一、 參加會議經過
我於 4 月 11 日搭機離開台灣，經日本成田機場轉機至美國舊金山市，因此我能夠有 一天的時間了解當地交通環境、調整時差，以及準備報告的內容。會議於 4 月 13 日開始，會議一共進行四日，直到 4 月 16 日結束，我於 4 月 17 日到加州大學柏克 萊分校進行實驗室參訪，並與該校神經科學中心與心理系的學者討論研究合作議題 與實驗內容，於 4 月 18 日搭機返台。
二、與會心得
The 2013 annual meeting of the Cognitive Neuroscience Society (CNS2013) was carried out in April 13 ‐ April 16 2013 in San Francisco, US. The main topic for the meeting is cognitive neuroscience research. There were 3 keynote lectures, a talk for George A.
Miller Prize in Cognitive Neuroscience, a talk for Distinguished Career Contributions Award, and 3 talks for Young Investigator award. Moreover, there were 6 invited symposiums and 5 mini symposiums, and more than 1000 poster presentations this year.
I summerise the main ideas from the talks which are related to my research here.
In the keynote talks, Professor William Newsome firstly presented their recent work in decision making using single‐unit recording in monkeys (14 April). He combined a dynamic system analysis and a dynamic recurrent network model to investigate neural mechanisms underlying gating and integration in a context‐dependent decision‐making task. He showed that both task‐relevant and irrelevant sensory responses (e.g. colour or motion) were observed in the frontal eye field (FEF) within the prefrontal cortex that
3
supports decision‐making processing according to task goals. In addition, gating of context‐specific information in FEF occurs very late in the process. Their findings suggest that FEF plays a critical role in filtering out irrelevant sensory information. Next, Professor Patricia Kuhl presented her work in the field of language, especially in early language learning, using MRI, DTI, and MEG in infants and normal adults (15 April).
Finally, Professor Joseph LeDoux presented his perspectives and findings in human emotion experiences and survival mechanisms (16 April). He also emphasised that the capacity to detect/respond to threat stimuli may not be the analogues to the capacity to consciously experience them. However, the capacities to detect and experience to threats were both important in survival.
In the talk of Distinguished Career Contributions Award (15 April), Professor Robert Knight firstly presented his amazing work in prefrontal cortex (PFC) and its role in human cognition. Abundant evidence exists that the PFC is a source of top‐down signals to guide and integrate human behaviours according to behavioural expectations. He demonstrated a direct recording of the electrocorticogram (ECoG) from the cortical surface in neurosurgical patients and showed that gamma frequency oscillations (> 60 Hz) were generated in the PFC to host cognitive operations. More importantly, he found a functional coupling between high‐ (> 80 Hz, gamma power) and low‐frequency (4‐8 Hz, theta phase) bands of ongoing ECoG signals. He also showed that different behavioural tasks can be related to different patterns of theta‐gamma coupling. His findings provide a clear evidence that phase and amplitude cross frequency coupling in regulating large scale brain networks during cognitive processing. Secondly, Professor Knight showed their data about human peri‐Sylvian language areas which mediated feedback control of
4
vocal pitches (e.g. speak sound) using ECoG recordings. Their data exhibited a clear correlation in ECoG responses between peri‐Sylvian neural network (e.g. listening to sound pitches) and ventral premotor areas (e.g. generating vocal responses). This neural network may underlie auditory feedback control of pitch.
I am particularly interested in the issue of network of attention. In the invited symposium (16 April), Professor Michael Posner, who is the Godfather in the field of human attention and cognition, reviewed his findings in attention network. Moreover, he showed this attention network is related to individual differences and can be modulated by practice. Secondly, Professor Sabine Kastner tested connectivity between pulvinar (within the thalamus) and cortical areas in both functional and structural aspects. She showed that thalamus can regulate visual processing and orient attention.
Next, Professor Earl Miller demonstrated their recent findings in top‐down versus bottom‐up control of attention in the prefrontal and parietal cortices in monkeys. Their data showed strong neuronal synchronisation between prefrontal and parietal (e.g. LIP) cortices in lower frequency ranges (22‐34 Hz) during top‐down attention but in higher frequency ranges (35‐55 Hz) during bottom‐up attention. Also, he showed that early activity in frontal neurons reflected target location in a visual search task during top‐down attention; however, early activity in parietal neurons was observed during bottom‐up attention. Finally, Professor Anna Christina Nobre showed how long‐term contextual memories can guide subsequent attention and modulate target‐related visual processing in a series of experiments developed in her laboratory. The original paradigm was based on the original contextual cueing experiments. The main behavioural finding was that participants were able to use their memories to guide their attention to learned
5
locations within complex scenes, as revealed by faster reaction times to target stimuli appearing in valid (learned) as opposed to neutral (unlearned) locations. In addition, the validity effect was stronger in the memory‐orienting than for visual‐orienting. fMRI activations revealed a common frontal‐parietal network for attentional guidance between memory‐ and visually‐guided orienting of attention, which is commonly seen during spatial‐attention tasks. Furthermore, for memory‐based attention, there was a selective activation of the hippocampus, which is involved in the retrieval of object‐context associations as described previously, and this activity correlated with the ensuing behavioural advantage for valid memory cues. Overall, the results of this novel paradigm replicate and extend the findings of the classic contextual cueing paradigms, and reveal a selective interaction between systems involved in memory and attention when humans use their long‐term memories to guide their attention in cluttered environments. The hippocampus may indeed have a ‘proactive’ role by which incoming perceptual information can be biased based on previous experiences.
I am also interested in decision making. In the invited symposium (16 April), Professor Matthew Rushworth designed experiments to study the neural mechanisms of foraging in human. They showed that decision making and foraging may depend on distinct neural substrates in ventromedial prefrontal cortex (vmPFC) and anterior cingulate cortex (ACC) using distinct reference frames. Choice variables were represented in invariant reference to foraging for alternatives that were correlated to ACC activity.
However, activity in the vmPFC reflected encoding of specific pre‐defined choices. They concluded that the ACC carries three signals for decision making: representations of value of options, averaged values of alternatives, and the cost of actions.
6
In addition to the issues, neuroimaging methods are also very critical to the field of cognitive neuroscience. In recent years, fMRI pattern classification methods have led to rapid advances in the field. In a mini‐symposium of analysing patterns of brain activity (14 April), Professors Frank Tong, John Serences, and Jack Gallant reviewed their recent advances in brain and mind reading (or decoding) in the issues of visual cognition and short‐term memory (Frank Tong), attention (John Serences), and semantic categorisation (Jack Gallant). This decoding technique revealed how different types of information, visual, auditory, intentional, sematic, etc., are represented in specific brain areas. Taken together, their evidence showed not only functional organization of these representations but also brain activity patterns that can be predictive of future actions.
7
三、發表論文全文或摘要
Searching for targets within visual short‐term memory versus perceptual representations: A magnetoencephalograhic study
Bo‐Cheng Kuo1, Duncan E. Astle2,4, Gaia Scerif2, Mark W. Woolrich3, Anna Christina Nobre2,3
1 Department of Psychology, National Chengchi University, Taipei, Taiwan
2 Department of Experimental Psychology, University of Oxford, Oxford, United Kingdom
3 Oxford Centre for Human Brain Activity, University of Oxford, Oxford, United Kingdom
4 Medical Research Council Cognition and Brain Sciences Unit, Cambridge, United Kingdom
Abstract
Recent event‐related potential (ERP) studies have revealed that searching for relevant items from within visual short‐term memory (VSTM) representations involves spatiotopically specific top‐down biasing of neural activity in a manner similar to that occurs during visual search for items within perceptual representations. However, the underlying neural substrates of VSTM and perceptual search were not fully investigated. Here our aim was to localise the neural source of brain activity associated with spatiotopic selection of targets in VSTM representations by using magnetoencephalograhy (MEG). Participants (N = 11) performed a visual and VSTM search task. Participants viewed a sample shape and a search array of different shapes. Their task was to respond whether the sample shape was present in the search array. In visual search trials, the sample shape appeared before the search array, and participants searched for the target item within the perceptual array. In VSTM trials, the sample shape appeared after the search array, and participants searched for the target item within representations held in VSTM. In accordance with previous ERP findings, we found the mN2pc for both VSTM and visual search from event‐related magnetic field recording. MEG
8
source localization using beamforming analysis revealed that the mN2pc was related to posterior visual areas for both search tasks. Moreover, the mN2pc for VSTM search also reflected neural activity in the frontal cortex. Taken together, these findings bolster the notion that top‐down biasing in VSTM may share properties with spatially specific attentional mechanisms that bias perceptual processing in favour of the relevant information.
四、建議
(1) 希望當局能更重視認知神經科學基礎研究，以此作為發展應用領域的基礎。
(2) 給予議題面與方法面相等的資源投入與發展鼓勵。
五、攜回資料名稱及內容 會議手冊。