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.
The objective of this study was to evaluate the performance of a new version of quantusFLM®, a software tool for prediction of neonatal respiratory morbidity (NRM) by ultrasound, which incorporates a fully automated fetal lung delineation based on Deep Learning techniques. A set of 790 fetal lung ultrasound images obtained at 24 + 0–38 + 6 weeks’ gestation was evaluated. Perinatal outcomes and the occurrence of NRM were recorded. quantusFLM® version 3.0 was applied to all images to automatically delineate the fetal lung and predict NRM risk. The test was compared with the same technology but using a manual delineation of the fetal lung, and with a scenario where only gestational age was available. The software predicted NRM with a sensitivity, specificity, and positive and negative predictive value of 71.0%, 94.7%, 67.9%, and 95.4%, respectively, with an accuracy of 91.5%. The accuracy for predicting NRM obtained with the same texture analysis but using a manual delineation of the lung was 90.3%, and using only gestational age was 75.6%. To sum up, automated and non-invasive software predicted NRM with a performance similar to that reported for tests based on amniotic fluid analysis and much greater than that of gestational age alone.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Objectives: To evaluate a novel Artificial Intelligence (AI) method for the detection of malignant skin lesions from dermoscopic images.
Methods: 58,457 dermoscopic images available online from the International Skin Imaging Collaboration (ISIC) Archive were downloaded. These images were acquired from different centers worldwide by recognized dermatologists and show varied clinical outcomes belonging to different types of benign and malign skin lesions. A state-of-the-art AI skin lesion classifier based on Deep Learning was designed. The method, fully automated, first locates and segments the nevus in the image and then classifies it into either benign or malign type.
Results: 1,631 images (2.8%) were discarded due to bad quality. A total of 56,826 images were finally used. Two thirds of the images (37,688) were used to train the classifier, leaving the remaining 19,138 images for validation. In this set, malignant lesions had a prevalence of 15.4% (2,956/19,138). The AI skin lesion classifier reached an area under the curve (AUC) of 87.4%. Optimal cut-off point in terms of accuracy resulted in an 85.9% accuracy (16,439/19,138) and sensibility of 89.6% (2,648/ 2,956) at 85.2% (13,791/16,182) specificity. Negative predictive value (NPV) was 97.8% (13,791/14,099). Other training/validation splits were also evaluated, showing similar results.
Conclusions: A novel AI method showed promising results as skin lesion classifier from dermoscopic images. Its high NPV value could make it suited for high-risk patient screening. A large clinical study to confirm these results is needed and will be pursued.
The objective of this study was to evaluate a novel automated test based on ultrasound cervical texture analysis to predict spontaneous Preterm Birth (sPTB) alone and in combination with Cervical Length (CL). General population singleton pregnancies between 18 + 0 and 24 + 6 weeks’ gestation were assessed prospectively at two centers. Cervical ultrasound images were evaluated and the occurrence of sPTB before weeks 37 + 0 and 34 + 0 were recorded. CL was measured on-site. The automated texture analysis test was applied offline to all images. Their performance to predict the occurrence of sPTB before 37 + 0 and 34 + 0 weeks was evaluated separately and in combination on 633 recruited patients. AUC for sPTB prediction before weeks 37 and 34 respectively were as follows: 55.5% and 65.3% for CL, 63.4% and 66.3% for texture analysis, 67.5% and 76.7% when combined. The new test improved detection rates of CL at similar low FPR. Combining the two increased detection rate compared to CL alone from 13.0 to 30.4% for sPTB < 37 and from 14.3 to 42.9% sPTB < 34. Texture analysis of cervical ultrasound improved sPTB detection rate compared to cervical length for similar FPR, and the two combined together increased significantly prediction performance. This results should be confirmed in larger cohorts.
To evaluate the concordance of the risk of neonatal respiratory morbidity (NRM) assessed by quantitative ultrasound lung texture analysis (QuantusFLM) between twin fetuses of the same pregnancy. Prospective study conducted in twin pregnancies. Fetal ultrasound lung images were obtained at 26.0–38.6 weeks of gestation. Categorical (high or low) and continuous results of the risk of NRM were compared between twins. Fetal ultrasound lung images from 131 pairs (262 images) of twins were included. The images were classified into three gestational age ranges: Group 1 (26.0–29.6 weeks, 78 images, 39 pairs [29.8%]); Group 2 (30.0–33.6 weeks, 98 images, 49 pairs [37.4%]) and Group 3 (34.0–38.6 weeks, 86 images, 43 pairs [32.8%]). Concordance was good in Groups 1 and 3 and moderate in Group 2. In Groups 2 and 3 at least one fetus presented high-risk results in 26.5% and 11.6% of twin pairs, respectively. Only gestational age < 32 weeks, gestational diabetes mellitus, and spontaneous conception were associated with a high risk of NRM in Group 2. There was good concordance of the risk of NRM between twins < 30.0 weeks and > 34.0 weeks. From 30.0 to 33.6 weeks 26.5% of the twin pairs had discordant results, with moderate concordance of the risk of NRM.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.