平均表觀傳播係數相關性研究及其在性別差異之應用

Relationships among Mean Apparent Propagator MRI Indices and Their Applications to Sex Differentiation

魏榆臻(Yu-Chen Wei)

指導教授 ： 曾文毅

國立臺灣大學/醫學院/醫療器材與醫學影像研究所/碩士(2018年)

https://doi.org/10.6342/NTU201800883

摘要

這項研究是橫跨不同擴散磁振造影模型之擴散係數的系統化相關性分析，目的是為了更瞭解平均表觀傳播磁振造影(Mean apparent propagator MRI, MAP-MRI)所產生的平均表觀傳播磁振造影係數以及其在男女大腦微結構差異觀察之應用。平均表觀傳播磁振造影是近年來由Özaeslan團隊提出來重建擴散磁振造影資料的方法，經由平均表觀傳播磁振造影，我們可以得到許多不同的平均表觀磁振造影係數，其中包含綜合非等項性指標(Generalized fractional anisotropy, GFA)、非等項性指標(Fractional anisotropy, FA)、軸向擴散值(Axial diffusivity, AD)、徑向擴散值(Radial diffusivity, RD)、平均擴散值(Mean diffusivity, MD)、表觀平均長度(Mean apparent length, AML)、表觀平均面積直徑(Effective diameter of mean apparent area, AMAD)、表觀平均體積直徑(Effective diameter of mean apparent volume, AMVD)、平行非高斯性(Parallel non-Gaussianity, NGP)、垂直非高斯性(Orthogonal non-Gaussianity, NGO)以及非高斯性(Non-Gaussianity, NG)等11種平均表觀傳播磁振造影參數。相較於前面5種擴散係數，我們對於其他擴散係數較不熟悉，為了要更瞭解這些係數，我們針對這些係數進行了一個系統化的相關性分析，藉以瞭解係數彼此之間之相關性以及關聯，並將這些係數應用在觀察男女大腦微結構差異上，來探討男女之間的差異。
在這項研究中，我們總共使用了122名健康成人(男女各61人)的擴散頻譜磁振造影影像，根據擴散峰度磁振造影以及Özaeslan團隊提出的平均表觀傳播磁振造影來分別計算分析影像，並推導出14個擴散參數圖後，經由神經束基礎自動分析術(Tract-based automatic analysis, TBAA)來獲得每位參與者這14種係數在76條大腦神經纖維束之曲線。在實驗一中，我們在每條神經纖維束上為每對擴散係數之間進行相關性的分析，而兩個擴散係數之間的相關性的定義為在所有神經纖維束上量測到之相關性Z值的平均。有了這些擴散係數彼此之間的相關性之後，我們更進一步地使用了階層式群聚分析來將這些擴散係數進行分群。在實驗二中，我們使用了平均表觀傳播磁振造影係數(包含了綜合非等項性指標、非等項性指標、軸向擴散值、徑向擴散值、平均擴散值、觀平均面積直徑、表觀平均體積直徑和表觀平均體積直徑)對這61名男性和61名女性參與者進行了逐步的雙樣本t檢驗。
在實驗一中，我們發現這些根據這些係數彼此之間的相關性，我們可以將這些係數分為三群。第一群有非高斯性、垂直非高斯性和平行非高斯性，第二群有軸向擴散值、表觀平均長度以及軸向峰度值(Axial kurtosis, AK)，最後第三群則包含了綜合非等項性指標、非等項性指標、徑向擴散值、平均擴散值、表觀平均體積直徑、表觀平均體積直徑、徑向峰度值(Radial kurtosis, RK)以及平均峰度值(Mean kurtosis, MK)。在實驗二中，我們發現男性在左側鉤束(uncinate fasciculus)、兩側的內側蹄系(medial lemniscus)、左側的中央後丘腦放射鈎(thalamic radiation postcentral)和海馬迴胼胝體纖維(callosal fiber of hippocampus)表現較好，然而女性在右側的聽覺丘腦放射鈎(thalamic radiation auditory)以及背後側前額葉皮質胼胝體纖維(callosal fiber of dorsal lateral prefrontal cortex)表現較好。
根據這些係數的定義，AD和AML的相關性、RD和AMAD的相關性以及MD和AMVD的相關性的結果很符合我們的預期。在另一方面，峰度係數和非高斯性係數的低相關性可能與這兩種係數在描述非高斯部分訊號的方式不同有關，峰度係數的描述方式同時結合了擴散張量和峰度張量。在實驗二中，根據這些神經纖維束所連接的區域與功能，由於男生在兩側的內側蹄系表現較好，所以我們認為男生可能在身體協調性上表現比女生好，而女生因為在右側的聽覺丘腦放射鈎在聽覺表現較好，因此我們推測女生的聽覺敏感度可能表現得比男生好。
這是第一項利用擴散頻譜磁振造影資料來進行不同平均表觀傳播磁振造影係數之間的相關性之研究，以及第一項利用擴散頻譜磁振造影資料所產生的平均表觀傳播磁振造影係數所進行的男女大腦為結構差異分析觀察研究。有了對於這些係數的知識，可以讓我們在未來使用擴散係數時避免共線性的問題。雖然我的結果也和同領域的其他研究結果有點不同，但這些結果也對於男女腦部為結構差異研究提供了良好的基礎，未來也可以針對行為方面的表現來探究男女之差異。

關鍵字

擴散張量磁振造影 ； 擴散頻譜造影 ； 擴散峰度磁振造影 ； 平均表觀傳播系數 ； 性別差異

並列摘要

This study is a systematic correlation analysis across different diffusion MRI models to better understand the new diffusion indices derived from Mean Apparent Propagator (MAP) MRI and the application to sex differentiation. MAP-MRI has been recently proposed to reconstruct diffusion indices from high angular resolution diffusion data including generalized fractional anisotropy (GFA), fractional anisotropy (FA), axial diffusivity (AD), radial diffusivity (RD), mean diffusivity (MD), mean apparent length (AML), effective diameter of mean apparent area (AMAD), effective diameter of mean apparent volume (AMVD), parallel non-Gaussianity (NGP), orthogonal non-Gaussianity (NGO), and non-Gaussianity (NG). As compared to the first 5 indices, the other indices are less understood. To understand these diffusion indices, this study has two parts. First, we performed a systematic correlation study to clarify the relationships between them (Experiment 1). Second, we applied these MAP-MRI indices to compare the brain microstructure differentiation between males and females (Experiment 2).
Diffusion models, including diffusion kurtosis imaging (DKI) and MAP-MRI, were evaluated using DSI datasets obtained in 122 healthy adults which were used to build a diffusion spectrum imaging template. We used whole brain tract-based automatic analysis to obtain profiles of the 14 diffusion indices along 76 tract bundles for each DSI dataset. In experiment 1, systematic correlation analysis was performed between each pairs of diffusion indices on each tract bundle. The definition of the correlation between two indices was the average of the z-values in each tract bundle. With the correlation between each pairs of diffusion indices, we also applied a hierarchical clustering analysis to group these indices. In experiment 2, we compared 61 female and 61 male participants using stepwise two-sample t-test on the MAP-MRI indices.
In experiment 1, we found the associations can be clustered into three groups. First, NG, NGO and NGP, second, AD, AML and axial kurtosis (AK), third, GFA, FA, RD, MD, AMAD, AMVD, radial kurtosis (RK) and mean kurtosis (MK). In experiment 2, men showed better microstructural property in the left uncinate fasciculus, bilateral medial lemnisci, the left thalamic radiation postcentral and callosal fibers of hippocampus, while women showed better microstructural property in the right thalamic radiation auditory and callosal fibers of the dorsal lateral prefrontal cortex.
According to the definition of these indices the relationships of AD and AML, RD and AMAD, MD and AMVD have met our expectation. The reason why diffusion kurtosis indices are less similar to non-Gaussianity might be because that the characterization of non-Gaussianity is different from diffusion kurtosis. The characterization of diffusion kurtosis indices combines information from both tensor and kurtosis parts, whereas non-Gaussianity only considers the non-Gaussian part. In experiment 2, according to the connected ROIs of the tract and the function of the tract, we speculate that males have greater coordination of body than females because of the better microstructural property in bilateral medial lemnisci, and females have greater hearing sensitivity than males because of the better microstructural property in the right auditory radiation.
This is the first study to clarify associations among different MAP-MRI derived diffusion indices in DSI datasets, and the first study to investigate the sexual differentiation using MAP-MRI derived diffusion indices obtained with DSI. With the knowledge of the correlation results from Experiment 1, we can select the indices of interest and avoid the problem of collinearity. Our results in Experiment 2 agree partially with previous studies on sexual dimorphism and provide a sound basis for further studies on behavioral and possible clinical correlates.

並列關鍵字

Diffusion tensor imaging ； Diffusion spectrum imaging ； Diffusion kurtosis imaging ； mean apparent propagator ； sexual differentiation

參考文獻

Chen, Y. J., Lo, Y. C., Hsu, Y. C., Fan, C. C., Hwang, T. J., Liu, C. M., . . . Tseng, W. Y. (2015). Automatic whole brain tract-based analysis using predefined tracts in a diffusion spectrum imaging template and an accurate registration strategy. Hum Brain Mapp, 36(9), 3441-3458. doi:10.1002/hbm.22854

Ozarslan, E., Koay, C. G., Shepherd, T. M., Komlosh, M. E., Irfanoglu, M. O., Pierpaoli, C., & Basser, P. J. (2013). Mean apparent propagator (MAP) MRI: a novel diffusion imaging method for mapping tissue microstructure. Neuroimage, 78, 16-32. doi:10.1016/j.neuroimage.2013.04.016

Wedeen, V. J., Wang, R. P., Schmahmann, J. D., Benner, T., Tseng, W. Y., Dai, G., . . . de Crespigny, A. J. (2008). Diffusion spectrum magnetic resonance imaging (DSI) tractography of crossing fibers. Neuroimage, 41(4), 1267-1277. doi:10.1016/j.neuroimage.2008.03.036

Alexander, A. L., Hurley, S. A., Samsonov, A. A., Adluru, N., Hosseinbor, A. P., Mossahebi, P., . . . Field, A. S. (2011). Characterization of cerebral white matter properties using quantitative magnetic resonance imaging stains. Brain Connect, 1(6), 423-446. doi:10.1089/brain.2011.0071

Allen, J. S., Damasio, H., Grabowski, T. J., Bruss, J., & Zhang, W. (2003). Sexual dimorphism and asymmetries in the gray-white composition of the human cerebrum. Neuroimage, 18(4), 880-894.