Sleep scoring is an important step for the detection of sleep disorders and usually performed by visual analysis. Since manual time consuming, machine-learning based approaches have been proposed. Though efficient, these algorithms are black-box in nature difficult to interpret clinicians. In this paper, we propose a deep learning architecture multi-modal scoring, investigate model's decision making process, compare reasoning with annotation guidelines AASM manual. Our architecture, called...

Abstract Aim To predict end‐stage renal disease (ESRD) in patients with type 2 diabetes by using machine‐learning models multiple baseline demographic and clinical characteristics. Materials methods In total, 11 789 nephropathy from three trials, RENAAL (n = 1513), IDNT 1715) ALTITUDE 8561), were used this study. Eighteen characteristics as predictors to train ESRD (doubling of serum creatinine and/or ESRD). We the area under receiver operator curve (AUC) assess prediction performance...

To investigate which metabolic pathways are targeted by the sodium-glucose co-transporter-2 inhibitor dapagliflozin to explore molecular processes involved in its renal protective effects.An unbiased mass spectrometry plasma metabolomics assay was performed on baseline and follow-up (week 12) samples from EFFECT II trial patients with type 2 diabetes non-alcoholic fatty liver disease receiving 10 mg/day (n = 19) or placebo 6). Transcriptomic signatures tubular compartments were identified...

Purpose Supply chains are facing several challenges due to disruptions and changing situations such as COVID-19 the need for increased levels of resilience is more important than ever. This paper focuses on exploring impact artificial intelligence (AI) supply chain (SCR) through a review existing literature. To address gap AI SCR, this study focused answering following two research questions: (1) What role technologies in SCR? (2) key ethical social implications that arise process enhancing...

Parkinson's disease (PD) is an advanced neurodegenerative condition distinguished by the rapid decline of dopamine neurons in midbrain, leading to imbalance and acetylcholine levels, precipitating associated symptoms. The main objective this work was fabricate solid lipid nanoparticles (SLNs) loaded with S-carboxymethyl-L-cystine (SC) for enhanced delivery brain. This study examines impact these SLNs on rotenone (RT) caused both rat zebrafish models. process loading SC into achieved through...

To develop a personalizable algorithm to discriminate between sedation levels in ICU patients based on heart rate variability.Multicenter, pilot study.Several ICUs at Massachusetts General Hospital, Boston, MA.We gathered 21,912 hours of routine electrocardiogram recordings from heterogenous group 70 adult patients. All included the study were mechanically ventilated and receiving sedatives.As "ground truth" for developing our method, we used Richmond Agitation Sedation Scale scores grouped...

Atomic decomposition (AD) can be used to efficiently decompose an arbitrary signal. In this paper, we present a method detect neonatal electroencephalogram (EEG) seizure based on AD via orthogonal matching pursuit using novel, application-specific, dictionary. The dictionary consists of pseudoperiodic Duffing oscillator atoms which are designed coherent with the epochs. relative structural complexity (a measure rate convergence AD) is as sole feature for detection. proposed was tested large...

Over- and under-sedation are common in the ICU, contribute to poor ICU outcomes including delirium. Behavioral assessments, such as Richmond Agitation-Sedation Scale (RASS) for monitoring levels of sedation Confusion Assessment Method (CAM-ICU) detecting signs delirium, often used. As an alternative, brain with electroencephalography (EEG) has been proposed operating room, but is challenging implement due differences between critical illness elective surgery, well duration sedation. Here we...

Abstract Study Objectives Dexmedetomidine-induced electroencephalogram (EEG) patterns during deep sedation are comparable with natural sleep patterns. Using large-scale EEG recordings and machine learning techniques, we investigated whether dexmedetomidine-induced indeed mimics Methods We used from three sources in this study: 8,707 overnight 30 dexmedetomidine clinical trial EEG. levels were assessed using the Modified Observer’s Assessment of Alertness/Sedation (MOAA/S) score. extracted 22...

Objective: This study was performed to evaluate how well states of deep sedation in 'CU patients can be detected from the frontal electroencephalogram (EEG) using features based on method atomic decomposition (AD). Methods: We analyzed a clinical dataset 20 min EEG recordings per patient 44 mechanically ventilated adult receiving sedatives an intensive care unit ('CU) setting. Several derived AD signal were used discriminate between awake and sedated states. trained support vector machine...

To explore the potential value of heart rate variability features for automated monitoring sedation levels in mechanically ventilated ICU patients.Multicenter, pilot study.Several ICUs at Massachusetts General Hospital, Boston, MA.Electrocardiogram recordings from 40 adult patients receiving sedatives an setting were used to develop and test proposed system.Richmond Agitation-Sedation Scale scores acquired prospectively assess patient as ground truth. Richmond grouped into four levels,...

To assess the potential of supervised machine-learning techniques to identify clinical variables for predicting short-term and long-term glycated haemoglobin (HbA1c) response after insulin treatment initiation in patients with type 2 diabetes mellitus (T2DM).We included T2DM from Groningen Initiative Analyse Type Treatment (GIANTT) database who started between 2007 2013 had a minimum follow-up years. Short- responses at 6 (±2) 24 months initiation, respectively, were assessed. Patients...

Brain monitors tracking quantitative brain activities from electroencephalogram (EEG) to predict hypnotic levels have been proposed as a labor-saving alternative behavioral assessments. Expensive clinical trials are required validate any newly developed processed EEG monitor for every drug and combinations of drugs due drug-specific patterns. There is need an alternative, efficient, economical method.Using deep learning algorithms, we novel data-repurposing framework sleep rhythms. We used...

In primary care, identifying patients with type 2 diabetes (T2D) who are at increased risk of hypoglycaemia is important for the prevention hypoglycaemic events. We aimed to develop a screening tool based on machine learning identify such using routinely available demographic and medication data.

Abstract The objective of this study is to evaluate the feasibility a disease-specific deep learning (DL) model based on minimum intensity projection (minIP) for automated emphysema detection in low-dose computed tomography (LDCT) scans. LDCT scans 240 individuals from population-based cohort Netherlands (ImaLife study, mean age ± SD = 57 6 years) were retrospectively chosen training and internal validation DL model. For independent testing, 125 lung cancer screening USA (NLST 64 5 used....

With aging, patients with diabetic kidney disease (DKD) show progressive decrease in function. We investigated whether the deviation of biological age (BA) from chronological (CA) due to DKD can be used (denoted as Kidney Age Index; KAI) quantify function using machine learning algorithms.Three large datasets were this study develop KAI. The algorithms trained on PREVEND dataset healthy subjects (N = 7963) 13 clinical markers predict CA. model was then BA RENAAL 1451) and IDNT 1706)....

Over and under-sedation are common in critically ill patients admitted to the Intensive Care Unit. Clinical assessments provide limited time resolution based on behavior rather than brain itself. Existing monitors have been developed primarily for non-ICU settings. Here, we use a clinical dataset from 154 ICU whom Richmond Agitation-Sedation Score is assessed about every 2 hours. We develop recurrent neural network (RNN) model discriminate between deep vs. no sedation, trained end-to-end raw...

We developed a simple and fully automated method for detecting artifacts in the R-R interval (RRI) time series of ECG that is tailored to intensive care unit (ICU) setting. From recordings 50 adult ICU-subjects we selected 60 epochs with valid R-peak detections containing leading missed or false positive detections. Next, calculated absolute value difference between two adjacent RRIs (adRRI), obtained empirical probability distributions adRRI values R-peaks artifacts. these, an optimal...

In this paper we examined the robustness of a feature-set based on time-frequency distributions (TFDs) for neonatal EEG seizure detection. This was originally proposed in literature detection using support vector machine (SVM). We tested performance with smoothed Wigner-Ville distribution and modified B as underlying TFDs. The system signal image processing features from TFD able to achieve median receiver operator characteristic area 0.96 (IQR 0.91-0.98) large clinical dataset 826 h data 18...

The development of automated methods electroencephalogram (EEG) seizure detection is an important problem in neonatology. This paper proposes improvements to a previously described method based on atomic decomposition by developing new time-frequency (TF) dictionary that highly coherent with the newborn EEG seizure. We compare performance proposed neonatal signals achieved using Gabor, Fourier and wavelet dictionaries. Through analysis real data, we show first, selection can influence...