIC RESONANCE IMAGING
(2023)
Article
Oncology
Ruiqi Yu, Ke-wen Jiang, Jie Bao, Ying Hou, Yinqiao Yi, Dongmei Wu, Yang Song, Chun-Hong Hu, Guang Yang, Yu-Dong Zhang
Summary: In this study, an artificial intelligence (AI)-aided Prostate Imaging Reporting and Data System (PI-RADS(AI)) was developed and validated for prostate cancer (PCa) diagnosis based on MRI. UNet-Seg and 3D-Resnet models were used to detect and segment prostate lesions, achieving an automatic AI-aided diagnosis for PCa. The results showed that PI-RADS(AI) outperformed or matched the performance of over 70% of general readers in the MRI assessment of PCa, providing diagnostic benefits to clinical practice.
BRITISH JOURNAL OF CANCER
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jia-Xiang Xin, Da-Xiu Wei, Yan Ren, Jun-Long Wang, Guang Yang, Huojun Zhang, Jianqi Li, Caixia Fu, Ye-Feng Yao
Summary: Selective probing of glutamate (Glu) and glutamine (Gln) signals in human brain in vivo is important for studying their metabolisms. Glu-/Gln- targeted pulse sequences were developed to selectively probe Glu and Gln signals. These sequences successfully separated Glu and Gln signals in healthy human brains, providing an effective method for distinguishing H-1-MR signals of Glu and Gln.
MAGNETIC RESONANCE IN MEDICINE
(2023)
Article
Multidisciplinary Sciences
Long Cui, Yang Song, Yida Wang, Rui Wang, Dongmei Wu, Haibin Xie, Jianqi Li, Guang Yang
Summary: This study proposes a new method to detect and remove motion artifacts in magnetic resonance images. A convolutional neural network (CNN) model is trained to filter motion-corrupted images, and the unaffected phase-encoding (PE) lines are used to reconstruct the final image using compressed sensing (CS). The results show that the proposed algorithm effectively alleviates motion artifacts.
Article
Oncology
Ziyu Le, Dongmei Wu, Xuming Chen, Lei Wang, Yi Xu, Guoqi Zhao, Chengxiu Zhang, Ying Chen, Ye Hu, Shengyu Yao, Tingfeng Chen, Jiangping Ren, Guang Yang, Yong Liu
Summary: The study aims to establish a predictive model for acute severe hematologic toxicity (HT) during radiotherapy in patients with cervical or endometrial cancer. It investigates the integration of clinical features and computed tomography (CT) radiomics features of the pelvic bone marrow (BM) to define a more precise model. The model combines clinical and radiomics features and achieves a higher area under the receiver operating characteristic curve (AUC) compared to models based on clinical or radiomics features alone. Further investigation is warranted to explore the value of pelvic BM radiomics in chemoradiotherapy-induced HT.
RADIOTHERAPY AND ONCOLOGY
(2023)
Article
Neurosciences
Yida Wang, Naying He, Chunyan Zhang, Youmin Zhang, Chenglong Wang, Pei Huang, Zhijia Jin, Yan Li, Zenghui Cheng, Yu Liu, Xinhui Wang, Chen Chen, Jingliang Cheng, Fangtao Liu, Ewart Mark Haacke, Shengdi Chen, Guang Yang, Fuhua Yan
Summary: A DL-based pipeline for automatic PD diagnosis is proposed, which uses QSM and T1W images to segment deep gray matter nuclei and distinguish PD from HC. The model achieved high accuracy in segmenting brain nuclei and showed promising performance in PD diagnosis with high AUC values on both internal and external testing cohorts.
HUMAN BRAIN MAPPING
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Haijie Wang, Yida Wang, He Zhang, Xuan Yin, Chenglong Wang, Yuanyuan Lu, Yang Song, Hao Zhu, Guang Yang
Summary: The purpose of this study was to develop a deep learning tool for antenatal diagnosis of prenatal placenta accreta spectrum (PAS) using T2-weighted MR images. By extracting the utero-placental boundary region image, the DL model achieved an AUC of 0.860 and 0.897 in the internal test and external test cohorts, respectively, significantly outperforming three radiologists (internal test AUC, 0.737-0.770). This fully automatic DL pipeline can improve the accuracy of PAS diagnosis using MRI for radiologists.
JOURNAL OF MAGNETIC RESONANCE IMAGING
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jing Zhang, Chenao Zhan, Chenxiu Zhang, Yang Song, Xu Yan, Yihao Guo, Tao Ai, Guang Yang
Summary: The purpose of this study was to develop an automatic computer-aided diagnosis (CAD) pipeline based on multiparametric magnetic resonance imaging (mpMRI) and investigate the role of different imaging features in the classification of breast cancer. The study included 222 histopathology-confirmed breast lesions and trained a neural network-based lesion segmentation model to extract radiomics features from DWI, T2WI, and DCE parametric maps. Models based on sequence combinations achieved higher diagnostic accuracy compared to BI-RADS scores, and the joint model combining radiomics and BI-RADS scores achieved the highest accuracy.
Article
Chemistry, Inorganic & Nuclear
Mingxi Jiang, Yajuan Zhang, Zihao Yang, Haibo Li, Jinliang Li, Jiabao Li, Ting Lu, Chenglong Wang, Guang Yang, Likun Pan
Summary: Metal ion doping is an effective method to improve the electrochemical performance of metal oxide anode materials for lithium-ion batteries. Machine learning models were built to predict the charging and discharging performance of doped oxide anode materials before synthesis, saving time and resources. The study found a correlation between the electronegativity of the dopant element and the capacity performance of the material.
INORGANIC CHEMISTRY FRONTIERS
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Qiong Ma, Yinqiao Yi, Tiejun Liu, Xinnian Wen, Fei Shan, Feng Feng, Qinqin Yan, Jie Shen, Guang Yang, Yuxin Shi
Summary: This study developed and evaluated a radiomics signature based on MRI to identify invisible changes of basal cisterns in tuberculous meningitis patients. The signature combined T2-weighted images and deep learning segmentation, providing a fully automatic and non-invasive tool for the diagnosis of TBM.
EUROPEAN RADIOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Funing Chu, Yun Liu, Qiuping Liu, Weijia Li, Zhengyan Jia, Chenglong Wang, Zhaoqi Wang, Shuang Lu, Ping Li, Yuanli Zhang, Yubo Liao, Mingzhe Xu, Xiaoqiang Yao, Shuting Wang, Cuicui Liu, Hongkai Zhang, Shaoyu Wang, Xu Yan, Ihab R. Kamel, Haibo Sun, Guang Yang, Yudong Zhang, Jinrong Qu
Summary: This study developed and validated an optimized model for predicting the survival of patients with esophageal squamous cell carcinoma (ESCC) based on the 1-mm-isotropic-3D contrast-enhanced StarVIBE MRI sequence and clinical risk factors. The combined model, which incorporated radiomics features and clinical factors, showed better predictive efficacy compared to radiomics models alone.
EUROPEAN RADIOLOGY
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jingyu Zhong, Chengxiu Zhang, Yangfan Hu, Jing Zhang, Yun Liu, Liping Si, Yue Xing, Defang Ding, Jia Geng, Qiong Jiao, Huizhen Zhang, Guang Yang, Weiwu Yao
Summary: This study implemented a pipeline to automatically segment the ROI and used a nomogram integrating the MRI-based radiomics score and clinical variables to predict responses to neoadjuvant chemotherapy (NAC) in osteosarcoma patients. The results showed that this automated approach could aid radiologists in identifying pathological good responders to NAC.
EUROPEAN RADIOLOGY
(2022)