ABSTRACT
BackgroundDecision scores and ethically mindful algorithms are being established to adjudicate mechanical ventilation in the context of potential resources shortage due to the current onslaught of COVID-19 cases. There is a need for a reproducible and objective method to provide quantitative information for those scores. PurposeTowards this goal, we present a retrospective study testing the ability of a deep learning algorithm at extracting features from chest x-rays (CXR) to track and predict radiological evolution. Materials and MethodsWe trained a repurposed deep learning algorithm on the CheXnet open dataset (224,316 chest X-ray images of 65,240 unique patients) to extract features that mapped to radiological labels. We collected CXRs of COVID-19-positive patients from two open-source datasets (last accessed on April 9, 2020)(Italian Society for Medical and Interventional Radiology and MILA). Data collected form 60 pairs of sequential CXRs from 40 COVID patients (mean age {+/-} standard deviation: 56 {+/-} 13 years; 23 men, 10 women, seven not reported) and were categorized in three categories: "Worse", "Stable", or "Improved" on the basis of radiological evolution ascertained from images and reports. Receiver operating characteristic analyses, Mann-Whitney tests were performed. ResultsOn patients from the CheXnet dataset, the area under ROC curves ranged from 0.71 to 0.93 for seven imaging features and one diagnosis. Deep learning features between "Worse" and "Improved" outcome categories were significantly different for three radiological signs and one diagnostic ("Consolidation", "Lung Lesion", "Pleural effusion" and "Pneumonia"; all P < 0.05). Features from the first CXR of each pair could correctly predict the outcome category between "Worse" and "Improved" cases with 82.7% accuracy. ConclusionCXR deep learning features show promise for classifying the disease trajectory. Once validated in studies incorporating clinical data and with larger sample sizes, this information may be considered to inform triage decisions.
ABSTRACT
Do recency processes associated with repetitive sensorimotor events modulate the magnitude and functional coupling of brain rhythmicity in human temporal cortex? Intracranial stereo electroencephalographic activity (SEEG; 256 Hz sampling rate) was recorded from hippocampus, and inferior (BA20) and middle (BA21) temporal cortex in four epilepsy patients. The repetitive events were represented by predicted imperative somatosensory stimuli (CNV paradigm) triggering hand movements ("repetitive visuomotor") or counting ("repetitive counting"). The non-repetitive events were "rare" (P3 paradigm) somatosensory stimuli triggering hand movements ("non-repetitive visuomotor") or counting ("non-repetitive counting"). Brain rhythmicity was indexed by event-related desynchronization/synchronization (ERD/ERS) of SEEG data, whereas the functional coupling was evaluated by spectral SEEG coherence between pairs of the mentioned areas. The frequency bands of interest were theta (4-8 Hz), alpha (8-12 Hz), beta (14-30 Hz), and gamma (32-46 Hz). Compared to the non-repetitive events, the "repetitive visuomotor" events showed a significant beta and gamma ERS in the hippocampus and a significant theta ERD in the inferior temporal cortex. Furthermore, the "repetitive visuomotor" events induced a task-specific significant gamma coherence among the examined areas. These results suggest that recency processes do modulate the magnitude and functional coupling of brain rhythmicity (especially gamma) in the human temporal cortex.