ABSTRACT
BACKGROUND: Chest radiographs are routinely performed in intensive care unit (ICU) to confirm the correct position of an endotracheal tube (ETT) relative to the carina. However, their interpretation is often challenging and requires substantial time and expertise. The aim of this study was to propose an externally validated deep learning model with uncertainty quantification and image segmentation for the automated assessment of ETT placement on ICU chest radiographs. METHODS: The CarinaNet model was constructed by applying transfer learning to the RetinaNet model using an internal dataset of ICU chest radiographs. The accuracy of the model in predicting the position of the ETT tip and carina was externally validated using a dataset of 200 images extracted from the MIMIC-CXR database. Uncertainty quantification was performed using the level of confidence in the ETT-carina distance prediction. Segmentation of the ETT was carried out using edge detection and pixel clustering. RESULTS: The interrater agreement was 0.18 cm for the ETT tip position, 0.58 cm for the carina position, and 0.60 cm for the ETT-carina distance. The mean absolute error of the model on the external test set was 0.51 cm for the ETT tip position prediction, 0.61 cm for the carina position prediction, and 0.89 cm for the ETT-carina distance prediction. The assessment of ETT placement was improved by complementing the human interpretation of chest radiographs with the CarinaNet model. CONCLUSIONS: The CarinaNet model is an efficient and generalizable deep learning algorithm for the automated assessment of ETT placement on ICU chest radiographs. Uncertainty quantification can bring the attention of intensivists to chest radiographs that require an experienced human interpretation. Image segmentation provides intensivists with chest radiographs that are quickly interpretable and allows them to immediately assess the validity of model predictions. The CarinaNet model is ready to be evaluated in clinical studies.
Subject(s)
Deep Learning , Humans , Trachea , Intubation, Intratracheal/methods , Radiography , Intensive Care UnitsABSTRACT
OBJECTIVE: Current guidelines consider obesity to be a relative contraindication to venoarterial extracorporeal membrane oxygenation (VA-ECMO) for refractory cardiogenic shock. The authors investigated the effect of body mass index (BMI) on clinical outcomes in patients treated with VA-ECMO for cardiogenic shock. DESIGN: This was a retrospective and observational study. SETTING: University hospital. PARTICIPANTS: The study comprised 150 adult patients who underwent VA-ECMO for cardiogenic shock. MEASUREMENTS AND MAIN RESULTS: The primary outcome was intensive care unit (ICU) mortality. Of the 150 included patients, 10 were underweight (BMI < 18.5 kg/m²), 62 were normal weight (BMI = 18.5-24.9 kg/m²), 34 were overweight (BMI = 25.0-29.9 kg/m²), 34 were obese class I (BMI = 30.0-34.9 kg/m²), and 10 were obese class II (BMI = 35.0-39.9 kg/m²). All-cause ICU mortality was 62% (underweight, 70%; normal weight, 53%; overweight, 65%; class I obese, 71%; class II obese, 70%). After multivariate logistic regression, BMI was not associated with ICU mortality (adjusted odds ratio [aOR] 0.99 [0.92-1.07], p = 0.8). Analysis by BMI category showed unfavorable mortality trends in underweight patients (aOR 3.58 [0.82-19.6], p = 0.11) and class I obese patients (aOR 2.39 [0.95-6.38], p = 0.07). No statistically significant differences were found among BMI categories in the incidences of complications. CONCLUSION: The results suggested that BMI alone should not be considered an exclusion criterion for VA-ECMO. The unfavorable trend observed in underweight patients could be the result of malnutrition.
Subject(s)
Extracorporeal Membrane Oxygenation , Shock, Cardiogenic , Adult , Body Mass Index , Extracorporeal Membrane Oxygenation/methods , Humans , Obesity/complications , Obesity/epidemiology , Overweight/complications , Retrospective Studies , Shock, Cardiogenic/etiology , Thinness/complications , Thinness/epidemiologyABSTRACT
Morel-Lavallée lesions are posttraumatic hemolymphatic collections related to shearing injury of variable mechanism (severe trauma or low-velocity crush injury), resulting in disruption of interfacial planes between subcutaneous soft tissue and muscle and perforating lymphatics and blood vessels. A 69-y-old woman, without medical history, was admitted to the emergency ward for important pain located from her behind to the anterior face of the thighs. Physical examination revealed large semirecent hematoma with a large soft fluctuant area with hypermobility of the skin and cutaneous hyperesthesia, spreading from her behind to the anterior face of her thigs. Her symptoms were typical of a Morel-Lavallée lesion. The ultrasound examination revealed hypoechoic heterogeneous compressible fluid containing fat debris with irregular margins and lobular shape, localized between subcutaneous fat and deep fascia, without flow on color Doppler, confirming the acute Morel-Lavallée lesion. After evacuation by needle puncture, the collection reappearance was probably due to hemolymphatic fluid corresponding to type 1 of the Mellado and Bencardino classification of Morel-Lavallée lesions. Patients and Physicians should be aware of the worsening effects of balneotherapy on skin hematoma to prevent dramatic extension of Morel-Lavallée lesions.
Subject(s)
Hematoma/surgery , Soft Tissue Injuries/surgery , Thigh/diagnostic imaging , Aged , Female , Hematoma/diagnostic imaging , Humans , Soft Tissue Injuries/diagnostic imaging , Ultrasonography, Doppler, Color/methodsABSTRACT
Deep learning algorithms have shown excellent performances in the field of medical image recognition, and practical applications have been made in several medical domains. Little is known about the feasibility and impact of an undetectable adversarial attacks, which can disrupt an algorithm by modifying a single pixel of the image to be interpreted. The aim of the study was to test the feasibility and impact of an adversarial attack on the accuracy of a deep learning-based dermatoscopic image recognition system.First, the pre-trained convolutional neural network DenseNet-201 was trained to classify images from the training set into 7 categories. Second, an adversarial neural network was trained to generate undetectable perturbations on images from the test set, to classifying all perturbed images as melanocytic nevi. The perturbed images were classified using the model generated in the first step. This study used the HAM-10000 dataset, an open source image database containing 10,015 dermatoscopic images, which was split into a training set and a test set. The accuracy of the generated classification model was evaluated using images from the test set. The accuracy of the model with and without perturbed images was compared. The ability of 2 observers to detect image perturbations was evaluated, and the inter observer agreement was calculated.The overall accuracy of the classification model dropped from 84% (confidence interval (CI) 95%: 82-86) for unperturbed images to 67% (CI 95%: 65-69) for perturbed images (Mc Nemar test, Pâ<â.0001). The fooling ratio reached 100% for all categories of skin lesions. Sensitivity and specificity of the combined observers calculated on a random sample of 50 images were 58.3% (CI 95%: 45.9-70.8) and 42.5% (CI 95%: 27.2-57.8), respectively. The kappa agreement coefficient between the 2 observers was negative at -0.22 (CI 95%: -0.49--0.04).Adversarial attacks on medical image databases can distort interpretation by image recognition algorithms, are easy to make and undetectable by humans. It seems essential to improve our understanding of deep learning-based image recognition systems and to upgrade their security before putting them to practical and daily use.
Subject(s)
Deep Learning , Diagnostic Errors , Image Interpretation, Computer-Assisted , Skin Diseases/diagnosis , Algorithms , Datasets as Topic , Humans , Models, Statistical , Neural Networks, Computer , Reproducibility of Results , Risk Factors , Sensitivity and Specificity , Skin/pathology , Skin Diseases/pathologyABSTRACT
Melioidosis is an emerging disease caused by the environmental Gram-negative bacillus Burkholderia pseudomallei. Melioidosis has been reported to be endemic mainly in northern Australia and Southeast Asia. Here, we report the first two cases of melioidosis on the Comorian island of Mayotte. We also describe four cases that occurred over a short period of time in patients who had traveled between Mayotte, Madagascar, and Reunion Island.
Subject(s)
Melioidosis/diagnosis , Melioidosis/epidemiology , Aged , Anti-Bacterial Agents/therapeutic use , Comoros/epidemiology , Humans , Madagascar/epidemiology , Male , Melioidosis/drug therapy , Middle Aged , Reunion/epidemiologyABSTRACT
Dengue disease is found worldwide and has been spreading exponentially in the past decades. Severe forms affect about 2% of patients, sometimes leading to organ failure. Recent studies have shown that fulminant dengue myocarditis is more common than previously thought, and fatal cases have been reported. The treatment of fulminant dengue myocarditis remains supportive and is very challenging in cases of cardiogenic shock. Here, we describe four cases presenting with fulminant dengue myocarditis treated with extracorporeal membrane oxygenation, among two survived.
Subject(s)
Extracorporeal Membrane Oxygenation/adverse effects , Myocarditis/therapy , Myocarditis/virology , Severe Dengue/complications , Shock, Cardiogenic/etiology , Adult , Female , Humans , Male , Middle Aged , Myocarditis/diagnosis , Myocarditis/pathologyABSTRACT
PURPOSE: Thrombocytopenia is a frequent and serious adverse event in patients treated with veno-arterial extracorporeal membrane oxygenation (VA-ECMO) for refractory cardiogenic shock. Similarly to postcardiac surgery patients, heparin-induced thrombocytopenia (HIT) could represent the causative underlying mechanism. However, the epidemiology as well as related mortality regarding HIT and VA-ECMO remains largely unknown. We aimed to define the prevalence and associated 90-day mortality of HIT diagnosed under VA-ECMO. METHODS: This retrospective study included patients under VA-ECMO from 20 French centers between 2012 and 2016. Selected patients were hospitalized for more than 3 days with high clinical suspicion of HIT and positive anti-PF4/heparin antibodies. Patients were classified according to results of functional tests as having either Confirmed or Excluded HIT. RESULTS: A total of 5797 patients under VA-ECMO were screened; 39/5797 met the inclusion criteria, with HIT confirmed in 21/5797 patients (0.36% [95% CI] [0.21-0.52]). Fourteen of 39 patients (35.9% [20.8-50.9]) with suspected HIT were ultimately excluded because of negative functional assays. Drug-induced thrombocytopenia tended to be more frequent in Excluded HIT at the time of HIT suspicion (p = 0.073). The platelet course was similar between Confirmed and Excluded HIT (p = 0.65). Mortality rate was 33.3% [13.2-53.5] in Confirmed and 50% [23.8-76.2] in Excluded HIT (p = 0.48). CONCLUSIONS: Prevalence of HIT among patients under VA-ECMO is extremely low at 0.36% with an associated mortality rate of 33.3%, which appears to be in the same range as that observed in patients treated with VA-ECMO without HIT. In addition, HIT was ultimately ruled out in one-third of patients with clinical suspicion of HIT and positive anti-PF4/heparin antibodies.