Ontology-based feature engineering in machine learning workflows for heterogeneous epilepsy patient records.

Biomedical ontologies are widely used to harmonize heterogeneous data and integrate large volumes of clinical data from multiple sources. This study analyzed the utility of ontologies beyond their traditional roles, that is, in addressing a challenging and currently underserved field of feature engineering in machine learning workflows. Machine learning workflows are being increasingly used to analyze medical records with heterogeneous phenotypic, genotypic, and related medical terms to improve patient care. We performed a retrospective study using neuropathology reports from the German Neuropathology Reference Center for Epilepsy Surgery at Erlangen, Germany. This cohort included 312 patients who underwent epilepsy surgery and were labeled with one or more diagnoses, including dual pathology, hippocampal sclerosis, malformation of cortical dysplasia, tumor, encephalitis, and gliosis. We modeled the diagnosis terms together with their microscopy, immunohistochemistry, anatomy, etiologies, and imaging findings using the description logic-based Web Ontology Language (OWL) in the Epilepsy and Seizure Ontology (EpSO). Three tree-based machine learning models were used to classify the neuropathology reports into one or more diagnosis classes with and without ontology-based feature engineering. We used five-fold cross validation to avoid overfitting with a fixed number of repetitions while leaving out one subset of data for testing, and we used recall, balanced accuracy, and hamming loss as performance metrics for the multi-label classification task. The epilepsy ontology-based feature engineering approach improved the performance of all the three learning models with an improvement of 35.7%, 54.5%, and 33.3% in logistics regression, random forest, and gradient tree boosting models respectively. The run time performance of all three models improved significantly with ontology-based feature engineering with gradient tree boosting model showing a 93.8% reduction in the time required for training and testing of the model. Although, all three models showed an overall improved performance across the three-performance metrics using ontology-based feature engineering, the rate of improvement was not consistent across all input features. To analyze this variation in performance, we computed feature importance scores and found that microscopy had the highest importance score across the three models, followed by imaging, immunohistochemistry, and anatomy in a decreasing order of importance scores. This study showed that ontologies have an important role in feature engineering to make heterogeneous clinical data accessible to machine learning models and also improve the performance of machine learning models in multilabel multiclass classification tasks.

Characterizing Brain Network Dynamics using Persistent Homology in Patients with Refractory Epilepsy.

Epilepsy is a common serious neurological disorder that affects more than 65 million persons worldwide and it is characterized by repeated seizures that lead to higher mortality and disabilities with corresponding negative impact on the quality of life of patients. Network science methods that represent brain regions as nodes and the interactions between brain regions as edges have been extensively used in characterizing network changes in neurological disorders. However, the limited ability of graph network models to represent high dimensional brain interactions are being increasingly realized in the computational neuroscience community. In particular, recent advances in algebraic topology research have led to the development of a large number of applications in brain network studies using topological structures. In this paper, we build on a fundamental construct of cliques, which are all-to-all connected nodes with a k-clique in a graph G (V, E), where V is set of nodes and E is set of edges, consisting of k-nodes to characterize the brain network dynamics in epilepsy patients using topological structures. Cliques represent brain regions that are coupled for similar functions or engage in information exchange; therefore, cliques are suitable structures to characterize the dynamics of brain dynamics in neurological disorders. We propose to detect and use clique structures during well-defined clinical events, such as epileptic seizures, to combine non-linear correlation measures in a matrix with identification of geometric structures underlying brain connectivity networks to identify discriminating features that can be used for clinical decision making in epilepsy neurological disorder.

Epilepsy-Connect: An Integrated Knowledgebase for Characterizing Alterations in Consciousness State of Pharmacoresistant Epilepsy Patients.

Alterations in consciousness state are a defining characteristic of focal epileptic seizures. Consequently, understanding the complex changes in neurocognitive networks which underpin seizure-induced alterations in consciousness state is important for advancement in seizure classification. Comprehension of these changes are complicated by a lack of data standardization; however, the use of a common terminological system or ontology in a patient registry minimizes this issue. In this paper, we introduce an integrated knowledgebase called Epilepsy-Connect to improve the understanding of changes in consciousness states during focal seizures of pharmacoresistant epilepsy patients. This registry catalogues over 809 seizures from 70 patients at University Hospital's Epilepsy Center who were undergoing stereotactic electroencephalography (SEEG) monitoring as part of an evaluation for surgical intervention. Although Epilepsy-Connect focuses on consciousness states, it aims to enable users to leverage data from an informatics platform to analyze epilepsy data in a streamlined manner. Epilepsy-Connect is available at https://bmhinformatics.case.edu/Epilepsyconnect/login/.