Background/AimTo automatically detect and classify the early stages of retinopathy of prematurity (ROP) using a deep convolutional neural network (CNN).MethodsThis retrospective cross-sectional study was conducted in a referral medical centre in Taiwan. Only premature infants with no ROP, stage 1 ROP or stage 2 ROP were enrolled. Overall, 11 372 retinal fundus images were compiled and split into 10 235 images (90%) for training, 1137 (10%) for validation and 244 for testing. A deep CNN was implemented to classify images according to the ROP stage. Data were collected from December 17, 2013 to May 24, 2019 and analysed from December 2018 to January 2020. The metrics of sensitivity, specificity and area under the receiver operating characteristic curve were adopted to evaluate the performance of the algorithm relative to the reference standard diagnosis.ResultsThe model was trained using fivefold cross-validation, yielding an average accuracy of 99.93%±0.03 during training and 92.23%±1.39 during testing. The sensitivity and specificity scores of the model were 96.14%±0.87 and 95.95%±0.48, 91.82%±2.03 and 94.50%±0.71, and 89.81%±1.82 and 98.99%±0.40 when predicting no ROP versus ROP, stage 1 ROP versus no ROP and stage 2 ROP, and stage 2 ROP versus no ROP and stage 1 ROP, respectively.ConclusionsThe proposed system can accurately differentiate among ROP early stages and has the potential to help ophthalmologists classify ROP at an early stage.
Birdwatching is a common hobby but to identify their species requires the assistance of bird books. To provide birdwatchers a handy tool to admire the beauty of birds, we developed a deep learning platform to assist users in recognizing 27 species of birds endemic to Taiwan using a mobile app named the Internet of Birds (IoB). Bird images were learned by a convolutional neural network (CNN) to localize prominent features in the images. First, we established and generated a bounded region of interest to refine the shapes and colors of the object granularities and subsequently balanced the distribution of bird species. Then, a skip connection method was used to linearly combine the outputs of the previous and current layers to improve feature extraction. Finally, we applied the softmax function to obtain a probability distribution of bird features. The learned parameters of bird features were used to identify pictures uploaded by mobile users. The proposed CNN model with skip connections achieved higher accuracy of 99.00 % compared with the 93.98% from a CNN and 89.00% from the SVM for the training images. As for the test dataset, the average sensitivity, specificity, and accuracy were 93.79%, 96.11%, and 95.37%, respectively.INDEX TERMS Bird image recognition, convolutional neural network, deep learning, mobile app.
Manual inspection and harvesting of ripening tomatoes is time consuming and labor intensive. Smart agriculture can emphasize the use of digital horticultural resources for farming and can increase farm sustainability; to that end, we proposed a fuzzy Mask R-CNN model to automatically identify the ripeness levels of cherry tomatoes. First, to annotate the images automatically, a fuzzy c-means model was used to maintain the spatial information of various foreground and background elements of the image. Then, a Hough transform method was applied to locate the specific geometric edge positions of the tomatoes. Each data point of the image space was annotated to a JavaScript Object Notation file. Second, annotated images were trained with Mask R-CNN to identify each tomato precisely. Finally, to prevent preharvest abscission of tomatoes, a hue-saturation-value color model and fuzzy inference rules were used to predict the ripeness of the tomatoes. A trigonometric function with Euclidian distance was calculated from the origin of calyx and stem to the bottom of the tomato to obtain the position of the pedicle head and dissect the fruit in a timely manner. For detection of 100 tomato images, Mask R-CNN achieved an accuracy of 98.00%. The ripeness classification of tomatoes achieved overall weighted precision and recall rates of 0.9614 and 0.9591, respectively. Thus, automatic tomato harvesting applications can empower farmers to make better decisions and enhance overall production efficiency and yield. INDEX TERMSAutomatic annotation, detection of tomato ripeness, fuzzy c-means, Mask Region-based Convolutional Neural Network (Mask R-CNN), hue-saturation-value (HSV) color model.
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.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.