To send content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about sending content to .
To send content items to your Kindle, first ensure firstname.lastname@example.org
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about sending to your Kindle.
Note you can select to send to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be sent to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
Numerous studies have applied novel multivariate statistical approaches to the analysis of brain alterations in patients with schizophrenia. However the diagnostic accuracy of the reported predictive models differs largely, making it difficult to evaluate the overall potential of these studies to inform clinical diagnosis.
We conducted a comprehensive literature search to identify all studies reporting performance of neuroimaging-based multivariate predictive models for the differentiation of patients with schizophrenia from healthy control subjects. The robustness of the results as well as the effect of potentially confounding continous variables (e.g. age, gender ratio, year of publication) was investigated.
The final sample consisted of n=37 studies studies including n=1491 patients with schizophrenia and n=1488 healthy controls. Metaanalysis of the complete sample showed a sensitivity of 80.7% (95%-CI: 77.0 to 83.9%) and a specificity of 80.2% (95%-CI: 83.3 to 76.7%). Separate analysis for the different imaging modalities showed similar diagnostic accuracy for the structural MRI studies (sensitivity 77.3%, specificity 78.7%), the fMRI studies (sensitivity 81.4%, specificity 82.4%) and resting-state fMRI studies (sensitivity 86.9%, specificity 80.3%). Moderator analysis showed significant effects of age of patients on sensitivity (p=0.021) and of positive-tonegative symptom ratio on specificity (p=0.028) indicating better diagnostic accuracy in older patients and patients with positive symptoms.
Our analysis indicate an overall sensitivity and overall specificity of around 80 % of neuroimaging-based predictive models for differentiating schizophrenic patients from healthy controls. The results underline the potential applicability of neuroimaging-based predictive models for the diagnosis of schizophrenia.
Previous studies have shown that structural brain changes are among the best-studied candidate markers for schizophrenia (SZ) along with global functional connectivity (FC) alterations of resting-state (RS) networks. Only few studies tried to combine these data domains to outperform unimodal pattern classification approaches. We aimed at distinguishing SZ patients from healthy controls (HC) at the single-subject level by applying multivariate pattern recognition analysis to both gray matter (GM) volume and FC measures.
The RS functional and structural MRI data from 74 HC and 71 patients with SZ were obtained from the publicly available COBRE database. The machine learning pipeline wrapped into repeated nested cross-validation was used to train a multi-modal diagnostic system and evaluate its generalization capacity in new subjects.
Both functional and structural classifiers were able to distinguish between HC and SZ patients with similar accuracies. The RS classifier was showing a slightly higher accuracy (75%) comparing to GM volume classifier (74.4%). Ensemble-based data fusion outperformed pattern classification based on single MRI modalities by reaching 76.6% accuracy, as determined by cross-validation. Further analysis showed that RS classification was less sensitive to age-related effects across the life span than GM volume.
Our findings suggest that age plays an important role in discriminating SZ patients from HC, but that RS is more robust towards age-differences compared to GM volume. Single neuroimaging modalities provide useful insight into brain function or structure, while multimodal fusion emphasizes the strength of each and provides higher accuracy in discriminating SZ patients from HC.
Everyday clinical routine is frequently challenged by difficulty to choose among differential diagnostic options, since many psychiatric disorders share similar phenotypes. E.g., borderline personality disorder (BPD) and schizophrenia (SZ) can both be associated with psychotic syndromes.
Our objective was to evaluate the effectiveness of combining sMRI data and pattern classification methods to differentiate between BPD and SZ.
We aim to introduce objective diagnostic measures to improve the reliability of clinical evaluations.
sMRI data of 114 female patients were used to train a multivariate disease classifier.
MR images were processed using voxel-based morphometry and high-dimensional registration to the MNI template. Grey matter volume maps were fed into a machine learning pipeline consisting of adjustment for possible age effects, PCA for dimensionality reduction and linear ν-support vector classification. Diagnostic performance of the classifier was determined by repeated nested 10-fold cross-validation.
We were able to correctly classify unseen test subjects’ diagnosis with 74% accuracy. Classification sensitivity and specificity was 74%. Volume reductions in SZ vs. BPD were predominantly located in the left peri- and intrasylvian regions, orbitofrontal regions, the nucleus caudatus and the right cerebellum. Volume reductions in BPD compared to SZ were found predominantly in the left cerebellum, in limbic areas and the left inferior occipital gyrus.
Our results suggest that SZ can be differentiated from BPD at the single-subject level using sMRI and pattern classification methods. In future, this method might enhance clinical evaluations and improve accuracy and reliability of differential diagnosis.
The clinical differentiation of schizophrenic and mood disorders is frequently challenged by co-occurring affective and psychotic symptoms. Thus, it has long been discussed whether these disease groups are subserved by common or distinct neurobiological surrogates.
The detection of diagnostic biomarkers for schizophrenic and mood disorders could facilitate clinical decision making in ambiguous cases.
To evaluated whether multivariate pattern classification of structural MRI enables the differential diagnostic classification of 158 patients with schizophrenia (SZ) and 104 patients with major depression (MD).
T1-weighted patient scans were processed using voxel-based morphometry. Diagnostic features were extracted from the age- and sex-adjusted GM maps using PCA and linear SVMs. Repeated nested cross-validation was emplyoed to assess the generalizability of diagnostic performance.
Cross-validated classification accuracy was 76% based on a discriminative pattern involving perisylvian, limbic, medial prefrontal and precuneal GM volume reductions in SZ vs. MD. GM volume reductions in MD vs. SZ were detected in the premotor, sensorimotor, parietal, cerebellar and brainstem structures. The 'SZ-likehood' of MD was correlated with the age of disease onset, leading to a significantly higher misclassification rate among MD patients with an age of onset between 15 and 30 yrs.
The findings suggest that SZ and MD can be identified at the single subject level using neuroanatomical pattern recognition. The decreased diagnostic separability of MD patients with an early disease onset may challenge the traditional nosological boundaries and may relate to higher levels of chronicity and unfavorable disease outcomes in this patient population.
Email your librarian or administrator to recommend adding this to your organisation's collection.