| 1. |
- de Pierrefeu, Amicie, et al.
(författare)
-
Identifying a neuroanatomical signature of schizophrenia, reproducible across sites and stages, using machine-learning with structured sparsity
- 2018
-
Ingår i: Acta Psychiatrica Scandinavica. - : John Wiley & Sons. - 0001-690X .- 1600-0447. ; 138, s. 571-580
-
Tidskriftsartikel (refereegranskat)abstract
- ObjectiveStructural MRI (sMRI) increasingly offers insight into abnormalities inherent to schizophrenia. Previous machine learning applications suggest that individual classification is feasible and reliable and, however, is focused on the predictive performance of the clinical status in cross‐sectional designs, which has limited biological perspectives. Moreover, most studies depend on relatively small cohorts or single recruiting site. Finally, no study controlled for disease stage or medication's effect. These elements cast doubt on previous findings’ reproducibility.MethodWe propose a machine learning algorithm that provides an interpretable brain signature. Using large datasets collected from 4 sites (276 schizophrenia patients, 330 controls), we assessed cross‐site prediction reproducibility and associated predictive signature. For the first time, we evaluated the predictive signature regarding medication and illness duration using an independent dataset of first‐episode patients.ResultsMachine learning classifiers based on neuroanatomical features yield significant intersite prediction accuracies (72%) together with an excellent predictive signature stability. This signature provides a neural score significantly correlated with symptom severity and the extent of cognitive impairments. Moreover, this signature demonstrates its efficiency on first‐episode psychosis patients (73% accuracy).ConclusionThese results highlight the existence of a common neuroanatomical signature for schizophrenia, shared by a majority of patients even from an early stage of the disorder.
|
|
| 2. |
- De Pierrefeu, Amicie, et al.
(författare)
-
Interpretable and stable prediction of schizophrenia on a large multisite dataset using machine learning with structured sparsity
- 2018
-
Ingår i: 2018 International Workshop on Pattern Recognition in Neuroimaging (PRNI). - : IEEE. - 9781538668597
-
Konferensbidrag (refereegranskat)abstract
- The use of machine-learning (ML) in neuroimaging offers new perspectives in early diagnosis and prognosis of brain diseases. Indeed, ML algorithms can jointly examine all brain features to capture complex relationships in the data in order to make inferences at a single-subject level. To deal with such high dimensional input and the associated risk of overfitting on the training data, a proper regularization (or feature selection) is required. Standard ℓ2-regularized predictors, such as Support Vector Machine, provide dense patterns of predictors. However, in the context of predictive disease signature discovery, it is now essential to understand the brain pattern that underpins the prediction. Despite ℓ1-regularized (sparse) has often been advocated as leading to more interpretable models, they generally lead to scattered and unstable patterns. We hypothesize that the integration of prior knowledge regarding the structure of the input images should improve the relevance and the stability of the predictive signature. Such structured sparsity can be obtained by combining together ℓ1 (possibly ℓ2) and Total variation (TV) penalties. We demonstrated the relevance of using ML with structured sparsity on a large multisite dataset of schizophrenia patients and controls. Using 3D maps of grey matter density, we obtained promising inter-site prediction performances. More importantly, we have uncovered a predictive signature of schizophrenia that is clinically interpretable and stable across resampling. This suggests that structured sparsity provides a major breakthrough over 'off-The-shelf' algorithms to perform a robust selection of important brain regions in the context of biomarkers discovery.
|
|
| 3. |
- de Zwarte, Sonja M. C., et al.
(författare)
-
Intelligence, educational attainment, and brain structure in those at familial high-risk for schizophrenia or bipolar disorder
- 2022
-
Ingår i: Human Brain Mapping. - : John Wiley & Sons. - 1065-9471 .- 1097-0193. ; 43:1, s. 414-430
-
Tidskriftsartikel (refereegranskat)abstract
- First-degree relatives of patients diagnosed with schizophrenia (SZ-FDRs) show similar patterns of brain abnormalities and cognitive alterations to patients, albeit with smaller effect sizes. First-degree relatives of patients diagnosed with bipolar disorder (BD-FDRs) show divergent patterns; on average, intracranial volume is larger compared to controls, and findings on cognitive alterations in BD-FDRs are inconsistent. Here, we performed a meta-analysis of global and regional brain measures (cortical and subcortical), current IQ, and educational attainment in 5,795 individuals (1,103 SZ-FDRs, 867 BD-FDRs, 2,190 controls, 942 schizophrenia patients, 693 bipolar patients) from 36 schizophrenia and/or bipolar disorder family cohorts, with standardized methods. Compared to controls, SZ-FDRs showed a pattern of widespread thinner cortex, while BD-FDRs had widespread larger cortical surface area. IQ was lower in SZ-FDRs (d = -0.42, p = 3 × 10-5 ), with weak evidence of IQ reductions among BD-FDRs (d = -0.23, p = .045). Both relative groups had similar educational attainment compared to controls. When adjusting for IQ or educational attainment, the group-effects on brain measures changed, albeit modestly. Changes were in the expected direction, with less pronounced brain abnormalities in SZ-FDRs and more pronounced effects in BD-FDRs. To conclude, SZ-FDRs and BD-FDRs show a differential pattern of structural brain abnormalities. In contrast, both had lower IQ scores and similar school achievements compared to controls. Given that brain differences between SZ-FDRs and BD-FDRs remain after adjusting for IQ or educational attainment, we suggest that differential brain developmental processes underlying predisposition for schizophrenia or bipolar disorder are likely independent of general cognitive impairment.
|
|
| 4. |
- Tahmasian, Masoud, et al.
(författare)
-
ENIGMA-Sleep : Challenges, opportunities, and the road map
- 2021
-
Ingår i: Journal of Sleep Research. - : Wiley. - 0962-1105 .- 1365-2869. ; 30:6
-
Forskningsöversikt (refereegranskat)abstract
- Neuroimaging and genetics studies have advanced our understanding of the neurobiology of sleep and its disorders. However, individual studies usually have limitations to identifying consistent and reproducible effects, including modest sample sizes, heterogeneous clinical characteristics and varied methodologies. These issues call for a large-scale multi-centre effort in sleep research, in order to increase the number of samples, and harmonize the methods of data collection, preprocessing and analysis using pre-registered well-established protocols. The Enhancing NeuroImaging Genetics through Meta-Analysis (ENIGMA) consortium provides a powerful collaborative framework for combining datasets across individual sites. Recently, we have launched the ENIGMA-Sleep working group with the collaboration of several institutes from 15 countries to perform large-scale worldwide neuroimaging and genetics studies for better understanding the neurobiology of impaired sleep quality in population-based healthy individuals, the neural consequences of sleep deprivation, pathophysiology of sleep disorders, as well as neural correlates of sleep disturbances across various neuropsychiatric disorders. In this introductory review, we describe the details of our currently available datasets and our ongoing projects in the ENIGMA-Sleep group, and discuss both the potential challenges and opportunities of a collaborative initiative in sleep medicine.
|
|
