A machine learning approach using EEG data to predict response to SSRI treatment for major depressive disorder.

OBJECTIVE: The problem of identifying, in advance, the most effective treatment agent for various psychiatric conditions remains an elusive goal. To address this challenge, we investigate the performance of the proposed machine learning (ML) methodology (based on the pre-treatment electroencephalogram (EEG)) for prediction of response to treatment with a selective serotonin reuptake inhibitor (SSRI) medication in subjects suffering from major depressive disorder (MDD). METHODS: A relatively small number of most discriminating features are selected from a large group of candidate features extracted from the subject's pre-treatment EEG, using a machine learning procedure for feature selection. The selected features are fed into a classifier, which was realized as a mixture of factor analysis (MFA) model, whose output is the predicted response in the form of a likelihood value. This likelihood indicates the extent to which the subject belongs to the responder vs. non-responder classes. The overall method was evaluated using a "leave-n-out" randomized permutation cross-validation procedure. RESULTS: A list of discriminating EEG biomarkers (features) was found. The specificity of the proposed method is 80.9% while sensitivity is 94.9%, for an overall prediction accuracy of 87.9%. There is a 98.76% confidence that the estimated prediction rate is within the interval [75%, 100%]. CONCLUSIONS: These results indicate that the proposed ML method holds considerable promise in predicting the efficacy of SSRI antidepressant therapy for MDD, based on a simple and cost-effective pre-treatment EEG. SIGNIFICANCE: The proposed approach offers the potential to improve the treatment of major depression and to reduce health care costs.

A pilot study to determine whether machine learning methodologies using pre-treatment electroencephalography can predict the symptomatic response to clozapine therapy.

OBJECTIVE: To investigate whether applying advanced machine learning (ML) methodologies to pre-treatment electroencephalography (EEG) data can predict the response to clozapine therapy in adult subjects suffering from chronic schizophrenia. METHODS: Pre-treatment EEG data are collected in 23+14 schizophrenic adults. Treatment outcome, after at least one year follow-up, is determined using clinical ratings by a trained clinician blind to EEG results. First, a feature selection scheme is employed to select a reduced subset of features extracted from the subjects' EEG that is most statistically relevant to our treatment-response prediction. These features are then entered into a classifier, which is realized in the form of a kernel partial least squares regression method that performs response prediction. Various scales, including the positive and negative syndrome scale (PANSS) are used as treatment-response indicators. RESULTS: We determined that a set of discriminating EEG features do exist. A low-dimensional representation of the feature space showed significant clustering into clozapine responder and non-responder groups. The minimum level of performance of the proposed prediction methodology, tested over a range of conditions using the leave-one-out cross-validation method using the original 23 subjects, with further testing in an independent sample of 14 subjects, was 85%. CONCLUSIONS: These findings indicate that analysis of pre-treatment EEG data can predict the clinical response to clozapine in treatment resistant schizophrenia. SIGNIFICANCE: If replicated in a larger population, this novel approach to EEG analysis may assist the clinician in determining treatment-efficacy.

Antipsychotic drug use is correlated with CRP40/mortalin mRNA expression in the dorsolateral prefrontal cortex of human postmortem brain specimens.

Heat shock proteins act as intracellular chaperones by assisting with proper protein folding in response to various cellular stresses. In doing so, these proteins protect the cell from unwanted protein aggregation, which in turn, plays an important role in the pathogenesis of numerous disorders. Previous reports from our laboratory have described a 40 kDa catecholamine regulated heat shock-like protein (CRP40), an alternate gene product of the 70 kDa mitochondrial heat shock protein, mortalin. CRP40 shares an intimate association with dopaminergic activity, specifically as it pertains to dopamine dysregulation in schizophrenia. This study investigates human CRP40/mortalin mRNA expression within dorsolateral prefrontal cortex postmortem specimens from normal control, schizophrenic and bipolar patients obtained from the Stanley Medical Research Institute. Real-time polymerase chain reaction was carried out for all patient samples (n=105; n=35 per group) in a blinded manner. No significant alterations in CRP40/mortalin mRNA expression levels were observed between control, bipolar and schizophrenic patients. However, multiple regression demonstrated a distinct positive correlation between CRP40/mortalin mRNA expression and lifetime use of antipsychotic drugs within the schizophrenic patient profile, after controlling for important confounding factors. Thus, the data suggest that human CRP40/mortalin is modulated by dopaminergic activity and may act to protect neurons from excess catecholamine activity in regions of the brain associated with psychosis.

Going beyond: an adventure- and recreation-based group intervention promotes well-being and weight loss in schizophrenia.

OBJECTIVE: To undertake a preliminary study to assess the feasibility of clinical implementation and evaluate the effectiveness of a novel adventure- and recreation-based group intervention in the rehabilitation of individuals with schizophrenia. METHODS: In a 2-year, prospective, case-control study, 23 consecutively referred, clinically stabilized schizophrenia patients received the new intervention over an 8-month period; 31 patients on the wait list, considered the control group, received standard clinical care that included some recreational activities. Symptom severity, self-esteem, self-appraised cognitive abilities, and functioning were documented for both groups with standardized rating scales administered at baseline, on completion of treatment, and at 12 months posttreatment. RESULTS: Treatment adherence was 97%, and there were no dropouts. Patients in the study group showed marginal improvement in perceived cognitive abilities and on domain-specific functioning measures but experienced a significant improvement in their self-esteem and global functioning (P < 0.05), as well as a weight loss of over 12 lb. Improvement was sustained over 1 year with further occupational and social gains. CONCLUSION: In the context of overcoming barriers to providing early intervention for youth and preventing metabolic problems among older adults with schizophrenia, adventure- and recreation-based interventions could play a useful complementary role.

Artefacting reliability in QEEG topographic maps.

OBJECTIVE: To quantify the extent of disagreement among expert artefactors, to compare their results with a 'minimalist' approach where only gross artefacts were removed, and to relate the result to frequency and to cranial location. METHODS: Raw QEEG records for 12 subjects were artefacted by 6-expert, and one 'minimalist', artefactor. Standard errors (SEs) of measurement were computed for each of 20 1.2 Hz frequency bins in each of 20 electrode positions. RESULTS: SEs declined with frequency. SEs associated with the 'minimalist' were comparable to those of the experts. The high SEs in delta were confined to the frontal and frontotemporal regions. SEs were small and uniform over the cranium for frequencies greater than 5.2 Hz. CONCLUSIONS: Artefactor unreliability is a serious problem in the delta band because of disagreement on eye movement artefacts. The success of the 'minimalist' suggests that automated methodologies may be a feasible alternative to the use of expert technicians. SIGNIFICANCE: A novel statistical procedure proves helpful in elucidating the sources of artefactor error and points to possible remedies.