Automatic ARDS surveillance with chest X-ray recognition using convolutional neural networks.

OBJECTIVE: This study aims to design, validate and assess the accuracy a deep learning model capable of differentiation Chest X-Rays between pneumonia, acute respiratory distress syndrome (ARDS) and normal lungs. MATERIALS AND METHODS: A diagnostic performance study was conducted using Chest X-Ray images from adult patients admitted to a medical intensive care unit between January 2003 and November 2014. X-ray images from 15,899 patients were assigned one of three prespecified categories: "ARDS", "Pneumonia", or "Normal". RESULTS: A two-step convolutional neural network (CNN) pipeline was developed and tested to distinguish between the three patterns with sensitivity ranging from 91.8% to 97.8% and specificity ranging from 96.6% to 98.8%. The CNN model was validated with a sensitivity of 96.3% and specificity of 96.6% using a previous dataset of patients with Acute Lung Injury (ALI)/ARDS. DISCUSSION: The results suggest that a deep learning model based on chest x-ray pattern recognition can be a useful tool in distinguishing patients with ARDS from patients with normal lungs, providing faster results than digital surveillance tools based on text reports. CONCLUSION: A CNN-based deep learning model showed clinically significant performance, providing potential for faster ARDS identification. Future research should prospectively evaluate these tools in a clinical setting.

Adipocyte hypertrophy associates with in vivo postprandial fatty acid metabolism and adipose single-cell transcriptional dynamics.

Adipocyte hypertrophy is associated with metabolic complications independent of obesity. We aimed to determine: 1) the association between adipocyte size and postprandial fatty acid metabolism; 2) the potential mechanisms driving the obesity-independent, hypertrophy-associated dysmetabolism in vivo and at a single-cell resolution. Tracers with positron emission tomography were used to measure fatty acid metabolism in 40 men and women with normal or impaired glucose tolerance (NCT02808182), and single nuclei RNA-sequencing (snRNA-seq) to determine transcriptional dynamics of subcutaneous adipose tissue (AT) between individuals with AT hypertrophy vs. hyperplasia matched for sex, ethnicity, glucose-tolerance status, BMI, total and percent body fat, and waist circumference. Adipocyte size was associated with high postprandial total cardiac fatty acid uptake and higher visceral AT dietary fatty acid uptake, but lower lean tissue dietary fatty acid uptake. We found major shifts in cell transcriptomal dynamics with AT hypertrophy that were consistent with in vivo metabolic changes.

Effects of Image Quality on the Accuracy Human Pose Estimation and Detection of Eye Lid Opening/Closing Using Openpose and DLib.

OBJECTIVE: The application of computer models in continuous patient activity monitoring using video cameras is complicated by the capture of images of varying qualities due to poor lighting conditions and lower image resolutions. Insufficient literature has assessed the effects of image resolution, color depth, noise level, and low light on the inference of eye opening and closing and body landmarks from digital images. METHOD: This study systematically assessed the effects of varying image resolutions (from 100 × 100 pixels to 20 × 20 pixels at an interval of 10 pixels), lighting conditions (from 42 to 2 lux with an interval of 2 lux), color-depths (from 16.7 M colors to 8 M, 1 M, 512 K, 216 K, 64 K, 8 K, 1 K, 729, 512, 343, 216, 125, 64, 27, and 8 colors), and noise levels on the accuracy and model performance in eye dimension estimation and body keypoint localization using the Dlib library and OpenPose with images from the Closed Eyes in the Wild and the COCO datasets, as well as photographs of the face captured at different light intensities. RESULTS: The model accuracy and rate of model failure remained acceptable at an image resolution of 60 × 60 pixels, a color depth of 343 colors, a light intensity of 14 lux, and a Gaussian noise level of 4% (i.e., 4% of pixels replaced by Gaussian noise). CONCLUSIONS: The Dlib and OpenPose models failed to detect eye dimensions and body keypoints only at low image resolutions, lighting conditions, and color depths. CLINICAL IMPACT: Our established baseline threshold values will be useful for future work in the application of computer vision in continuous patient monitoring.

Total Postprandial Hepatic Nonesterified and Dietary Fatty Acid Uptake Is Increased and Insufficiently Curbed by Adipose Tissue Fatty Acid Trapping in Prediabetes With Overweight.

Excessive lean tissue uptake of fatty acids (FAs) is important in the development of insulin resistance and may be caused by impaired dietary FA (DFA) storage and/or increased nonesterified FA (NEFA) flux from adipose tissue intracellular lipolysis. Cardiac and hepatic total postprandial FA uptake of NEFA+DFA has, however, never been reported in prediabetes with overweight. In this study, 20 individuals with impaired glucose tolerance (IGT) and 19 participants with normal glucose tolerance (NGT) and normal fasting glucose underwent postprandial studies with whole-body positron emission tomography/computed tomography (PET/CT) with oral [18F]fluoro-thia-heptadecanoic acid and dynamic PET/CT with intravenous [11C]palmitate. Hepatic (97 [range 36-215] mmol/6 h vs. 68 [23-132] mmol/6 h, P = 0.03) but not cardiac (11 [range 4-24] mmol/6 h vs. 8 [3-20] mmol/6 h, P = 0.09) uptake of most sources of postprandial FA (NEFA + DFA uptake) integrated over 6 h was higher in IGT versus NGT. DFA accounted for lower fractions of total cardiac (21% [5-47] vs. 25% [9-39], P = 0.08) and hepatic (19% [6-52] vs. 28% [14-50], P = 0.04) uptake in IGT versus NGT. Increased adipose tissue DFA trapping predicted lower hepatic DFA uptake and was associated with higher total cardiac FA uptake. Hence, enhanced adipose tissue DFA trapping in the face of increased postprandial NEFA flux is insufficient to fully curb increased postprandial lean organ FA uptake in prediabetes with overweight (ClinicalTrials.gov; NCT02808182).

DeepImageTranslator: A free, user-friendly graphical interface for image translation using deep-learning and its applications in 3D CT image analysis.

The advent of deep-learning has set new standards in an array of image translation applications. At present, the use of these methods often requires computer programming experience. Non-commercial programs with graphical interface usually do not allow users to fully customize their deep-learning pipeline. Therefore, our primary objective is to provide a simple graphical interface that allows researchers with no programming experience to easily create, train, and evaluate custom deep-learning models for image translation. We also aimed to test the applicability of our tool in CT image semantic segmentation and noise reduction. DeepImageTranslator was implemented using the Tkinter library, the standard Python interface to the Tk graphical user interface toolkit; backend computations were implemented using data augmentation packages such as Pillow, Numpy, OpenCV, Augmentor, Tensorflow, and Keras libraries. Convolutional neural networks (CNNs) were trained using DeepImageTranslator. The effects of data augmentation, deep-supervision, and sample size on model accuracy were also systematically assessed. The DeepImageTranslator a simple tool that allows users to customize all aspects of their deep-learning pipeline, including the CNN, training optimizer, loss function, and the types of training image augmentation scheme. We showed that DeepImageTranslator can be used to achieve state-of-the-art accuracy and generalizability in semantic segmentation and noise reduction. Highly accurate 3D segmentation models for body composition can be obtained using training sample sizes as small as 17 images. In conclusion, an open-source deep-learning tool for accurate image translation with a user-friendly graphical interface was presented and evaluated. This standalone software can be downloaded at: https://sourceforge.net/projects/deepimagetranslator/.

Fat Cell Size: Measurement Methods, Pathophysiological Origins, and Relationships With Metabolic Dysregulations.

The obesity pandemic increasingly causes morbidity and mortality from type 2 diabetes, cardiovascular diseases and many other chronic diseases. Fat cell size (FCS) predicts numerous obesity-related complications such as lipid dysmetabolism, ectopic fat accumulation, insulin resistance, and cardiovascular disorders. Nevertheless, the scarcity of systematic literature reviews on this subject is compounded by the use of different methods by which FCS measurements are determined and reported. In this paper, we provide a systematic review of the current literature on the relationship between adipocyte hypertrophy and obesity-related glucose and lipid dysmetabolism, ectopic fat accumulation, and cardiovascular disorders. We also review the numerous mechanistic origins of adipocyte hypertrophy and its relationship with metabolic dysregulation, including changes in adipogenesis, cell senescence, collagen deposition, systemic inflammation, adipokine secretion, and energy balance. To quantify the effect of different FCS measurement methods, we performed statistical analyses across published data while controlling for body mass index, age, and sex.

Increased postprandial nonesterified fatty acid efflux from adipose tissue in prediabetes is offset by enhanced dietary fatty acid adipose trapping.

The mechanism of increased postprandial nonesterified fatty acid (NEFA) appearance in the circulation in impaired glucose tolerance (IGT) is due to increased adipose tissue lipolysis but could also be contributed to by reduced adipose tissue (AT) dietary fatty acid (DFA) trapping and increased "spillover" into the circulation. Thirty-one subjects with IGT (14 women, 17 men) and 29 with normal glucose tolerance (NGT, 15 women, 14 men) underwent a meal test with oral and intravenous palmitate tracers and the oral [18F]-fluoro-thia-heptadecanoic acid positron emission tomography method. Postprandial palmitate appearance (Rapalmitate) was higher in IGT versus NGT (P < 0.001), driven exclusively by Rapalmitate from obesity-associated increase in intracellular lipolysis (P = 0.01), as Rapalmitate from DFA spillover was not different between the groups (P = 0.19) and visceral AT DFA trapping was even higher in IGT versus NGT (P = 0.02). Plasma glycerol appearance was lower in IGT (P = 0.01), driven down by insulin resistance and increased insulin secretion. Thus, we found higher AT DFA trapping, limiting spillover to lean organs and in part offsetting the increase in Rapalmitate from intracellular lipolysis. Whether similar findings occur in frank diabetes, a condition also characterized by insulin resistance but relative insulin deficiency, requires further investigation (Clinicaltrials.gov: NCT04088344, NCT02808182).NEW & NOTEWORTHY We found higher adipose tissue dietary fatty acid trapping, limiting spillover to lean organs, that in part offsets the increase in appearance rate of palmitate from intracellular lipolysis in prediabetes. These results point to the adaptive nature of adipose tissue trapping and dietary fatty acid spillover as a protective mechanism against excess obesity-related palmitate appearance rate from intracellular adipose tissue lipolysis.