Adjusting the neuroimaging statistical inferences for nonstationarity.
In neuroimaging cluster-based inference has generally been found to be more powerful than voxel-wise inference. However standard cluster-based methods assume stationarity (constant smoothness), while under nonstationarity clusters are larger in smooth regions just by chance, making false positive ri...
| Main Authors: | , , |
|---|---|
| 格式: | Conference item |
| 出版: |
2009
|
| _version_ | 1826278519647240192 |
|---|---|
| author | Salimi-Khorshidi, G Smith, S Nichols, T |
| author_facet | Salimi-Khorshidi, G Smith, S Nichols, T |
| author_sort | Salimi-Khorshidi, G |
| collection | OXFORD |
| description | In neuroimaging cluster-based inference has generally been found to be more powerful than voxel-wise inference. However standard cluster-based methods assume stationarity (constant smoothness), while under nonstationarity clusters are larger in smooth regions just by chance, making false positive risk spatially variant. Hayasaka et al. proposed a Random Field Theory (RFT) based nonstationarity adjustment for cluster inference and validated the method in terms of controlling the overall family-wise false positive rate. The RFT-based methods, however, have never been directly assessed in terms of homogeneity of local false positive risk. In this work we propose a new cluster size adjustment that accounts for local smoothness, based on local empirical cluster size distributions and a two-pass permutation method. We also propose a new approach to measure homogeneity of local false positive risk, and use this method to compare the RFT-based and our new empirical adjustment methods. We apply these techniques to both cluster-based and a related inference, threshold-free cluster enhancement (TFCE). Using simulated and real data we confirm the expected heterogeneity in false positive risk with unadjusted cluster inference but find that RFT-based adjustment does not fully eliminate heterogeneity; we also observe that our proposed empirical adjustment dramatically increases the homogeneity and TFCE inference is generally quite robust to nonstationarity. |
| first_indexed | 2024-03-06T23:45:08Z |
| format | Conference item |
| id | oxford-uuid:70a21414-3b51-43c4-b33c-6feb24820037 |
| institution | University of Oxford |
| last_indexed | 2024-03-06T23:45:08Z |
| publishDate | 2009 |
| record_format | dspace |
| spelling | oxford-uuid:70a21414-3b51-43c4-b33c-6feb248200372022-03-26T19:38:27ZAdjusting the neuroimaging statistical inferences for nonstationarity.Conference itemhttp://purl.org/coar/resource_type/c_5794uuid:70a21414-3b51-43c4-b33c-6feb24820037Symplectic Elements at Oxford2009Salimi-Khorshidi, GSmith, SNichols, TIn neuroimaging cluster-based inference has generally been found to be more powerful than voxel-wise inference. However standard cluster-based methods assume stationarity (constant smoothness), while under nonstationarity clusters are larger in smooth regions just by chance, making false positive risk spatially variant. Hayasaka et al. proposed a Random Field Theory (RFT) based nonstationarity adjustment for cluster inference and validated the method in terms of controlling the overall family-wise false positive rate. The RFT-based methods, however, have never been directly assessed in terms of homogeneity of local false positive risk. In this work we propose a new cluster size adjustment that accounts for local smoothness, based on local empirical cluster size distributions and a two-pass permutation method. We also propose a new approach to measure homogeneity of local false positive risk, and use this method to compare the RFT-based and our new empirical adjustment methods. We apply these techniques to both cluster-based and a related inference, threshold-free cluster enhancement (TFCE). Using simulated and real data we confirm the expected heterogeneity in false positive risk with unadjusted cluster inference but find that RFT-based adjustment does not fully eliminate heterogeneity; we also observe that our proposed empirical adjustment dramatically increases the homogeneity and TFCE inference is generally quite robust to nonstationarity. |
| spellingShingle | Salimi-Khorshidi, G Smith, S Nichols, T Adjusting the neuroimaging statistical inferences for nonstationarity. |
| title | Adjusting the neuroimaging statistical inferences for nonstationarity. |
| title_full | Adjusting the neuroimaging statistical inferences for nonstationarity. |
| title_fullStr | Adjusting the neuroimaging statistical inferences for nonstationarity. |
| title_full_unstemmed | Adjusting the neuroimaging statistical inferences for nonstationarity. |
| title_short | Adjusting the neuroimaging statistical inferences for nonstationarity. |
| title_sort | adjusting the neuroimaging statistical inferences for nonstationarity |
| work_keys_str_mv | AT salimikhorshidig adjustingtheneuroimagingstatisticalinferencesfornonstationarity AT smiths adjustingtheneuroimagingstatisticalinferencesfornonstationarity AT nicholst adjustingtheneuroimagingstatisticalinferencesfornonstationarity |