Evaluation of an improved tool for non-invasive prediction of neonatal respiratory morbidity based on fully automated fetal lung ultrasound analysis

Burgos-Artizzu, Xavier P.; Perez‐Moreno, Alvaro; Coronado-Gutiérrez, David; Gratacós, Eduard; Palacio, Montse

doi:10.1038/s41598-019-38576-w

Cited by 42 publications

(35 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The most commonly used biomarkers in clinical practice for the screening of fetal abnormalities are fetal biometries, estimates of fetal weight, and/or Doppler blood flow 2 . For example, nuchal translucency measurement is the basis for the first trimester screening of fetal aneuploidies 3 , estimated fetal weight is used to detect abnormal growth 4 , fetal lungs can be used to predict neonatal respiratory morbidity 5 and uterine cervix can be used to determine the risk of a preterm delivery 6,7 .…”

mentioning

confidence: 99%

Evaluation of deep convolutional neural networks for automatic classification of common maternal fetal ultrasound planes

Burgos-Artizzu

Coronado-Gutiérrez

Valenzuela-Alcaráz

et al. 2020

Sci Rep

Self Cite

107

View full text Add to dashboard Cite

The goal of this study was to evaluate the maturity of current Deep Learning classification techniques for their application in a real maternal-fetal clinical environment. A large dataset of routinely acquired maternal-fetal screening ultrasound images (which will be made publicly available) was collected from two different hospitals by several operators and ultrasound machines. All images were manually labeled by an expert maternal fetal clinician. Images were divided into 6 classes: four of the most widely used fetal anatomical planes (Abdomen, Brain, Femur and Thorax), the mother's cervix (widely used for prematurity screening) and a general category to include any other less common image plane. fetal brain images were further categorized into the 3 most common fetal brain planes (Trans-thalamic, Trans-cerebellum, Trans-ventricular) to judge fine grain categorization performance. The final dataset is comprised of over 12,400 images from 1,792 patients, making it the largest ultrasound dataset to date. We then evaluated a wide variety of state-of-the-art deep convolutional neural networks on this dataset and analyzed results in depth, comparing the computational models to research technicians, which are the ones currently performing the task daily. Results indicate for the first time that computational models have similar performance compared to humans when classifying common planes in human fetal examination. However, the dataset leaves the door open on future research to further improve results, especially on fine-grained plane categorization.

show abstract

mentioning

confidence: 99%

Evaluation of deep convolutional neural networks for automatic classification of common maternal fetal ultrasound planes

Burgos-Artizzu

Coronado-Gutiérrez

Valenzuela-Alcaráz

et al. 2020

Sci Rep

Self Cite

107

View full text Add to dashboard Cite

show abstract

“…The thorax occupied two-thirds of the screen, and there were no significant shadows in fetal lungs deriving from ribs or spine. The image was saved as DICOM file and sent to the commercial Web site of QuantusFLM® (www.quantusflm.com) [4, 7].…”

Section: Methodsmentioning

confidence: 99%

Assessment of Fetal Lung Maturity Using Quantitative Ultrasound Analysis in Patients with Prelabor Rupture of Membranes

et al. 2020

View full text Add to dashboard Cite

Introduction: Prelabor rupture of membranes (PROM) is a frequent clinical situation, and the decision about the best time for delivery remains controversial, mainly due to the risk of neonatal respiratory morbidity (NRM). Assessment of fetal lung maturity using ultrasound, a safe method and widely used in current obstetrical practice, could change this scenario. This study was designed to evaluate the ability of quantitative ultrasound method QuantusFLM ® to predict NRM in patients with PROM and whether maternal BMI, gestational age, occurrence of the disease, and presence of oligohydramnios influenced the performance. Methods: Patients with singleton gestations, diagnosis of PROM, and gestational age between 24 and 38 weeks and 6 days were included. Fetal lung image was acquired by ultrasound within 48 h prior to delivery and analyzed by QuantusFLM ®. The results were then paired with neonatal outcomes to assess the program's ability to predict the NRM in this specific group. A logistic regression model was created to analyze factors that could affect the test results. Results: Fifty-four patients were included. Mean maternal BMI was 28.99 kg/ m 2 , and in 25 patients (46.2%), oligohydramnios was observed at the time of examination. Mean gestational age at delivery was 35 weeks and 4 days, and the NRM prevalence was of 18.5%. QuantusFLM ® predicted NRM with a 60% sensitivity, 79.5% specificity, 40% positive predictive value, 89.7% negative predictive value, and 75.6% accuracy. Maternal BMI, disease occurrence, presence of oligohydramnios, and gestational age did not interfere with the evaluation. Conclusion: This study demonstrates a good accuracy of QuantusFLM ® as a NRM predictor in patients with PROM, with particular reliability in identifying that pulmonary maturity has already occurred.

show abstract

“…A computational method was developed for ultrasound neonatal respiratory morbidity prediction system using CNNs [27] . Throughout the secure management of COVID-19 spreads, the usage of ultrasound is now important as it can facilitate the clinical assessment and lung imaging to be conducted simultaneously at a single doctor's bedside [28] .…”

Section: Literature Reviewmentioning

confidence: 99%

COVID-19 and Non-COVID-19 Classification using Multi-layers Fusion From Lung Ultrasound Images

Muhammad

Hossain

2021

Information Fusion

View full text Add to dashboard Cite

COVID-19 or related viral pandemics should be detected and managed without hesitation, since the virus spreads very rapidly. Often with insufficient human and electronic resources, patients need to be checked from stable patients using vital signs, radiographic photographs, or ultrasound images. Vital signs do not often offer the right outcome, and radiographic photos have a variety of other problems. Lung ultrasound (LUS) images can provide good screening without a lot of complications. This paper suggests a model of a convolutionary neural network (CNN) that has fewer learning parameters but can achieve strong accuracy. The model has five main blocks or layers of convolution connectors. A multi-layer fusion functionality of each block is proposed to improve the efficiency of the COVID-19 screening method utilizing the proposed model. Experiments are conducted using freely accessible LUS photographs and video datasets. The proposed fusion method has 92.5% precision, 91.8% accuracy, and 93.2% retrieval using the data collection. These efficiency metric levels are considerably higher than those used in any of the state-of-the-art CNN versions.

show abstract

Evaluation of an improved tool for non-invasive prediction of neonatal respiratory morbidity based on fully automated fetal lung ultrasound analysis

Cited by 42 publications

References 42 publications

Evaluation of deep convolutional neural networks for automatic classification of common maternal fetal ultrasound planes

Evaluation of deep convolutional neural networks for automatic classification of common maternal fetal ultrasound planes

Assessment of Fetal Lung Maturity Using Quantitative Ultrasound Analysis in Patients with Prelabor Rupture of Membranes

COVID-19 and Non-COVID-19 Classification using Multi-layers Fusion From Lung Ultrasound Images

Contact Info

Product

Resources

About