Lingling Deng¹, Huasheng Liu¹, Wen Liu¹, Yunjie Liao¹, Qi Liang¹, Chen Thomas Zhao², and Wei Wang^1
1Department of Radiology, the Third Xiangya Hospital, Central South University, Changsha, China, ²Philips Healthcare, Guangzhou, China

The structural covariance of connected gray matter has been demonstrated valuable in inferring large-scale structural brain networks. The alterations of grey matter structural covariance networks in prediabetes remains unclear. In this study, the topological features and robustness of gray matter structural covariance networks in prediabetes were examined. Results showed that the prediabetes group retained the small-worldness characteristics. The prediabetes group showed higher clustering coefficient, higher local efficiency and more vulnerable to random failure than healthy controls (HCs) group, suggesting that prediabetes disturbed the segregation of gray matter structural covariance networks, which provided new insights into the pathophysiology of this disease

Zeyang Li¹, Teng Ma¹, Lin-Feng Yan¹, and Guangbin Cui^1
1Hospital, Fourth Military Medical University, 569 Xinsi Road, Xi’an 710038, Shaanxi, China, Xi’an, China

Neuroimaging meta-analysis have identified abnormal neural activity alterations involved in type 2 diabetes mellitus (T2DM) patients, but there is no consistency and heterogeneity analysis between different brain imaging processing strategies. For the indicators obtained from varied post-processing methods reflect different neurophysiological and pathological characteristics, we further conducted a coordinated-based meta-analysis (CBMA) for two categories of neuroimaging literature grouped by similar data processing indicators. Compared to healthy controls, T2DM patients showed a significantly decreased brain activity in the right rolandic operculum, right supramarginal gyrus and right postcentral gyrus, providing a new non-invasive biomarker for T2DM neuropathy.

Yunge Zhang¹, Dongyue Zhou¹, Wei Zhao¹, Guoqiang Hu¹, Fengyu Cong¹, and Huanjie Li^1
1School of Biomedical Engineering, Faculty of Electronic and Electrical Engineering, Dalian University of Technology, Dalian, China

Atypical dynamic brain activities (dBAs) of people with autism spectrum disorder (ASD) were reported related to ASD symptoms. However, the most commonly used dynamic function connectivity method is affected by sliding window size. In this study, co-activate pattern analysis which keeps high temporal resolution and without the limitation of window size, is used to evaluate dBAs. It is found that the atypical dBAs of ASD is related to ASD core symptoms, and the atypical dBAs pattern of ASD is also affected by age.

Yanan Gao¹, Yanting Liu¹, Sihui Zhao¹, Hui Steve², Mikkelsen Mark², Edden A.E. Richard ², Chen Zhang³, and Bing Yu^4
1China Medical University, shenyang, China, ²Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, MD, USA F. M. Kirby Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, USA, Baltimore, MD, United States, ³MR Scientific Marketing, Siemens Healthineers, BeiJing, China, ⁴China Medical University, Shenyang, China

 A series of studies have suggested the cognitive deficits in patients with narcolepsy.  The aim of the current study is to explore the correlation between abnormal GABA+/Glx ratio in sleep state and abnormal cognitive function in patients with narcolepsy using Hadamard Encoding and Reconstruction of Mega-Edited Spectroscopy (HERMES) method. This study demonstrate that the ratio of GABA+/Glx in prefrontal lobe of narcolepsy patients during N2 sleep stage were higher than that of normal control group, which may be related to their abnormal cognitive functions.

Yayun Xiang¹, Xiaoxuan Dong², and Yongmei Li^1,3
1Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China, ²College of Computer & Information Science, Southwest University, Chongqing, China, Chongqing, China, ³The First Affiliated Hospital of Chongqing Medical University, Chongqing, China

The establishment and validation of accurate prognostic models in anti-NMDA receptor (NMDAR) encephalitis is lacking. This study aims to conduct an artifificial intelligence (AI) scheme to predict the prognosis of patients with anti-NMDAR encephalitis using clinical and machine learning features. We first bulid the clinical, deep learning and radiomics models, respectively. Then, we fuse the three schemes to build a fusion model and use an independent external dataset for further validation. The new fusion model significantly outperforms all other models. It demonstrates that applying AI method is an effective way to improve the performance of prognosis prediction in anti-NMDAR encephalitis.

Wenjing Zhang¹, Chengmin Yang¹, Zehong Cao², Feng Shi², and Su Lui^1
1Department of Radiology, West China Hospital of Sichuan University, Chengdu, China, ²Department of Research and Development, Shanghai United Imaging Intelligence Co., Ltd., Shanghai, China

To yield clinical utility in mental disorder identification individually, we used a multiple instance learning-based method to construct a digital model based on clinical MRI scans for automated detection of patients with psychiatric disorders. An accuracy of 84% was achieved in the primary dataset with 19453 subjects, and 76% in external dataset with 600 subjects. A higher sensitivity was achieved in identifying high-risk subjects than self-scaled questionnaires (71.1% vs 22.2%) in 148 prospectively recruited college students. With a complete workflow of development and validation, the constructed model is more practical to be translated in high-risk subject screening among vulnerable populations.

Yi-Tien Li^1,2, Yung-Chieh Chen², Yung-Li Chen³, Duen-Pang Kuo², and Cheng-Yu Chen^2
1Neuroscience Research Center, Taipei Medical University, Taipei, Taiwan, ²Translational Imaging Research Center, Taipei Medical University Hospital, Taipei, Taiwan, ³Department of Occupational Therapy, Kaohsiung Medical University, Kaohsiung, Taiwan

The objective of this study is to construct a framework for precise individualized prediction of post-concussive cognitive outcomes based on the early fMRI and neuropsychological biomarkers assessed at baseline to facilitate early therapeutic intervention and individualized rehabilitation strategies. Satisfactory predictions can be achieved for patients whose WM function did not recover after 3 months (accuracy = 87.5%), 6 months (accuracy = 83.3%), and 1 year (accuracy = 83.3%). The results prove the feasibility of using machine learning–based approaches to reveal predictive biomarkers related to poor post-concussive cognitive outcomes.

Wenjing Huang¹, Jing Zhang¹, Wanjun Hu¹, Pengfei Zhang¹, Jun Wang¹, and Shaoyu Wang^2
1Lanzhou University Second Hospital, Lanzhou, China, ²MR Scientific Marketing, Siemens Healthineers, ShangHai, China

Mild traumatic brain injury (mTBI) is not a static event, but a progressive injury. The neuroanatomical alterations of mTBI at the acute stage can be an initial step of damage leading to cognitive and emotional deficit. The potential mechanisms conferring these deficits are poorly understood. In this study, we employed graph theory to explore alterations of white matter connectome and to study their relationship with neuropsychological examinations. We found disrupted topologic manner of the brain’s structural connectome, and the changes of right middle occipital gyrus explain depressive symptoms in the acute phase of mTBI due to white matter network alterations.

Chenchen Hua¹, Lu Qiu², Leting Zhou³, Yi Zhuang², Bin Xu¹, Liang Wang³, Min Xu⁴, and Haoxiang Jiang^2
1Diagnostic Radiology, Wuxi People's Hospital Affiliated to Nanjing Medical University, Wuxi, China, ²Diagnostic Radiology, The Affiliated Wuxi Children's Hospital of Nanjing Medical University, Wuxi, China, ³Nephrology, Wuxi People's Hospital Affiliated to Nanjing Medical University, Wuxi, China, ⁴Human Anatomy, Medical School, Southeast University, Nanjing, China

Chronic kidney disease (CKD) is chronic progressive renal parenchymal damage caused by multiple factors. Evaluation of the degree of renal interstitial fibrosis (IF) is of great importance in treatment and prognosis prediction. Herein, we use multi-parametric magnetic resonance imaging (mpMRI) to assess renal IF noninvasively with intention to identify the optimal MRI biomarkers for clinical application. Our results show that multi-parameter prediction model using cortical longitudinal relaxation time (cT1) and cortical true diffusion coefficient (cDt) can effectively assess the degree of renal IF and support clinical diagnosis, treatment strategy and risk stratification in CKD patients.

Yu Shen¹, Yan Bai¹, Xianchang Zhang², Yaping Wu¹, Menghuan Zhang¹, and Meiyun Wang^1
1Henan Provincial People's Hospital, Zhengzhou, China, ²MR Collaboration, Siemens Healthineers Ltd, Beijing, China

The brain fibre-specific changes in patients who recovered from Coronavirus Disease 2019 (COVID-19) for one year but still had anosmia (reCOVID19_ano) deserves to be investigated. This study used fixel based analysis (FBA), which is a novel diffusion MRI post-processing technique that enables fibre tract-specific statistical analysis, to investigate the white matter alterations in reCOVID19_ano. The results revealed that the intra-axonal volume in reCOVID19_ano was significantly increased in several fibre tracts. We speculate that impaired fibre tracts may recover gradually in a compensatory mechanism after one year recovery in reCOVID19_ano patients.

Chien-Ying Lee^1,2, Chia-Wei Li³, Chien-Yuan Lin³, Wen-sheng Huang⁴, and Ko-Han Lin^1
1Department of Nuclear Medicine, Taipei Veterans General Hospital, Taipei, Taiwan, ²Department of Biomedical Imaging and Radiological Sciences, National Yang-Ming University, Taipei, Taiwan, ³GE Healthcare, Taipei, Taiwan, ⁴Department of Nuclear Medicine, Taipei Medical University Hospital, Taipei, Taiwan

¹⁸F-fluoro-Boronophenylalanine (FBPA) Positron emission tomography (PET) plays a crucial role for patient selection prior to boron neutron capture therapy. As integrated PET-MR is being used frequently in clinical research and routine, this study aimed to compare regional difference of brain standard uptake values on the PET quantification accuracy in FBPA PET-CT and PET-MR among lesions and normal tissues. Compared to SUVR in PET-CT, PET-MR data showed significant lower SUVR in lesion regions, while no significance was observed in the normal tissue regions. The result of this study may accelerate the routine use of PET-MR in tumoral quantification before BNCT.

Wenxiao Liu^1,2, Minglei Wang², Xiaocheng Wei³, Min Li³, Jianguo Zhao², Xin Ge¹, Xuhong Yang¹, Peng Yong¹, and Xiaodong Wang^2
1Ningxia Medical University, Yinchuan, China, ²General Hospital of Ningxia Medical University, Yinchuan, China, ³GE Healthcare, Beijing, China

In this study, we aim to investigate whether T1mapping and T2mapping can be used as diagnostic tools for screening MHE, and strive earlier time for diagnosis and treatment for these patients. It was concluded that T1 values of bilateral putamen, right globus pallidus and bilateral frontal lobe in MHE group were statistically significant compared with HC group and NMHE group and the T2 values of right putamen and right frontal lobe were statistically significant. In addition, the T2 values of the HC group in this study were highly correlated with Hallgren and Sourander’s autopsy results.

Hongxia Li¹, Ming Zhang², Hongjiang Wei², and Yiwen Wu^1
1Department of Neurology & Institute of Neurology, Ruijin Hospital, affiliated with Shanghai Jiao Tong University School of Medicine, Shanghai, China, ²School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China

We found increased susceptibility in patients compared with HCs that primarily involved cerebellar structures . The result of 2D-flatmap visualization showed that the involved subregions of iron accumulation were in the right Crus Ⅱ and right lobule Ⅶb. Interestingly, no significant differences in QSM value of basal ganglia, thalamus, or motor cortex were observed between ICD patients and HCs. Moreover, there was no significant difference between groups regarding the GMV in the cerebellum. Of consistent findings with standard structural MRI technique highlights the potential of probing tissue iron across the cerebellum with QSM as a disease biomarker

Xing Qian¹, Kok Pin Ng^2,3,4, Kwun Kei Ng¹, Fang Ji¹, Pedro Rosa-Neto ^5,6, Serge Gauthier ⁶, Nagaendran Kandiah ^2,3,4, and Juan Helen Zhou ^1,3,7,8
1Centre for Sleep and Cognition and Centre for Translational MR Research, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore, Singapore, Singapore, ²Department of Neurology, National Neuroscience Institute, Singapore, Singapore, Singapore, ³Duke-NUS Medical School, Singapore, Singapore, Singapore, ⁴Lee Kong Chian School of Medicine, Nanyang Technological University Singapore, Singapore, Singapore, Singapore, Singapore, ⁵Translational Neuroimaging Laboratory, The McGill University Research Centre for Studies in Aging, Montreal, Canada, Montreal, QC, Canada, ⁶Alzheimer’s Disease Research Unit, The McGill University Research Centre for Studies in Aging, McGill University, Montreal, Canada, Montreal, QC, Canada, ⁷Department of Electrical and Computer Engineering, National University of Singapore, Singapore, Singapore, Singapore, Singapore, ⁸Integrative Sciences and Engineering Programme (ISEP), National University of Singapore, Singapore, Singapore, Singapore, Singapore

While emerging evidence suggests the association between network neurodegeneration and memory varies with pathology in Alzheimer’s disease (AD), the trajectory of this relationship remains elusive. We stratified 708 participants into non-amyloid/non-tau, tau-only, and AD pathology groups and examined the associations between individual-level structural and metabolic network integrity and memory across cognitive stages (cognitively normal, mild cognitive impairment, and probable AD) in each pathology group. The associations of hippocampal and default mode networks with memory exhibited differential pathology-dependent trajectories across cognitive stages. Our findings pave the way for early interventions and stage-dependent remedies to modify disease trajectory and improve clinical outcomes.

Yingying Dong¹, Yu Luo², Mingming Wang², Quan Tao², and He Wang^3
1Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China, ²Department of Radiology, Shanghai Fourth People's Hospital, School of Medicine, Tongji University, Shanghai, China, ³1Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University 2Human Phenome Institute, Fudan University, Shanghai, China

According to a large amount of emergency and follow-up data, we propose an optimized threshold of the apparent diffusion coefficient to identify the ischemic core with fair accuracy, which is crucial for the decision of the treatment scheme and operation mode.