Evaluation of artificial intelligence systems for assisting neurologists with fast and accurate annotations of scalp electroencephalography data.

BACKGROUND: Assistive automatic seizure detection can empower human annotators to shorten patient monitoring data review times. We present a proof-of-concept for a seizure detection system that is sensitive, automated, patient-specific, and tunable to maximise sensitivity while minimizing human annotation times. The system uses custom data preparation methods, deep learning analytics and electroencephalography (EEG) data. METHODS: Scalp EEG data of 365 patients containing 171,745 s ictal and 2,185,864 s interictal samples obtained from clinical monitoring systems were analysed as part of a crowdsourced artificial intelligence (AI) challenge. Participants were tasked to develop an ictal/interictal classifier with high sensitivity and low false alarm rates. We built a challenge platform that prevented participants from downloading or directly accessing the data while allowing crowdsourced model development. FINDINGS: The automatic detection system achieved tunable sensitivities between 75.00% and 91.60% allowing a reduction in the amount of raw EEG data to be reviewed by a human annotator by factors between 142x, and 22x respectively. The algorithm enables instantaneous reviewer-managed optimization of the balance between sensitivity and the amount of raw EEG data to be reviewed. INTERPRETATION: This study demonstrates the utility of deep learning for patient-specific seizure detection in EEG data. Furthermore, deep learning in combination with a human reviewer can provide the basis for an assistive data labelling system lowering the time of manual review while maintaining human expert annotation performance. FUNDING: IBM employed all IBM Research authors. Temple University employed all Temple University authors. The Icahn School of Medicine at Mount Sinai employed Eren Ahsen. The corresponding authors Stefan Harrer and Gustavo Stolovitzky declare that they had full access to all the data in the study and that they had final responsibility for the decision to submit for publication.

A comparative study of fatty acid profile and formation of biofilm in Geobacillus gargensis exposed to variable abiotic stress.

Understanding bacterial fatty acid (FA) profile has a great taxonomic significance as well as clinical importance for diagnosis issues. Both the composition and nature of membrane FAs change under different nutritional, biotic and (or) abiotic stresses, and environmental stress. Bacteria produce both odd-carbon as well as branched-chain fatty acids (BCFAs). This study was designed to examine the effect of abiotic pressure, including salinity, temperature, pH, and oxinic stress on the growth, development, and FA profile in thermophilic Geobacillus gargensis. Under these stresses, 3 parametric ratios, 2-methyl fatty acids/3-methyl fatty acids (iso-/anteiso-FAs), BCFAs/straight-chain saturated fatty acids (SCSFA), and SCSFAs/straight-chain unsaturated fatty acids (SCUFA), in addition to total lipids affected by variable stresses were measured. Our results indicate that the ratio of total iso-/anteiso-FAs increased at the acidic pH range of 4.1-5.2 and decreased with increasing pH. The reverse was true for salt stress when iso-/anteiso-FAs ratio increased with salt concentration. The BCFAs/SCSFAs and SCSFAs/SCUFAs ratios increased at neutral and alkaline pH and high salt concentration, reduced incubation time, and comparatively high temperature (55-65 °C) of the growth medium. The bacterial total lipid percentage deceased with increasing salt concentration, incubation period, but it increased with temperature. The formation of extracellular polymeric substances was observed under all stress conditions and with the addition of sodium dodecyl sulfate (2 and 5 mmol/L) to the growth medium. The membrane phospholipid composition of the bacterium was analyzed by thin-layer chromatography.