some questions about reconstruction, correction, loading ROIs, and registration
Elijah
Dear Dr. Yeh
I am a graduate student new to diffusion image processing and DSI Studio. I have a few questions and would like to seek your opinions.
1. I noticed that there are several differences between DSI studio version-2012.07 and version-2021.05. Specifically, there are several reconstruction options not available in the 2021 version, such as DSI Hamming filter, ODF Tessellation, half-sphere scheme, and Number of fibers resolved. I wonder if the options are hidden and could be activated somewhere, and if not, whether skipping the options would have any impact on the reconstruction results. I also wonder what they mean, but couldn’t find any documentation describing them, so if you have any simple reference on them, please let me know.
2. I wonder if there is any tool for correction for head motion, eddy current distortion, and susceptibility-induced distortion in the DSI Studio. Our lab’s diffusion-weighted imaging data was acquired using a twice-refocused balanced spin-echo diffusion echo-planar imaging sequence with the half-sphere sampling scheme (a total of 102 diffusion gradient vectors was applied).
3. I try to seek any potential streamlines between the locus coeruleus and trans-entorhinal/para-hippocampal cortex. For this purpose, I load some ROI NIFTI files provided from open resources (one is locus coeruleus atlas [1], the other is locus coeruleus-transentorhinal cortex bundle atlas [2]). As far as I know, all of them are registered into the MNI152 space; therefore, I used [Regions Misc][Open MNI region] to load them into the native diffusion space on fiber tracking step. However, one of them [1] was loaded successfully into the native space in this way while the other was not (and no error message popped out, either). Even though [2] could be established by loading it as slices ([Slices][Insert MNI Images]) and making the ROI indirectly ([Regions Misc][Modify Region][Threshold]), I am still uncertain if it is a correct way, and wonder why I couldn’t load it directly.
4. I am also curious about the differences between loading a T1-weighted image by using [Slices][Insert T1w/T2w Images] on the fiber tracking step and registering a diffusion image into the T1-weighted image by [Edit][Resample to T1w/T2w] on the reconstruction step. I discover that the image resolution reconstructed by the latter was higher, and the shape of the ROI loaded was more plausible, but I don’t know if it is the better way to perform fiber tracking.
I will very appreciate your help, thank you.
Frank Yeh
I just found Gmail sendtseveral discussion threads to my SPAM folder, and I missed your thread.
I am a graduate student new to diffusion image processing and DSI Studio. I have a few questions and would like to seek your opinions.
1. I noticed that there are several differences between DSI studio version-2012.07 and version-2021.05. Specifically, there are several reconstruction options not available in the 2021 version, such as DSI Hamming filter, ODF Tessellation, half-sphere scheme, and Number of fibers resolved. I wonder if the options are hidden and could be activated somewhere, and if not, whether skipping the options would have any impact on the reconstruction results. I also wonder what they mean, but couldn’t find any documentation describing them, so if you have any simple reference on them, please let me know.
2. I wonder if there is any tool for correction for head motion, eddy current distortion, and susceptibility-induced distortion in the DSI Studio. Our lab’s diffusion-weighted imaging data was acquired using a twice-refocused balanced spin-echo diffusion echo-planar imaging sequence with the half-sphere sampling scheme (a total of 102 diffusion gradient vectors was applied).
3. I try to seek any potential streamlines between the locus coeruleus and trans-entorhinal/para-hippocampal cortex. For this purpose, I load some ROI NIFTI files provided from open resources (one is locus coeruleus atlas [1], the other is locus coeruleus-transentorhinal cortex bundle atlas [2]). As far as I know, all of them are registered into the MNI152 space; therefore, I used [Regions Misc][Open MNI region] to load them into the native diffusion space on fiber tracking step. However, one of them [1] was loaded successfully into the native space in this way while the other was not (and no error message popped out, either). Even though [2] could be established by loading it as slices ([Slices][Insert MNI Images]) and making the ROI indirectly ([Regions Misc][Modify Region][Threshold]), I am still uncertain if it is a correct way, and wonder why I couldn’t load it directly.
4. I am also curious about the differences between loading a T1-weighted image by using [Slices][Insert T1w/T2w Images] on the fiber tracking step and registering a diffusion image into the T1-weighted image by [Edit][Resample to T1w/T2w] on the reconstruction step. I discover that the image resolution reconstructed by the latter was higher, and the shape of the ROI loaded was more plausible, but I don’t know if it is the better way to perform fiber tracking.
I will very appreciate your help, thank you.