The Neural Basis of Sound Processing - Implications for the Perception of Speech
Sophie SCOTT

Probing Synaptic and Plasticity Deficits in Neurodevelopmental Disorders
Jacque Pak Kan IP

Mobile Exoneuromusculoskeleton for Telerehabilitation after Stroke
Xiaoling HU

Stroke is a leading cause of long-term disability in adults. However, regular and intensive physical training is hard to be accessed by discharged patients because of limited outpatient resources and interruptions by pandemics, e.g., COVID-19. Effective and self-help telerehabilitation with remote professional supervision is needed for long-term post-stroke rehabilitation. A hybrid robotic arm was invented for post-stroke upper limb rehabilitation at different joints, i.e., the elbow, wrist and fingers. It integrated the advantages of neuromuscular electrical stimulation, soft artificial muscle and exoskeleton in one system (exoneuromusculoskeleton), with the purposes to improve the robot-assisted rehabilitative effectiveness after stroke and to provide a novel rehabilitation service, i.e., self-help and mobile telerehabilitation for outpatients. The invention received a series of international awards and has been recognized as one of the five key developments which will impact the rehabilitation robotic global market in the near future. The technology and the service model are undergoing the process of commercialization, in collaboration with local companies and hospitals mainly located in the Greater Bay Area of China.

Cerebral Hemodynamics in Intracranial Atherosclerotic Disease
Cindy LENG

Background and objectives: Intracranial atherosclerotic disease (ICAD) is an important cause of ischemic stroke, particularly in East Asian populations. Those with an ischemic stroke attributed to ICAD (i.e., symptomatic ICAD) face a considerable risk of stroke relapse despite guideline-recommended, best medical treatment. We aimed to investigate the clinical relevance of cerebral hemodynamics in symptomatic ICAD.

Methods: In a series of studies, we used a computational fluid dynamics (CFD) model constructed based on CT angiography images to simulate blood flow across a symptomatic ICAD lesion. Translesional pressure ratio (PR) and translesional wall shear stress ratio (WSSR) were obtained in the CFD model to quantify the relative changes of these two hemodynamic metrics across a symptomatic ICAD lesion, when a low PR (relatively large pressure gradient across the lesion) and high WSSR (significantly elevated wall shear stress upon the lesion) indicate a hemodynamically more severe ICAD lesion.

Results: We found that symptomatic ICAD patients with a low PR and a high WSSR had a significantly higher risk of recurrent ischemic stroke in the same territory than those with normal PR and WSSR, despite best medical treatment, with hypoperfusion and/or artery-to-artery embolism as the probable mechanisms of most of the recurrent strokes. We have also identified a significant association between wall shear stress measures across symptomatic ICAD lesion and evolution of luminal stenosis in the lesion in medically treated patients. Moreover, we have also associated low PR and high WSSR with more severe covert cerebral small vessel disease burden in the ipsilateral hemisphere, in those with symptomatic ICAD in the anterior circulation.

Conclusions: Cerebral hemodynamics plays an important role in governing lesion evolution and the mechanisms and risk of stroke in ICAD, which may also underlie the development/progression of cerebral small vessel disease.

Direct Oral Anticoagulants - Insights from International and Local Real World Data
Bonaventure IP

Hong Kong Lower-grade Glioma Registry
Peter Y.M. WOO

Molecular Pathology of IDH Mutant Gliomas: The Unanswered Questions
Ho-Keung NG

Prediction of Molecular Pathology in Glioma from MRI Using AI
Zhenyu ZHANG

Stroke is a leading cause of long-term disability in adults. However, regular and intensive physical training is hard to be accessed by discharged patients because of limited outpatient resources and interruptions by pandemics, e.g., COVID-19. Effective and self-help telerehabilitation with remote professional supervision is needed for long-term post-stroke rehabilitation. A hybrid robotic arm was invented for post-stroke upper limb rehabilitation at different joints, i.e., the elbow, wrist and fingers. It integrated the advantages of neuromuscular electrical stimulation, soft artificial muscle and exoskeleton in one system (exoneuromusculoskeleton), with the purposes to improve the robot-assisted rehabilitative effectiveness after stroke and to provide a novel rehabilitation service, i.e., self-help and mobile telerehabilitation for outpatients. The invention received a series of international awards and has been recognized as one of the five key developments which will impact the rehabilitation robotic global market in the near future. The technology and the service model are undergoing the process of commercialization, in collaboration with local companies and hospitals mainly located in the Greater Bay Area of China.

Translational Therapy of Gliomas - Looking Beyond Conventional Targets
Vinay K. PUDUVALLI

Trials of CODEL & CATNON and the Up-to-date Standard of Care Treatment
Martin J van den BENT

REM Sleep Behavioral Disorder and a- Synucelinopathy - A Window to Prevent Neurodegeneration
Yun Kwok WING

Rapid eye movement (REM) sleep behavior disorder (RBD) is a distinct parasomnia characterized by repetitive occurrence of dream enactment behaviors related to vivid dreams and the loss of REM sleep related atonia (preservation of electromyography activity during REM sleep). Since the first discovery of RBD by Schenck and his colleagues in 1986, accumulating evidence has suggested that RBD is a prodromal stage of alpha-synucleinopathy, as over 90% of patients with idiopathic RBD would eventually develop alpha-synucleinopathies including Parkinson’s disease, dementia with Lewy body, and multiple system atrophy within 15 years of RBD diagnosis. By exploring the clinical profiles, developing assessment tools and diagnostic criteria, investigating the pathological mechanisms, determining the conversion rate to alpha-synucleinopathy, and identifying the neurodegenerative biomarkers in RBD, the ultimate aim is to delineate the progression of neurodegeneration and to develop neuroprotective therapy in preventing the neurodegeneration. In this talk, Prof. Wing will summarize the findings of previous studies on RBD and neurodegeneration and discuss future research directions and related challenges especially related to early diagnosis, high risk relatives, gut-brain hypothesis and psychiatric comorbidties and differential/converging motor and cognition pathways in the synucleinopathy outcome.

Gastrointestinal Dysfunction and Microbiota in Parkinson's Disease
Filip SCHEPERJANS

Gastrointestinal dysfunction affects up to 80% of Parkinson’s disease (PD) patients and may precede the onset of motor symptoms by years. It can heavily impact quality of life and pharmacological treatment. Neurodegenerative changes in the enteric nervous system can be found in earliest stages of PD. The colonic mucosa of PD patients shows an impaired barrier function, inflammation, oxidative stress, and bacterial invasion. There is accumulating evidence for an intense bidirectional interaction between gut microbiota and the brain influencing neuronal activity, behavior, as well as levels of neurotransmitter receptors, neurotrophic factors, and neuroinflammation. Pathways involving gut inflammation, amyloid cross-seeding, and short chain fatty acids may directly link gut microbiota and neurodegeneration.
Alterations of gut microbiome composition in PD have been shown in multiple studies suggesting that gut microbiota are related to PD, motor and non-motor symptoms, and disease progression. Recent research has revealed how gut microbiota interact with the host in PD through metabolic, immunologic, and epigenetic mechanisms. Gut microbiota changes have also been found in the prodromal phase of PD, namely subjects with isolated REM sleep behavior disorder (iRBD), and have been linked to several risk and prodromal markers of PD. Furthermore, increased exposure to antibiotics may increase PD risk and gut bacteria may impact drug absorption in PD. Thus, the gut and microbiota emerge as a crucial determinant of PD phenotype across different disease stages and may thus provide new diagnostic and therapeutic targets for a more individualized and disease-modifying treatment of PD.

Olfactory Function in Parkinson's Disease
Thomas HUMMEL

There is evidence from numerous studies, using both psychophysical and electrophysiological approaches, that impairment of olfaction is a characteristic and early feature of Parkinson´s disease. Deficits in the sense of smell precede clinical motor symptoms by years and can be used to assess risk for developing Parkinson´s disease in otherwise asymptomatic individuals. Recent data indicate that >95% of patients with Parkinson´s disease present with significant olfactory loss. Olfactory dysfunction is a reliable marker of the disease which is helpful in early and differential diagnosis.

A Tale of Two Frames - from Leksell G to Leksell Vantage
Danny T.M. CHAN

Directional Deep Brain Stimulation under General Anesthesia
Shin-Yuan CHEN

Background: Stereotactic implantation of electrode into deep seated nucleus may be resulted with substantial error and causing side-effects. Segmented directional deep brain stimulation (D-DBS) device may widen the therapeutic threshold. Because of fear and advanced symptoms of the patient, we have performed DBS under anesthesia for decades. Here we report our preliminary outcome on D-DBS.
Methods: From 2019/03 ~ 2021/06, a total of 51 cases (50 Parkinson’s disease, PD and 1 tardive dystonia) were underwent directional deep brain stimulation (D-DBS) at Hualien Tzu Chi General Hospital. Surgery was performed under inhalation general anesthesia and assisted with intraoperative microelectrode recording. Post-op image fusion and acute stimulation test were performed. The least possible direction was choosing for chronic stimulation.
Result: Twenty-eight PD patients (n=56 implants) were followed by a mean of 8.4±4.1months (3~17months). UPDRS part 3 was improved by a mean of 33±13%, LEDD was reduced by 51±27%. One patient committed self-mutilation during hospitalization with one-segment (direction) stimulation and was resulted with intracranial hemorrhage, whom recovered well with conservative treatment. Otherwise, no revision nor infection. Initial setting with one segment stimulation was 82%(n=46), two segments 11%(n=6), and all segments (omnidirectional) 7%(n=4). At the time of follow-up, the incidence of one segment stimulation was reduced from 82% to 27%, two segments stimulation was increased from 11% to 23% and omnidirectional was increased from 7% to 50%. The mean stimulation amplitude(mA) and impedance were: one segment stimulation 1.95±0.33/1992±428; two segments stimulation 2.46±0.59/1428±124; omnidirectional 2.06±0.70/1185±199.
Conclusion: DBS under general anesthesia is feasible and the treatment outcome was not inferior to the traditional awake surgery. The advent of directional stimulation may widen the therapeutic window. However, omnidirectional stimulation is suggested to be the initial setting for chronic stimulation, if acute stimulation test up to 4mA has no side effects. Raising the amplitude slowly in one segment (direction) stimulation, to avoid possible over stimulation side effects.

Application of Intraoperative CT in Robot-Assisted Deep Brain Stimulation
Xiaodong CAI

To introduce the application of intraoperative computed tomography (iCT) in Robot-Assisted Deep Brain Stimulation surgery and discuss the advantage and disadvantage of iCT based on our center’s experience.

Interventional MRI (iMRI) - guided Deep Brain Stimulation (DBS)
Paul S. LARSON

In the past, the gold standard for implantation of deep brain stimulator (DBS) electrodes was awake surgery with some form of physiologic mapping to confirm proper electrode placement. Techniques for physiologic mapping include single or multi-unit microelectrode recording, local field potential recording, micro- or macro-stimulation with observation of physiologic response or a combination of these techniques. At many centers in the United States and elsewhere, surgeons have moved to purely image-guided placement under general anesthesia. This so called “asleep DBS” technique was initially developed using interventional MRI (iMRI), but other groups have adopted interventional CT (iCT) as an alternative to MRI-guided surgery. An overview of the iMRI DBS procedure and its rationale will be provided, as well as a discussion of the pros and cons of awake versus asleep surgery, iMRI vs iCT and the clinical outcomes with iMRI-guided surgery as compared to traditional awake surgery. We will also explore future applications of iMRI for brain disorders beyond DBS.

Movement disorders and Their Stereotactic Surgical Treatment: A Retrospective Study of 5126 Patients at a Single Clinical Center over 22 Years
Yongjie LI

A Biological Definition of Alzheimer's Disease: NIA-AA Research Framework Present and Future
Clifford R. JACK

Machine Learning Derived Biomarker Based on MRI for Alzheimer's Disease
Vincent MOK

Structural magnetic resonance imaging (MRI) is currently the preferred first-line brain imaging modality for the evaluation of cognitive impairment. MRI is very useful in detecting cerebrovascular disease (e.g., silent infarcts) or in ruling out rare brain diseases (e.g., normal pressure hydrocephalus) that may account for the cognitive impairment. It can also serve as an imaging biomarker for Alzheimer’s disease (AD) as the pattern and severity of neurodegeneration (e.g., medial temporal lobe atrophy [MTA]) associated with AD can also be captured by MRI. Simple visual scale (e.g., Scheltens scale, score of 0 to 4) has been used to grade the severity of MTA over the past few decades. Although it can be easily applied in real-life clinical setting and has good diagnostic accuracy in diagnosing Alzheimer's dementia, it has only fair accuracy for diagnosing prodromal AD and low inter- and intra-rater reliability among inexperienced raters. In recent years, with further advancement in MRI post-processing technology, fully automated and accurate quantification of regional brain volumes (e.g., hippocampus) becomes feasible. Moreover, some post-processing pipelines were further translated into clinical applicable software that can be integrated into the management workflow of real-life clinical setting. These pipelines commonly include easy uploading of the imaging data to designated cloud platforms where data will be analysed, and automated generation of volumetric results back to the end users (e.g., radiologist) quickly. Although test-retest reliability of current software is usually excellent, different software may yield variable inter-scans reliability. Studies have consistently shown that volumetric measure outperform visual assessment in the diagnosis of prodromal AD or in the prognostication of mild cognitive impairment (MCI). In addition, given small change in hippocampal volume can be quantified over time, volumetry may be used to monitor disease progression or as surrogate outcome in clinical trials.

Apart from MTA, atrophy affecting other brain regions (e.g., entire temporal lobe, parietal and frontal lobe) is also present even at the preclinical stage and increases further as the disease continues to progress. With the advent of machine learning technology, difference in the various features detectable by structural MRI (e.g., regional volumes, tissue density, morphological feature of the cortical surface) between the “diseased” and “normal” brains can be recognised and be transformed into algorithms that can aid disease classification. Many studies published over the past two decades demonstrated the feasibility of applying machine learning technology upon MR scans for the differentiation between “AD dementia” and “normal” or between “progressive MCI” and “stable MCI”. Few studies had further derived a continuous “index” or “score” (e.g., from 0 to 100%) that reflects the similarity of the individual brain’s atrophy pattern to that of subjects with Alzheimer’s dementia. Many studies showed that machine learning-derived MRI-based biomarker based on multi-structure volumetric measures outperformed that of single volumetric measure in the diagnosis of AD or prognostication of MCI. In this lecture, we will report one of world first software that is able to generate not only multi-structure brain volumetry, but also a machine-learning derived integrated “index” that has been applied and validated in real-life clinical setting for the management of patients presenting with cognitive impairment.

Retinal Biomarkers for Alzheimer's Disease
Carol CHEUNG

Identifying biomarkers of Alzheimer's disease (AD) will accelerate the understanding of its pathophysiology, facilitate screening and risk stratification, and aid in developing new therapies. Developments in non-invasive retinal imaging technologies, including optical coherence tomography (OCT), OCT angiography and digital retinal photography, have provided a means to study neuronal and vascular structures in the retina in people with AD. Both qualitative and quantitative measurements from these retinal imaging technologies have been shown to be associated with cognitive function impairment and risk of AD. The development of computer algorithms for respective retinal imaging methods has further enhanced the potential of retinal imaging as a viable tool for rapid, early detection and screening of AD. In this talk, I present an update of current retinal imaging techniques and their potential applications in AD research. The directions in this expanding field will also be discussed.

Chemical Exchange Saturation Transfer (CEST) MRI for Alzheimer's Disease
Kannie CHAN

Gastrointestinal dysfunction affects up to 80% of Parkinson’s disease (PD) patients and may precede the onset of motor symptoms by years. It can heavily impact quality of life and pharmacological treatment. Neurodegenerative changes in the enteric nervous system can be found in earliest stages of PD. The colonic mucosa of PD patients shows an impaired barrier function, inflammation, oxidative stress, and bacterial invasion. There is accumulating evidence for an intense bidirectional interaction between gut microbiota and the brain influencing neuronal activity, behavior, as well as levels of neurotransmitter receptors, neurotrophic factors, and neuroinflammation. Pathways involving gut inflammation, amyloid cross-seeding, and short chain fatty acids may directly link gut microbiota and neurodegeneration.
Alterations of gut microbiome composition in PD have been shown in multiple studies suggesting that gut microbiota are related to PD, motor and non-motor symptoms, and disease progression. Recent research has revealed how gut microbiota interact with the host in PD through metabolic, immunologic, and epigenetic mechanisms. Gut microbiota changes have also been found in the prodromal phase of PD, namely subjects with isolated REM sleep behavior disorder (iRBD), and have been linked to several risk and prodromal markers of PD. Furthermore, increased exposure to antibiotics may increase PD risk and gut bacteria may impact drug absorption in PD. Thus, the gut and microbiota emerge as a crucial determinant of PD phenotype across different disease stages and may thus provide new diagnostic and therapeutic targets for a more individualized and disease-modifying treatment of PD.

scONE-seq: A One-tube Single-cell Multi-omics Method Enables Simultaneous Dissection of Molecular Phenotype and Genotype Heterogeneity from Frozen Tumors
Angela R. WU

Genomic and transcriptomic heterogeneity both play important roles in normal cellular function as well as in disease development. To be able to characterize these different forms of cellular heterogeneity in diverse sample types, we developed scONE-seq, which enables simultaneous transcriptome and genome profiling in a one-tube reaction. Previous single-cell-whole-genome-RNA-sequencing (scWGS-RNA-seq) methods require physical separation of DNA and RNA, often by physical separation of the nucleus from the cytoplasm. These methods are labor-intensive and technically demanding, time-consuming, or require special devices, and they are not applicable to frozen samples that cannot generate intact single-cell suspensions. scONE-seq is a one-tube reaction, thus is highly scalable and is the first scWGS-RNA-seq method compatible with frozen biobanked tissue. We benchmarked scONE-seq against existing methods using cell lines and lymphocytes from a healthy donor, and we applied it to a 2-year-frozen astrocytoma sample profiling over 1,200 nuclei, subsequently identifying a unique transcriptionally normal-like tumor clone. scONE-seq makes it possible to perform large-scale single-cell multi-omics interrogation with ease on the vast quantities of biobanked tissue, which could transform the scale of future multi-omics single-cell cancer profiling studies.

Identification of Early Predictors of Glioblastoma Evolution under Therapy
Jiguang WANG

Clonal evolution of cancer is a major challenge leading to treatment failure, but the molecular mechanisms of how cancer cells evolve and gain the capability of surviving intensive chemo- and/or radio- therapies remain elusive. Therefore, it is critically important to characterize the spatial and temporal dynamics of cancer cells and thereby mathematically modelling this process via big data integration. We have been working on diffuse gliomas, the most common and aggressive forms of primary tumors in adult brain whose treatment outcome is still very poor. Current therapies inevitably lead to tumor recurrence and the recurrent gliomas commonly become treatment resistance and incurable. Analyzing longitudinal and single-cell multi-omics data on this disease, our team aims to address the following questions: a) why cancer cells always display complex patterns of intratumoral heterogeneity; b) what is the temporal order of multiple somatic mutations detected in various cancer clones; c) how to predict the evolutionary path and clinical response of cancer cells under a certain therapy based on the sign seen earlier; and d) what are the key factors in tumor and its microenvironment that shape cancer evolution and determine cancer cell response under clinical intervention. In the process of addressing these questions, we will be able to unravel the mysteries of cancer evolution and it might provide a theoretical foundation for designing new means of treatment or diagnostics for better precision cancer medicine via targeting cancer dynamics.

The Epigenetic Mechanism and Clinical Translation of Metabolic Profiling in IDH Mutated Gliomas
Harvey W. HUA

IDH mutation was first found in gliomas by Duke group in 2008, and over 80% lower grade gliomas (grade II and III) harbored IDH mutation. IDH mutation could specifically induce 2-HG accumulation, which functionally altered the epigenetics in gliomas. IDH mutated gliomas are quite unique in clinical, as they favor young patients and frontal lobe. IDH mutated gliomas also could benefit from chemotherapy and total neurosurgical resection. So, it is important to extend the application of this special metabolic change in clinic management.
Previously, we found proline could be synchronously upregulated with 2-HG accumulation (Cell reports, 2018). We also investigated the metabolic profiling between IDH mutated and wild type gliomas. This metabolic reshape could be applied in PET-CT (BMC Cancer, 2021) and MRS for IDH mutation identification. In the meantime, metabolic profiling could also be rapidly detected intraoperatively with mass spectrometer (Lab Inv, 2018). We could identify 2-HG accumulation in 3 minutes with Mini-beta mass spectrometer intraoperatively, which was trained with 115 patients and validated with 75 patients. Meanwhile, real time metabolic photograph and delineation could be obtained in theatre.

Non-mitotic Mode of Proliferation in Primary Cancer Cells and Cell Lines
Hannah H. XUE

Meningeal hemangiopericytomas (HPC) are rare tumors of the meninges, aggressive versions of solitary fibrous tumors of the dura with high risk of reoccurrence. To understand the factors underlying the relapse of these rare tumors, comparative analysis at the genomic and cellular levels was performed based on AluScan sequencing on a recurrent chemotherapy-treated (Case 1) and benign (Case 2) cases. In genomic level, both cases showed CG to TG mutations along with TG to CG ones, particularly in the second-recurrence tissue of Case 1 (C1T3). At the cellular level, high frequency of multinucleation was observed in C1T3, along with increased non-mitotic modes of proliferation including cannibalism, shedding and tunneling when mitosis was suppressed. Furthermore, cell lines under cisplatin or epirubicin treatment showed variable cellular responses that included both mitotic and non-mitotic events. These findings suggest that when mitosis was reduced, aberrant proliferative prokaryotic characteristics become manifest which could contribute to cancer relapse. Therefore, an improved understanding of deviations from standard mitotic proliferation may represent an important aspect of cancer cell behavior that requires further investigation.

Stronger Association was shown between Cognition with Pulsatility Index than with Conventional Imaging Markers in Community Elderly without Dementia
Maggie L.M. LAM

Application of Pre-Hospital Identification and Notification of Stroke Patients (Pre-act) Initiative to the Management of Haemorrhagic Stroke
Queenie H.W. WONG

Dynamic Changes in Automatic Generated Quantitative Biomarkers - A Longitudinal Study
Xiang FAN

Biomarkers for the Detection of Progressive Early Alzheimer's Disease
Vera CAI

Glioblastoma Patient Survival Predictors treated by the Hong Kong Hospital Authority: A Multicenter Retrospective 15-year Review
Chi Chun LAM

Integrative Medicine as an Alternative Approach to Reduce Fatigue and Inflammation in Multiple Sclerosis
Angel NG

A Novel Brain Biopsy Cannula for Safe and Adequate Tissue Acquisition
Yitian XIAN

Mapping Structural Disconnection to Behavioral Dysexecutive Symptoms in Subcortical Stroke
Lisha WANG

The Development of DBS Device in China and Future Direction
Luming LI

Clinical Application of Directional Lead and Virtual Clinic to DBS Patients
Binith J. CHEERAN

The past decade has seen rapid progress in Deep Brain Stimulation Technology. This lecture will focus on one hardware (Directional Leads) and one software advance (Remote Programming), from ideation to implementation and clinical evidence development. The lecture will also offer the speakers personal views on future developments in the space.
Learning Goals:
1.Directional Lead design, biophysics and clinical evidence review
2.Remote Programming necessity, utility and clinical evidence review
Future developments: Remote monitoring, Sensing, Analytics

Personalised DBS Programming: Delivering Patient Specific Stimulation by Combining High-resolution Stimulation Control with Image Guided Programming Capabilities
Peter J. YOO

Workflow Considerations for BrainSense Technology
Sol T. De JESUS

The Spatial and Temporal Dynamics of Microglia Cells Following Transient Focal Cerebral Ischemia
Cyrus W. C. CHENG

Pax6/C-MYB Mediates Neuroinflammation and Amyloid Beta Clearance in a Term2 Dependent Manner in Alzheimer's Disease
Yi ZHANG

Preliminary Evaluation of the Therapeutic Potential of MIR-30C MIMIC Treatment in Acute Ischemic Stroke: Dose-Response and Therapeutic Time Window
Rong JIN

Current Management of Atypical Glioma
Martin J van den BENT