• Research Progress
  • On September 4th, 2018, SCIENCE ADVANCES published an research article entitled “Trancranial Brain Atlas”. The study was accomplished by Professor Chaozhe Zhu’s group at the State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University. This work introduced a concept and methodological framework of a new type of brain atlas specific for transcranial brain mapping technologies.

    Brain atlases are maps to descript the relationship between the functional and structural organization of human brain. In construction of brain atlases, researchers first define a 3-D standard brain space. Then anatomical or functional information of human brain observed in specific researches, including information of cytoarchitecture, function, network and gene expression, are mapped to locations in the standard brain space in form of labels. In this way, brain atlases provide a common platform to integrate the isolated observations different researches, towards a systemic understanding on human brain. On the other hand, in researches of brain mapping, prior knowledge provided by brain atlases are also important accordance in defining region of interest (ROI), identifying nodes for network analysis or target for brain stimulation. Brain mapping researches are conducted mainly with two kinds of invasive technologies. The first is the stereo imaging techniques represented by MRI and PET, the image of which is obtained in 3-D brain space and can be directly transform into the standard brain space via spatial normalization. The second in the transcranial technique represented by fNIRS and TMS. These techniques image or stimulate the lateral cortex via devices set on the 2-D surface of scalp, from where the 3-D standard brain space of brain atlases is invisible. This gap hinders the brain atlases from providing guidance for transcranial techniques, and therefore the neuroscience researches or clinical applications conducted with these techniques.

    The address this issue, in the article published in SCIENCE ADVANCES,  the authors proposed a concept of transcranial brain atlas (TBA, Fig.1) and methodological framework for its construction. First, they defined a Cranial Proportional Coordinates system (CPC system), based on cranial fiducial landmarks and a proportional measurement. The coordinate system provided a coordinates system to quantify the whole space of the cranial surface to set the trancranial devices. Second, based on a 114-subjects structural MRI data base, the modeled the probabilistic mapping between the CPC space and the standard brain space. Last, based on a two-step Markov model, they solved the probabilistic mapping to the label space of brain atlases.
    Fig. 1 Concept of trancranial brain atlas.

    Four TBAs were constructed in the study based on brain atlases widely adopted in neuroscience researches, they were the brain atlases of LPBA40, AAL, Tailarach and Craddock400. Results of the construction were validated. First, as demonstrated by the analysis of slit-half reliability, TBAs constructed from different MRI data bases were highly reproducible. Second, as demonstrated by the analysis of group-individual prediction, localization results told by individual structural MRI were predicted by the group-level TBA with high accuracy. Third, the group-level TBA also predicted the localization result of cross-racial subjects with high accuracy. Finally, TBA constructed from atlas AAL was applied to guide the optobe setting in a fNIRS based finger-tapping experiment. Compared with traditional 10-20 system, guidance of TBA significantly improved the performance of optode setting in both the coverage of ROI and the consistency of measured cortical location across the subjects. As a result, the intended brain activation evoked by the finger-tapping task was observed with higher sensitivity under the guidance of TBA.
    Fig. 2 颅骨比例坐标系统(CPC 系统)。(A)CPC系统在个体头壳上的实例。(B)CPC坐标系统的平面投影表示。
    Fig. 3 TBA_LPBA , TBA constructed from brain atlas LPBA40.(A)Lobe-level maximum probability map.(B)Lobe-level maximum likelihood label map.(C)Gyrus-level maximum probability map.(D)Gyrus-level maximum likelihood label map(E)Gyrus-level maximum likelihood label map rendered in a perspective projection.
    Fig. 4 Comparison between TBA guided (top) vs. traditional 10-20 system guided (bottom) optode placement. (A) Intended probe locations on the scalp. (B) Consistency of channel locations (C) Group-level HbO activation during finger-tapping task. (D) ROI coverage of each probe placement.

    This study proposed a novel brain atlas that can be directly applied for transcranial techniques. It provide a possibility for researchers to predict the invisible brain labels from visible scalp location, with a spatial accuracy at level of millimeter which is comparable to the spatial resolution of fNIRS and TMS. Combined with technologies like real-time positioning system or augment-reality, TBAs also provide a potential as a intuitional and accurate navigation technique, which would promote the application of transcranial techniques in researches of awake infants, real-life social interaction and neuro-ergonomics and in clinical treatment for neurological and psychiatric disorders.

    This article is contributed by Xiang Xiao from Prof. Chaozhe Zhu’s group, at State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University. The authors acknowledge Prof. Li Liu and Prof. Jiang Qiu’s contribution of sharing the data. This research is funded by the National Natural Science Foundation of China (61431002) and National 973 project (2014CB846100).

    Link of the article: