Identifying signal-dependent information about the preictal state: A comparison across ECoG, EEG and EKG using deep learning.

BACKGROUND: The inability to reliably assess seizure risk is a major burden for epilepsy patients and prevents developing better treatments. Recent advances have paved the way for increasingly accurate seizure preictal state detection algorithms, primarily using electrocorticography (ECoG). To develop seizure forecasting for broad clinical and ambulatory use, however, less complex and invasive modalities are needed. Algorithms using scalp electroencephalography (EEG) and electrocardiography (EKG) have also achieved better than chance performance. But it remains unknown how much preictal information is in ECoG versus modalities amenable to everyday use - such as EKG and single channel EEG - and how to optimally extract that preictal information for seizure prediction. METHODS: We apply deep learning - a powerful method to extract information from complex data - on a large epilepsy data set containing multi-day, simultaneous recordings of EKG, ECoG, and EEG, using a variety of feature sets. We use the relative performance of our algorithms to compare the preictal information contained in each modality. RESULTS: We find that single-channel EKG contains a comparable amount of preictal information as scalp EEG with up to 21 channels and that preictal information is best extracted not with standard heart rate measures, but from the power spectral density. We report that preictal information is not preferentially contained in EEG or ECoG channels within the seizure onset zone. CONCLUSION: Collectively, these insights may help to devise future prospective, minimally invasive long-term epilepsy monitoring trials with single-channel EKG as a particularly promising modality.

STEM/Non-STEM Divide Structures Undergraduate Beliefs About Gender and Talent in Academia.

Research and popular debate on female underrepresentation in academia has focused on STEM fields. But recent work has offered a unifying explanation for gender representation across the STEM/non-STEM divide. This proposed explanation, called the field-specific ability beliefs (FAB) hypothesis, postulates that, in combination with pervasive stereotypes that link men but not women with intellectual talent, academics perpetuate female underrepresentation by transmitting to students in earlier stages of education their beliefs about how much intellectual talent is required for success in each academic field. This theory was supported by a nationwide survey of U.S. academics that showed both STEM and non-STEM fields with fewer women are also the fields that academics believe require more brilliance. We test this top-down schema with a nationwide survey of U.S. undergraduates, assessing the extent to which undergraduate beliefs about talent in academia mirror those of academics. We find no evidence that academics transmit their beliefs to undergraduates. We also use a second survey "identical to the first but with each field's gender ratio provided as added information" to explicitly test the relationship between undergraduate beliefs about gender and talent in academia. The results for this second survey suggest that the extent to which undergraduates rate brilliance as essential to success in an academic field is highly sensitive to this added information for non-STEM fields, but not STEM fields. Overall, our study offers evidence that, contrary to FAB hypothesis, the STEM/non-STEM divide principally shapes undergraduate beliefs about both gender and talent in academia.