help > Result interpretation help!
Showing 1-6 of 6 posts
Jun 11, 2021 10:06 AM | Aishvarya Shri Rajasimman - AIIMS, Bhubaneswar
Result interpretation help!
Dear all,
After analysis of resting state fmri between cases and controls in conn I have doubts regarding the interpretation.
After setting the contrast to [1 -1], i.e., cases> controls, for each selected seed there are few significant connections that are being displayed in either blue or red color (Attaching a picture below).
Please guide me,
here in the tabulation below (with contrast of [1 -1]),
1. Does the beta value and T score here represent the difference between cases and controls?
2. What does the plus or minus sign of T score indicate?
3. What is the significance of the color of the displayed connection ? (Because for a same seed, I could observe the color is red irrespective of setting cases> controls or controls>cases)
I would be grateful. Please guide me through.
Regards,
Aishvarya
After analysis of resting state fmri between cases and controls in conn I have doubts regarding the interpretation.
After setting the contrast to [1 -1], i.e., cases> controls, for each selected seed there are few significant connections that are being displayed in either blue or red color (Attaching a picture below).
Please guide me,
here in the tabulation below (with contrast of [1 -1]),
1. Does the beta value and T score here represent the difference between cases and controls?
2. What does the plus or minus sign of T score indicate?
3. What is the significance of the color of the displayed connection ? (Because for a same seed, I could observe the color is red irrespective of setting cases> controls or controls>cases)
I would be grateful. Please guide me through.
Regards,
Aishvarya
Jun 11, 2021 10:06 PM | Alfonso Nieto-Castanon - Boston University
RE: Result interpretation help!
Dear Aishvarya
The beta/T stats represent the results from a two-sample t-test comparing cases>controls, so the signs of the beta/T values will be positive if connectivity is higher in cases compared to controls and negative otherwise (the stats are two-sided / two-tailed so the p-values will not change when evaluating controls>cases but the sign of the beta/T statistics will of course flip). The color of the lines, on the other hand, simply represents the sign of the connectivity on average in both cases and controls, so that will not change when testing cases>controls vs. controls>cases (in your case, for example, FP and Putamen show positive connectivity on average, with higher connectivity in controls compared to cases; in the example GUI shown in your attachment you may switch your contrast from 'Difference cases > controls' to 'Any effects (F-test)' if you want to see a bar display of the connectivity within cases and controls separately)
Hope this helps
Alfonso
Originally posted by Aishvarya Shri Rajasimman:
The beta/T stats represent the results from a two-sample t-test comparing cases>controls, so the signs of the beta/T values will be positive if connectivity is higher in cases compared to controls and negative otherwise (the stats are two-sided / two-tailed so the p-values will not change when evaluating controls>cases but the sign of the beta/T statistics will of course flip). The color of the lines, on the other hand, simply represents the sign of the connectivity on average in both cases and controls, so that will not change when testing cases>controls vs. controls>cases (in your case, for example, FP and Putamen show positive connectivity on average, with higher connectivity in controls compared to cases; in the example GUI shown in your attachment you may switch your contrast from 'Difference cases > controls' to 'Any effects (F-test)' if you want to see a bar display of the connectivity within cases and controls separately)
Hope this helps
Alfonso
Originally posted by Aishvarya Shri Rajasimman:
Dear all,
After analysis of resting state fmri between cases and controls in conn I have doubts regarding the interpretation.
After setting the contrast to [1 -1], i.e., cases> controls, for each selected seed there are few significant connections that are being displayed in either blue or red color (Attaching a picture below).
Please guide me,
here in the tabulation below (with contrast of [1 -1]),
1. Does the beta value and T score here represent the difference between cases and controls?
2. What does the plus or minus sign of T score indicate?
3. What is the significance of the color of the displayed connection ? (Because for a same seed, I could observe the color is red irrespective of setting cases> controls or controls>cases)
I would be grateful. Please guide me through.
Regards,
Aishvarya
After analysis of resting state fmri between cases and controls in conn I have doubts regarding the interpretation.
After setting the contrast to [1 -1], i.e., cases> controls, for each selected seed there are few significant connections that are being displayed in either blue or red color (Attaching a picture below).
Please guide me,
here in the tabulation below (with contrast of [1 -1]),
1. Does the beta value and T score here represent the difference between cases and controls?
2. What does the plus or minus sign of T score indicate?
3. What is the significance of the color of the displayed connection ? (Because for a same seed, I could observe the color is red irrespective of setting cases> controls or controls>cases)
I would be grateful. Please guide me through.
Regards,
Aishvarya
Jun 12, 2021 06:06 AM | Aishvarya Shri Rajasimman - AIIMS, Bhubaneswar
RE: Result interpretation help!
Originally posted by Alfonso Nieto-Castanon:
Dear Aishvarya
The beta/T stats represent the results from a two-sample t-test comparing cases>controls, so the signs of the beta/T values will be positive if connectivity is higher in cases compared to controls and negative otherwise (the stats are two-sided / two-tailed so the p-values will not change when evaluating controls>cases but the sign of the beta/T statistics will of course flip). The color of the lines, on the other hand, simply represents the sign of the connectivity on average in both cases and controls, so that will not change when testing cases>controls vs. controls>cases (in your case, for example, FP and Putamen show positive connectivity on average, with higher connectivity in controls compared to cases; in the example GUI shown in your attachment you may switch your contrast from 'Difference cases > controls' to 'Any effects (F-test)' if you want to see a bar display of the connectivity within cases and controls separately)
Hope this helps
Alfonso
The beta/T stats represent the results from a two-sample t-test comparing cases>controls, so the signs of the beta/T values will be positive if connectivity is higher in cases compared to controls and negative otherwise (the stats are two-sided / two-tailed so the p-values will not change when evaluating controls>cases but the sign of the beta/T statistics will of course flip). The color of the lines, on the other hand, simply represents the sign of the connectivity on average in both cases and controls, so that will not change when testing cases>controls vs. controls>cases (in your case, for example, FP and Putamen show positive connectivity on average, with higher connectivity in controls compared to cases; in the example GUI shown in your attachment you may switch your contrast from 'Difference cases > controls' to 'Any effects (F-test)' if you want to see a bar display of the connectivity within cases and controls separately)
Hope this helps
Alfonso
Dear Alfonso,
Thank you so much! Doubts cleared :))
I should also mention that your name is the one
which I see in most of the replies. Thanks from all of us for being
there and helping us out.
Regards,
Aishvarya
Jun 26, 2021 02:06 PM | Aishvarya Shri Rajasimman - AIIMS, Bhubaneswar
RE: Result interpretation help!
Originally posted by Aishvarya Shri Rajasimman:
In ROI analysis, the picture on right shows connections. However, the left and right are not as radiological right and left. Is it how the display is or there is some error that I have run into? In the picture below, seed is left occipital pole , but the picture depicts on right. Please clarify.
Regards,
Aishvarya
Dear all,
After analysis of resting state fmri between cases and controls in conn I have doubts regarding the interpretation.
After setting the contrast to [1 -1], i.e., cases> controls, for each selected seed there are few significant connections that are being displayed in either blue or red color (Attaching a picture below).
Please guide me,
here in the tabulation below (with contrast of [1 -1]),
1. Does the beta value and T score here represent the difference between cases and controls?
2. What does the plus or minus sign of T score indicate?
3. What is the significance of the color of the displayed connection ? (Because for a same seed, I could observe the color is red irrespective of setting cases> controls or controls>cases)
I would be grateful. Please guide me through.
Regards,
Aishvarya
Dear All, After analysis of resting state fmri between cases and controls in conn I have doubts regarding the interpretation.
After setting the contrast to [1 -1], i.e., cases> controls, for each selected seed there are few significant connections that are being displayed in either blue or red color (Attaching a picture below).
Please guide me,
here in the tabulation below (with contrast of [1 -1]),
1. Does the beta value and T score here represent the difference between cases and controls?
2. What does the plus or minus sign of T score indicate?
3. What is the significance of the color of the displayed connection ? (Because for a same seed, I could observe the color is red irrespective of setting cases> controls or controls>cases)
I would be grateful. Please guide me through.
Regards,
Aishvarya
In ROI analysis, the picture on right shows connections. However, the left and right are not as radiological right and left. Is it how the display is or there is some error that I have run into? In the picture below, seed is left occipital pole , but the picture depicts on right. Please clarify.
Regards,
Aishvarya
Jun 28, 2021 09:06 PM | Alfonso Nieto-Castanon - Boston University
RE: Result interpretation help!
Dear Aishvarya,
All default displays in CONN follow neurological (not radiological) convention, so left (in the image) is left (hemisphere), right is right (i.e. axial images displayed as if viewed from top, coronal images displayed as if viewed from back)
Hope this helps
Alfonso
Originally posted by Aishvarya Shri Rajasimman:
Dear All,
In ROI analysis, the picture on right shows connections. However, the left and right are not as radiological right and left. Is it how the display is or there is some error that I have run into? In the picture below, seed is left occipital pole , but the picture depicts on right. Please clarify.
Regards,
Aishvarya
All default displays in CONN follow neurological (not radiological) convention, so left (in the image) is left (hemisphere), right is right (i.e. axial images displayed as if viewed from top, coronal images displayed as if viewed from back)
Hope this helps
Alfonso
Originally posted by Aishvarya Shri Rajasimman:
Dear All,
In ROI analysis, the picture on right shows connections. However, the left and right are not as radiological right and left. Is it how the display is or there is some error that I have run into? In the picture below, seed is left occipital pole , but the picture depicts on right. Please clarify.
Regards,
Aishvarya
Jul 3, 2023 01:07 PM | sukhbinder Kumar
changing display from neurological to radiological convention
Dear All
I understand the defaut image display in Conn toolbox is neurological (left is left and right is right). How do I chnage this setting if I want to use radiological convention for display (that is left is right and right is left). Any help would be highly appreciated.
Many Thanks
Sukhbinder