磁共振頻譜(MRS)是一種非破壞性技術，不需要使用放射性物質或有害藥物，可以利用非侵入的方式獲取氫核自旋的共振訊號，分析物質的分子結構、組成和代謝物含量，通常在做MRS實驗時會搭配單點式頻譜(SVS)技術獲取單一體素內的頻譜訊號，單一體素是我們感興趣的範圍(VOI)，不同的VOI位置會影響到想要研究的頻譜訊號，以至於VOI的定位非常重要，而研究人員常常需要手動定位VOI，在過程中人為疏失是難以避免的情況，為了降低此類情況發生，自動定位變得不可或缺。
在過往研究中，我們在MATLAB的環境開發了一個自動定位系統，使用者必須透過MATLAB才能開啟使用，對於使用者來說若需額外安裝套件才能使用此應用程式，可能增加了不必要的繁瑣步驟，本研究致力於跳脫原先MATLAB環境，程式語言改為Python將系統打包成一個獨立執行檔，讓使用者不用透過其他套件就能使用且應用程式大小不會佔據太大空間。
為了驗證改寫在Python系統的自動定位演算法，我們將受試者T1影像對位到腦模板空間，中間會生成對位轉換器，此轉換器能將Python系統創建的voxel(〖VOI〗_Tem)反轉回T1空間(〖VOI〗_T1)，計算〖VOI〗_T1的數據再生成一個voxel(〖VOI〗_(cube_T1))，〖VOI〗_(cube_T1)再對位回腦模板空間得到〖VOI〗_(cube_Tem)。研究發現Python系統在腦模板生成的voxel(〖VOI〗_Tem)與〖VOI〗_(cube_Tem)的相似性極高，證明本研究的自動定位演算法有著很高的準確性。

Single voxel spectroscopy (SVS) is a technique for obtaining spectral signals within a single voxel of interest (VOI). The location of the VOI can affect the spectroscopic signals being studied, making VOI positioning very important. In previous research, we developed an automatic positioning system in the MATLAB environment. Users had to open it through MATLAB, and if they needed to install additional packages to use the application, it was redundant. This study aims to break away from the original MATLAB environment and change the programming language to Python, packaging the system into a standalone executable file. This allows users to use the application without the need for other packages, and the application size will not take up too much space.
To verify the automatic localization algorithm rewritten in Python system, we registered the T1 images of the subjects to the brain template space, generating a registration transformer in the process. This transformer can inverse the voxel (〖VOI〗_Tem) created by the Python system back to the T1 space (〖VOI〗_T1), calculate the data of 〖VOI〗_T1, and generate another voxel (〖VOI〗_(cube_T1)). 〖VOI〗_(cube_T1) was then registered to the brain template space to obtain 〖VOI〗_(cube_Tem). The study found that the similarity between 〖VOI〗_Tem and 〖VOI〗_(cube_Tem) was extremely high, proving the high accuracy of the automatic localization algorithm in this study.

[1] Babikian, T., Freier, M. C., Ashwal, S., Riggs, M. L., Burley, T., & Holshouser, B. A. (2006). MR spectroscopy: Predicting long‐term neuropsychological outcome following pediatric TBI. Journal of Magnetic Resonance Imaging: An Official Journal of the International Society for Magnetic Resonance in Medicine, 24(4), 801-811.
[2] van der Graaf, M. In vivo magnetic resonance spectroscopy: basic methodology and clinical applications. Eur Biophys J 39, 527–540 (2010). https://doi.org/10.1007/s00249-009-0517-y
[3] IMAIOS. (n.d.). Magnetic Resonance Spectroscopy - Single Voxel Spectroscopy. Retrieved from https://www.imaios.com/en/e-mri/magnetic-resonance-spectroscopy/single-voxel-spectroscopy
[4] Ricci PE, Pitt A, Keller PJ, Coons SW, Heiserman JE. Effect of voxel position on single-voxel MR spectroscopy findings. AJNR Am J Neuroradiol. 2000 Feb;21(2):367-74. PMID: 10696025; PMCID: PMC7975341.
[5] Tsuzuki, D., Jurcak, V., Singh, A. K., Okamoto, M., Watanabe, E., & Dan, I. (2007). Virtual spatial registration of stand-alone fNIRS data to MNI space. Neuroimage, 34(4), 1506-1518.
[6] Lead-DBS. (n.d.). About the MNI spaces. Retrieved from https://www.lead-dbs.org/about-the-mni-spaces/
[7] Collins, L., Zijdenbos, A. P., Kollokian, V., Sled, J. G., Kabani, N. J., Holmes, C. J., & Evans, A. C. The MNI brain and the Talairach atlas. NeuroImage 1998.
[8] happyhorizion. (n.d.). MNI Spac. Retrieved from https://blog.csdn.net/happyhorizion/article/details/79579453
[9] McGill University. (n.d.). Colin27. Retrieved from https://www.mcgill.ca/bic/software/tools-data-analysis/anatomical-mri/atlases/colin-27
[10] Kim JS, Lee JM, Lee YH, Kim JS, Kim IY, Kim SI. Intensity based affine registration including feature similarity for spatial normalization. Comput Biol Med. 2002 Sep;32(5):389-402. doi: 10.1016/s0010-4825(02)00024-0. PMID: 12102756.
[11] Friston, K.J.; Ashburner, J.; Frith, C.D.; Poline, J.-B.; Heather, J.D.; Frackowiak, R.S.J. Spatial registration and normalization of images. Hum. Brain Mapp. 1995, 3, 165–189.
[12] Garyfallidis E, Brett M, Amirbekian B, Rokem A, Van Der Walt S, Descoteaux M, Nimmo-Smith I. Dipy, a library for the analysis of diffusion MRI data, Front Neuroinformat 2014;8. https://doi.org/10.3389/fninf.2014.00008.
[13] S. Ille, H. Zhang, N. Stassen, M. Schwendner, A. Schroeder, B. Meyer, S. Krieg, Connectome analysis based on intraoperative awake mapping reveals new insights on language network organization,Brain and Spine,Volume 2, Supplement 2, 2022, 101274, ISSN 2772-5294, https://doi.org/10.1016/j.bas.2022.101274. (https://www.sciencedirect.com/science/article/pii/S2772529422004155)
[14] R. Beare, B. C. Lowekamp, Z. Yaniv, “Image Segmentation, Registration and Characterization in R with SimpleITK”, J Stat Software, 86(8), https://doi.org/10.18637/jss.v086.i08, 2018.
[15] Mezer, A., et al. (2016). "Multicenter reliability of diffusion tensor imaging." Brain connectivity, 6(9), 164-180.
[16] Yu-Jhou Gong. Automatic position of MR spectroscopy voxels.Taipei: National Taiwan University of Science and Technology;2015.
[17] PO-YU PENG.Toolbox for Automatic Localization of Volume of Interest in Single Voxel MRS. Taipei: National Taiwan University of Science and Technology;2016.
[18] Meier, B. (2019). Python GUI Programming Cookbook: Develop functional and responsive user interfaces with tkinter and PyQt5. Packt Publishing Ltd.
[19] Weng JS, Huang TY. Deriving a robust deep-learning model for subcortical brain segmentation by using a large-scale database: Preprocessing, reproducibility, and accuracy of volume estimation. NMR Biomed. 2022 Nov 23:e4880. doi: 10.1002/nbm.4880. (https://doi.org/10.1002/nbm.4880)
[20] MathWorks. (n.d.). Application Compiler App. Retrieved from https://www.mathworks.com/help/compiler/applicationcompiler-app.html