Dataset Information

Applying machine learning in motor activity time series of depressed bipolar and unipolar patients compared to healthy controls.

ABSTRACT: Current practice of assessing mood episodes in affective disorders largely depends on subjective observations combined with semi-structured clinical rating scales. Motor activity is an objective observation of the inner physiological state expressed in behavior patterns. Alterations of motor activity are essential features of bipolar and unipolar depression. The aim was to investigate if objective measures of motor activity can aid existing diagnostic practice, by applying machine-learning techniques to analyze activity patterns in depressed patients and healthy controls. Random Forrest, Deep Neural Network and Convolutional Neural Network algorithms were used to analyze 14 days of actigraph recorded motor activity from 23 depressed patients and 32 healthy controls. Statistical features analyzed in the dataset were mean activity, standard deviation of mean activity and proportion of zero activity. Various techniques to handle data imbalance were applied, and to ensure generalizability and avoid overfitting a Leave-One-User-Out validation strategy was utilized. All outcomes reports as measures of accuracy for binary tests. A Deep Neural Network combined with SMOTE class balancing technique performed a cut above the rest with a true positive rate of 0.82 (sensitivity) and a true negative rate of 0.84 (specificity). Accuracy was 0.84 and the Matthews Correlation Coefficient 0.65. Misclassifications appear related to data overlapping among the classes, so an appropriate future approach will be to compare mood states intra-individualistically. In summary, machine-learning techniques present promising abilities in discriminating between depressed patients and healthy controls in motor activity time series.

SUBMITTER: Jakobsen P

PROVIDER: S-EPMC7446864 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Applying machine learning in motor activity time series of depressed bipolar and unipolar patients compared to healthy controls.

Jakobsen Petter P Garcia-Ceja Enrique E Riegler Michael M Stabell Lena Antonsen LA Nordgreen Tine T Torresen Jim J Fasmer Ole Bernt OB Oedegaard Ketil Joachim KJ

PloS one 20200824 8

Current practice of assessing mood episodes in affective disorders largely depends on subjective observations combined with semi-structured clinical rating scales. Motor activity is an objective observation of the inner physiological state expressed in behavior patterns. Alterations of motor activity are essential features of bipolar and unipolar depression. The aim was to investigate if objective measures of motor activity can aid existing diagnostic practice, by applying machine-learning techn ...[more]

PMID: 32833958

Similar Datasets

Project description:ObjectiveThe clinical response of patients with bipolar disorder to medical treatment is variable. A better understanding of the underlying neural circuitry involved in bipolar treatment responsivity subtypes may provide insight into treatment resistance and aid in identifying an effective surgical target for deep brain stimulation (DBS) specific to the disorder. Despite considerable imaging research related to the disease, a paucity of comparative imaging analyses of treatment responsiveness exists. There are also no DBS targets designed expressly for patients with bipolar disorder. Therefore, the authors analyzed cingulum bundle axonal connectivity in relation to cortico-striatal-thalamo-cortical (CSTC) loops implicated in bipolar disorder across subjects who are responsive to treatment (RSP) and those who are refractory to therapy (REF), compared to healthy controls (HCs).MethodsTwenty-five subjects with bipolar disorder (13 RSP and 12 REF), diagnosed using the Mini International Neuropsychiatric Interview and classified with standardized rating scales, and 14 HCs underwent MRI with diffusion sequences for probabilistic diffusion-weighted tractography analysis. Image processing and tractography were performed using MRTrix. Region of interest (ROI) masks were created manually for 10 anterior cingulum bundle subregions, including surgical targets previously evaluated for the treatment of bipolar disorder (cingulotomy and subgenual cingulate DBS targets). Cortical and subcortical ROIs of brain areas thought to be associated with bipolar disorder and described in animal tract-tracing models were created via FreeSurfer. The number of axonal projections from the cingulum bundle subregion ROIs to cortical/subcortical ROIs for each group was compared.ResultsSignificant differences were found across groups involving cingulum bundle and CSTC loops. Subjects in the RSP group had increased connections from rostral cingulum bundle to medial orbitofrontal cortex, which is part of the limbic CSTC loop, whereas subjects in the REF group had increased connectivity from rostral cingulum bundle to thalamus. Additionally, compared to HCs, both RSP and REF subjects had decreased cingulum bundle dorsal connectivity (dorsal anterior/posterior cingulate, dorsomedial/lateral frontal cortex) and increased cingulum bundle ventral connectivity (subgenual cingulate, frontal pole, lateral orbitofrontal cortex) involving limbic and associative CSTC loops.ConclusionsFindings demonstrate that bipolar treatment responsivity may be associated with significant differences in cingulum bundle connectivity in relation to CSTC loops, which may help identify a surgical target for bipolar disorder treatment via DBS in the future.

Dataset Information

Applying machine learning in motor activity time series of depressed bipolar and unipolar patients compared to healthy controls.

Publications

Applying machine learning in motor activity time series of depressed bipolar and unipolar patients compared to healthy controls.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets