VBM 的輸出

100 views
Skip to first unread message

andy...@gmail.com

unread,
Aug 21, 2017, 1:12:42 AM8/21/17
to 成大心智影像研究中心-線上諮詢服務
想請問vbm能否輸出數值資料,以利後續用別的軟體進行分析?例如像Freesurfer 輸出的檔案?

葉丁瑞

unread,
Aug 21, 2017, 3:48:43 AM8/21/17
to YQ Chen, 成大心智影像研究中心-線上諮詢服務
永銓您好

你若是要extract ROI的話,建議用CAT12。


丁瑞

Yeh Ting-Jui
Member of Mind Research and Imaging  Center, National  Cheng Kung University
No.1,University Road,Tainan City 701,Taiwan
       +886(6)2353535#6269
Fax: +886(6)2004848
Email : dingruey88@gmail.com

2017-08-21 13:12 GMT+08:00 <andy...@gmail.com>:
想請問vbm能否輸出數值資料,以利後續用別的軟體進行分析?例如像Freesurfer 輸出的檔案?

--
您已訂閱「Google 網上論壇」的「成大心智影像研究中心-線上諮詢服務」群組,因此我們特別傳送這封郵件通知您。
如要取消訂閱這個群組並停止接收來自這個群組的郵件,請傳送電子郵件到 fmrincku+unsubscribe@googlegroups.com
如需更多選項,請前往:https://groups.google.com/d/optout

andy...@gmail.com

unread,
Aug 22, 2017, 3:14:33 AM8/22/17
to 成大心智影像研究中心-線上諮詢服務, andy...@gmail.com
Ding-Ruey Yeh於 2017年8月21日星期一 UTC+8下午3時48分43秒寫道:
> 永銓您好
>
>
> 你若是要extract ROI的話,建議用CAT12。
>
>
> http://dbm.neuro.uni-jena.de/cat12/CAT12-Manual.pdf
>
>
> 丁瑞
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> Yeh Ting-Jui
> Member of Mind Research and Imaging  Center, National  Cheng Kung University
> No.1,University Road,Tainan City 701,Taiwan
> Tel: +886(6)2008114#13;
>        +886(6)2353535#6269
> Fax: +886(6)2004848
> Email : dingruey88...@gmail.com
>
>
> 2017-08-21 13:12 GMT+08:00 <andy...@gmail.com>:
> 想請問vbm能否輸出數值資料,以利後續用別的軟體進行分析?例如像Freesurfer 輸出的檔案?
>
>
>
> --
>
> 您已訂閱「Google 網上論壇」的「成大心智影像研究中心-線上諮詢服務」群組,因此我們特別傳送這封郵件通知您。
>
> 如要取消訂閱這個群組並停止接收來自這個群組的郵件,請傳送電子郵件到 fmrincku+u...@googlegroups.com
>
> 如需更多選項,請前往:https://groups.google.com/d/optout

您好, 我想再請問如果我想要在VBM、Freesurfer 或cat 軟體上分析資料時把template 改成我自己的, 這樣可行嗎? 如果可行該怎麼做呢?

葉丁瑞

unread,
Aug 22, 2017, 3:48:58 AM8/22/17
to YQ Chen, 成大心智影像研究中心-線上諮詢服務
過去曾想過這個問題,改成自己的template,沒有成功。

建議於該軟體的forum上詢問。



丁瑞

Yeh Ting-Jui
Member of Mind Research and Imaging  Center, National  Cheng Kung University
No.1,University Road,Tainan City 701,Taiwan
       +886(6)2353535#6269
Fax: +886(6)2004848
Email : dingruey88@gmail.com

andy...@gmail.com

unread,
Aug 22, 2017, 4:03:02 AM8/22/17
to 成大心智影像研究中心-線上諮詢服務, andy...@gmail.com
學長您好:

在這個論壇上沒有找到可以發問的地方, 請問是需要有權限嗎?

Ding-Ruey Yeh於 2017年8月22日星期二 UTC+8下午3時48分58秒寫道:
> 過去曾想過這個問題,改成自己的template,沒有成功。
>
>
> 建議於該軟體的forum上詢問。
>
>
> freesurfer forum:http://www.mail-archive.com/frees...@nmr.mgh.harvard.edu/
>
>
>
>
> 丁瑞
>
>
>
>
>
>
>
>
>
>
>
>
>
> Yeh Ting-Jui
> Member of Mind Research and Imaging  Center, National  Cheng Kung University
> No.1,University Road,Tainan City 701,Taiwan
> Tel: +886(6)2008114#13;
>        +886(6)2353535#6269
> Fax: +886(6)2004848
> Email : dingruey88...@gmail.com

葉丁瑞

unread,
Aug 22, 2017, 4:56:17 AM8/22/17
to YQ Chen, 成大心智影像研究中心-線上諮詢服務
使用說明在這裡。


Using Freesurfer
To post a message to all the list members, send email to frees...@nmr.mgh.harvard.edu.

You can subscribe to the list, or change your existing subscription, in the sections below.

丁瑞


Yeh Ting-Jui
Member of Mind Research and Imaging  Center, National  Cheng Kung University
No.1,University Road,Tainan City 701,Taiwan
       +886(6)2353535#6269
Fax: +886(6)2004848
Email : dingruey88@gmail.com

andy...@gmail.com

unread,
Sep 11, 2017, 3:02:10 AM9/11/17
to 成大心智影像研究中心-線上諮詢服務
andy...@gmail.com於 2017年8月21日星期一 UTC+8下午1時12分42秒寫道:
> 想請問vbm能否輸出數值資料,以利後續用別的軟體進行分析?例如像Freesurfer 輸出的檔案?

您好:

我想再請問我從VBM 做完前處理non-linear only匯出的檔案(m0wp1XX)去抽取某個ROI 裡面的值去求平均得到的數字, 這個值是代表體積嗎? 這個值的單位又是什麼?

葉丁瑞

unread,
Sep 11, 2017, 3:50:21 AM9/11/17
to YQ Chen, 成大心智影像研究中心-線上諮詢服務
您好


 vbm進行 Estimate and write分析會若選用Tissue Probability Maps (TPMs)跑出m0wp1* for gray matter and m0wp2* for white matter的檔案。

所以我想你取得的數值應該是Probability Maps,數值介於0-1之間。

丁瑞

Yeh Ting-Jui
Member of Mind Research and Imaging  Center, National  Cheng Kung University
No.1,University Road,Tainan City 701,Taiwan
       +886(6)2353535#6269
Fax: +886(6)2004848
Email : dingruey88@gmail.com

Ding-Ruey Yeh

unread,
Sep 11, 2017, 5:25:12 AM9/11/17
to 成大心智影像研究中心-線上諮詢服務
第27頁
“Modulated normalized” gives two different options:下面敘述

若你要計算absolute amount,在GM、WM、CSF的Modulated normalized選擇Affine+non-linear,

然後寫matlab 圈選ROI,"加總"GM、WM和CSF的volume,就會得出,ROI裡面GM、WM和CSF的體積大小,單位為mm^3。

給你參考

reade...@gmail.com

unread,
Oct 16, 2017, 4:09:55 AM10/16/17
to 成大心智影像研究中心-線上諮詢服務
Ding-Ruey Yeh於 2017年9月11日星期一 UTC+8下午3時50分21秒寫道:
> 您好
>
>
> 依照 VBM的手冊:http://dbm.neuro.uni-jena.de/vbm8/vbm8-manual.pdf
>
>
>
>  vbm進行 Estimate and write分析會若選用Tissue Probability Maps (TPMs)跑出m0wp1* for gray matter and m0wp2* for white matter的檔案。
>
>
>
> 所以我想你取得的數值應該是Probability Maps,數值介於0-1之間。
>
>
> 丁瑞
>
>
>
>
>
>
>
>
>
>
>
>
>
> Yeh Ting-Jui
> Member of Mind Research and Imaging  Center, National  Cheng Kung University
> No.1,University Road,Tainan City 701,Taiwan
> Tel: +886(6)2008114#13;
>        +886(6)2353535#6269
> Fax: +886(6)2004848
> Email : dingruey88...@gmail.com
>
>
> 2017-09-11 15:02 GMT+08:00 <andy...@gmail.com>:
> andy...@gmail.com於 2017年8月21日星期一 UTC+8下午1時12分42秒寫道:
>
> > 想請問vbm能否輸出數值資料,以利後續用別的軟體進行分析?例如像Freesurfer 輸出的檔案?
>
>
>
> 您好:
>
>
>
> 我想再請問我從VBM 做完前處理non-linear only匯出的檔案(m0wp1XX)去抽取某個ROI 裡面的值去求平均得到的數字, 這個值是代表體積嗎? 這個值的單位又是什麼?
>
>
>
> --
>
> 您已訂閱「Google 網上論壇」的「成大心智影像研究中心-線上諮詢服務」群組,因此我們特別傳送這封郵件通知您。
>
> 如要取消訂閱這個群組並停止接收來自這個群組的郵件,請傳送電子郵件到 fmrincku+u...@googlegroups.com
>
> 如需更多選項,請前往:https://groups.google.com/d/optout

請問一下
這步驟應該已經經過modulated
這樣是否已經不能算是Probability Maps了?
VBM的手冊是這樣說
produces tissue class images in alignment with the template, but multiplies the voxel values by the non-linear components only.

這樣算是probability乘上voxel values?

另外一條回覆提到如果要算absolute amount要用affine+non-linear,可以得到mm^3

那如果是non-linear的情況下手冊說allows comparing the absolute amount of tissue corrected for individual brain sizes
這樣是得到類似residual的數字嗎..(也就是無單位)

葉丁瑞

unread,
Oct 16, 2017, 5:21:38 AM10/16/17
to M.H. Yang, 成大心智影像研究中心-線上諮詢服務
您好

您可以再整理一下你的問題嗎?感覺有點亂了。

以vbm的結果來說,您應該先注意的是有無nomalization。(在GM、WM、CSF的native space選擇:none,為無normalization)

加總ROI裡的每個voxel大小*機率,即可以得到volume size。

邏輯上是這樣子計算。

丁瑞


Yeh Ting-Jui
Member of Mind Research and Imaging  Center, National  Cheng Kung University
No.1,University Road,Tainan City 701,Taiwan
       +886(6)2353535#6269
Fax: +886(6)2004848
Email : dingruey88@gmail.com

Message has been deleted

reade...@gmail.com

unread,
Oct 16, 2017, 10:36:55 PM10/16/17
to 成大心智影像研究中心-線上諮詢服務
Ding-Ruey Yeh於 2017年10月16日星期一 UTC+8下午5時21分38秒寫道:

> 您好
>
>
> 您可以再整理一下你的問題嗎?感覺有點亂了。
>
>
> 以vbm的結果來說,您應該先注意的是有無nomalization。(在GM、WM、CSF的native space選擇:none,為無normalization)
>
>
> 加總ROI裡的每個voxel大小*機率,即可以得到volume size。
>
>
> 邏輯上是這樣子計算。
>
>
> 丁瑞
>
>
>


這邊看不懂
https://imgur.com/xHCRgSD.jpg
這邊指的normalization不同於圖中的normalization嗎?

還是這邊的意思是,想要得到volume size就要在writing option多讓他輸出native space才能算?

葉丁瑞

unread,
Oct 17, 2017, 3:47:22 AM10/17/17
to M.H. Yang, 成大心智影像研究中心-線上諮詢服務
您好

VBM的modulation分兩種:1. Affine + non-linear, 2. Non-linear Only。做群組比較的VBM分析時,每個T1 image基本分析應採用2(non-linear only)。

single case VBM分析時,如果用non-linear only的資料,體積算出來是錯誤的,要使用Affine + non-linear(mwep1xx.nii)。

因為non-linear only的modulation過程,僅考慮非線性變形量,沒有進行affine剛體轉換裡面的scaling(放大/縮小),所以體積大小的轉換資訊沒有被加進modulated影像裡(頭大頭小這件事),而造成計算絕對體積時出現錯誤。

相反的,如果要進行正規VBM分析,因為是一群人的平均,或是一群人比另一群人,要將頭大小這件事忽略不考慮,亦即,normalized with brain size,所以反而應該使用non-linear only的選項。即是m0wrp1xx.nii。

**0wrp1xx.nii是Probability Map? 它是組織機率,乘上一個體積形變量數值經過segmentation,VBM會產生三個影像,分別是GM/WM/CSF的”機率圖”,所以這三個map的上的數值應該都是0-1。代表是該種tissue的機率。但單單是這樣的資料是無法拿來進行體積比較的分析,而需要進一步將這個組織機率圖,乘上一個體積形變量數值。VBM裡稱之為modulation。

non-linear紀錄的是非線性"局部變形"的量,沒有考慮總體積的觀念。

丁瑞

Yeh Ting-Jui
Member of Mind Research and Imaging  Center, National  Cheng Kung University
No.1,University Road,Tainan City 701,Taiwan
       +886(6)2353535#6269
Fax: +886(6)2004848
Email : dingruey88@gmail.com

Reply all
Reply to author
Forward
0 new messages