scholarly journals Automated detection of early-stage ROP using a deep convolutional neural network

2020 ◽  
pp. bjophthalmol-2020-316526
Author(s):  
Yo-Ping Huang ◽  
Haobijam Basanta ◽  
Eugene Yu-Chuan Kang ◽  
Kuan-Jen Chen ◽  
Yih-Shiou Hwang ◽  
...  
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Early Stage ◽  
Characteristic Curve ◽  
Cross Sectional Study ◽  
Deep Convolutional Neural Network ◽  
Cross Sectional ◽  
Early Stages ◽  
Sensitivity Specificity ◽  
Stage 1

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.

scholarly journals Application of a deep convolutional neural network in the diagnosis of neonatal ocular fundus hemorrhage

2018 ◽  
Vol 38 (6) ◽  
Author(s):  
Binbin Wang ◽  
Li Xiao ◽  
Yang Liu ◽  
Jing Wang ◽  
Beihong Liu ◽  
...  
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
High Performance ◽  
Binary Classification ◽  
Characteristic Curve ◽  
Classification Problem ◽  
Retinal Hemorrhage ◽  
Multiclass Classification ◽  
Deep Convolutional Neural Network ◽  
Cross Sectional

There is a disparity between the increasing application of digital retinal imaging to neonatal ocular screening and slowly growing number of pediatric ophthalmologists. Assistant tools that can automatically detect ocular disorders may be needed. In present study, we develop a deep convolutional neural network (DCNN) for automated classification and grading of retinal hemorrhage. We used 48,996 digital fundus images from 3770 newborns with retinal hemorrhage of different severity (grade 1, 2 and 3) and normal controls from a large cross-sectional investigation in China. The DCNN was trained for automated grading of retinal hemorrhage (multiclass classification problem: hemorrhage-free and grades 1, 2 and 3) and then validated for its performance level. The DCNN yielded an accuracy of 97.85 to 99.96%, and the area under the receiver operating characteristic curve was 0.989–1.000 in the binary classification of neonatal retinal hemorrhage (i.e., one classification vs. the others). The overall accuracy with regard to the multiclass classification problem was 97.44%. This is the first study to show that a DCNN can detect and grade neonatal retinal hemorrhage at high performance levels. Artificial intelligence will play more positive roles in ocular healthcare of newborns and children.

Identification of Tomato Leaf Disease Detection using Pretrained Deep Convolutional Neural Network Models

2020 ◽  
Vol 21 (4) ◽  
pp. 625-635
Author(s):  
Anandhakrishnan T ◽  
Jaisakthi S.M Murugaiyan
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Early Stage ◽  
Computing Time ◽  
Network Models ◽  
Deep Convolutional Neural Network ◽  
Neural Network Models ◽  
Leaf Disease ◽  
Proposed Model ◽  
Deep Cnn

In this paper, we proposed a plant leaf disease identification model based on a Pretrained deep convolutional neural network (Deep CNN). The Deep CNN model is trained using an open dataset with 10 different classes of tomato leaves We observed that overall architectures which can increase the best performance of the model. The proposed model was trained using different training epochs, batch sizes and dropouts. The Xception has attained maximum accuracy compare with all other approaches. After an extensive simulation, the proposed model achieves classification accuracy better. This accuracy of the proposed work is greater than the accuracy of all other Pretrained approaches. The proposed model is also tested with respect to its consistency and reliability. The set of data used for this work was collected from the plant village dataset, including sick and healthy images. Models for detection of plant disease should predict the disease quickly and accurately in the early stage itself so that a proper precautionary measures can be applied to avoid further spread of the diseases. So, to reduce the main issue about the leaf diseases, we can analyze distinct kinds of deep neural network architectures in this research. From the outcomes, Xception has a constantly improving more to enhance the accuracy by increasing the number of epochs, without any indications of overfitting and decreasein quality. And Xception also generated a fine 99.45% precision in less computing time.

scholarly journals Skin Lesion Classification With Deep Convolutional Neural Network: Process Development and Validation

10.2196/18438 ◽  
2020 ◽  
Vol 3 (1) ◽  
pp. e18438
Author(s):  
Arnab Ray ◽  
Aman Gupta ◽  
Amutha Al
Keyword(s):  
Neural Network ◽  
Skin Cancer ◽  
Skin Lesion ◽  
Convolutional Neural Network ◽  
Process Development ◽  
Early Stage ◽  
Deep Convolutional Neural Network ◽  
Fine Tuning ◽  
Network Process ◽  
Lesion Classification

Background Skin cancer is the most common cancer and is often ignored by people at an early stage. There are 5.4 million new cases of skin cancer worldwide every year. Deaths due to skin cancer could be prevented by early detection of the mole. Objective We propose a skin lesion classification system that has the ability to detect such moles at an early stage and is able to easily differentiate between a cancerous and noncancerous mole. Using this system, we would be able to save time and resources for both patients and practitioners. Methods We created a deep convolutional neural network using an Inceptionv3 and DenseNet-201 pretrained model. Results We found that using the concepts of fine-tuning and the ensemble learning model yielded superior results. Furthermore, fine-tuning the whole model helped models converge faster compared to fine-tuning only the top layers, giving better accuracy overall. Conclusions Based on our research, we conclude that deep learning algorithms are highly suitable for classifying skin cancer images.

scholarly journals A Study on Hypertension among Persons Living with HIV attending ART Centers in TamilNadu- A Descriptive Cross-Sectional Study

2020 ◽  
Vol 11 (4) ◽  
pp. 7078-7082
Author(s):  
Krishna Prasanth B ◽  
Mahalakshmi K ◽  
Kalpana S ◽  
Anantha Eashwar V M
Keyword(s):  
Tamil Nadu ◽  
Early Stage ◽  
Cross Sectional Study ◽  
Sectional Study ◽  
Cross Sectional ◽  
Mortality And Morbidity ◽  
Persons Living With Hiv ◽  
Education And Awareness ◽  
Stage 1 ◽  
Study Participants

People suffering from immunosuppressive conditions like Human Immunodeficiency Virus (HIV) are more prone to suffer from non-communicable diseases like hypertension, which is not identified and treated at an earlier stage, can lead to significant mortality and morbidity in them. The study design was a cross-sectional study done in select Anti-Retroviral Therapy (ART) centers in Government hospitals in Tamil Nadu during a period from 2017-2018. Data regarding their HIV status, treatment history and Body Mass Index (BMI) were recorded and Blood Pressure (BP) was recorded by using mercury sphygmomanometer using standard guidelines. Data was entered in Microsoft excel and analyzed by using SPSS version 22 software. The study population comprised of 75% males and 25% females. Mean age of study participants was 45+8.2 years and the mean BMI was 22+3.4. The prevalence of hypertension among HIV affected individuals was found to be 14.63%. According to JNC criteria, 34% were having Stage 1 hypertension and 33% had Stage 2 hypertension and only 23% were on treatment. Significant association was found between increasing age, gender, BMI and hypertension. Health education and awareness creation has to be created among HIV patients on maintaining a healthy diet and lifestyle so that, obesity can be prevented or reduced which could play an important role in NCD’s like hypertension and also hypertensive patients have to be identified and treated at early stage to reduce the morbidity and mortality.

scholarly journals Classification of plant seedlings using deep convolutional neural network architectures

2022 ◽  
Vol 2161 (1) ◽  
pp. 012006
Author(s):  
Namratha Makanapura ◽  
C Sujatha ◽  
Prakash R Patil ◽  
Padmashree Desai
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Weed Management ◽  
Early Stage ◽  
Vital Role ◽  
Deep Convolutional Neural Network ◽  
Network Architectures ◽  
Weed Species ◽  
Neural Network Architectures ◽  
Plant Seedlings

Abstract Weed management has a vital role in applications of agriculture domain. One of the key tasks is to identify the weeds after few days of plant germination which helps the farmers to perform early-stage weed management to reduce the contrary impacts on crop growth. Thus, we aim to classify the seedlings of crop and weed species. In this work, we propose a plant seedlings classification using the benchmark plant seedlings dataset. The dataset contains the images of 12 different species where three belongs to plant species and the other nine belongs to weed species. We implement the classification framework using three different deep convolutional neural network architectures, namely ResNet50V2, MobileNetV2 and EfficientNetB0. We train the models using transfer learning and compare the performance of each model on a test dataset of 833 images. We compare the three models and demonstrate that the EfficientNetB0 performs better with an average F1-Score of 96.26% and an accuracy of 96.52%.

scholarly journals ACP-MHCNN: An Accurate Multi-Headed Deep-Convolutional Neural Network to Predict Anticancer peptides

2020 ◽  
Author(s):  
Sajid Ahmed ◽  
Rafsanjani Muhammod ◽  
Sheikh Adilina ◽  
Zahid Hossain Khan ◽  
Swakkhar Shatabda ◽  
...  
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Peptide Identification ◽  
Deep Convolutional Neural Network ◽  
Primary Methods ◽  
Anticancer Peptides ◽  
Negative Side Effects ◽  
Sensitivity Specificity ◽  
Identification Techniques ◽  
Therapeutic Alternatives

AbstractAlthough advancing the therapeutic alternatives for treating deadly cancers has gained much attention globally, still the primary methods such as chemotherapy have significant downsides and low specificity. Most recently, Anticancer peptides (ACPs) have emerged as a potential alternative to therapeutic alternatives with much fewer negative side-effects. However, the identification of ACPs through wet-lab experiments is expensive and time-consuming. Hence, computational methods have emerged as viable alternatives. During the past few years, several computational ACP identification techniques using hand-engineered features have been proposed to solve this problem. In this study, we propose a new multi headed deep convolutional neural network model called ACP-MHCNN, for extracting and combining discriminative features from different information sources in an interactive way. Our model extracts sequence, physicochemical, and evolutionary based features for ACP identification through simultaneous interaction with different numerical peptide representations while restraining parameter overhead. It is evident through rigorous experiments using cross-validation and independent-dataset that ACP-MHCNN outperforms other models for anticancer peptide identification by a substantial margin. ACP-MHCNN outperforms state-of-the-art model by 6.3%, 8.6%, 3.7%, 4.0%, and 0.20 in terms of accuracy, sensitivity, specificity, precision, and MCC respectively. ACP-MHCNN and its relevant codes and datasets are publicly available at: https://github.com/mrzResearchArena/Anticancer-Peptides-CNN.

When Time is of the Essence: Preliminary Findings for a Quick Administration of the Dot Counting Test

2019 ◽  
Author(s):  
K Chase Bailey ◽  
Troy A Webber ◽  
Jacob I Phillips ◽  
Lindsay D R Kraemer ◽  
Janice C Marceaux ◽  
...  
Keyword(s):  
Classification Accuracy ◽  
Clinical Sample ◽  
Characteristic Curve ◽  
Cross Sectional Study ◽  
Receiver Operator Characteristic Curve ◽  
Performance Validity ◽  
Free Standing ◽  
Cross Sectional ◽  
Validity Measures ◽  
Sensitivity Specificity

Abstract Objective Performance validity research has emphasized the need for briefer measures and, more recently, abbreviated versions of established free-standing tests to minimize neuropsychological evaluation costs/time burden. This study examined the accuracy of multiple abbreviated versions of the Dot Counting Test (“quick” DCT) for detecting invalid performance in isolation and in combination with the Test of Memory Malingering Trial 1 (TOMMT1). Method Data from a mixed clinical sample of 107 veterans (80 valid/27 invalid per independent validity measures and structured criteria) were included in this cross-sectional study; 47% of valid participants were cognitively impaired. Sensitivities/specificities of various 6- and 4-card DCT combinations were calculated and compared to the full, 12-card DCT. Combined models with the most accurate 6- and 4-card combinations and TOMMT1 were then examined. Results Receiver operator characteristic curve analyses were significant for all 6- and 4-card DCT combinations with areas under the curve of .868–.897. The best 6-card combination (cards, 1-3-5-8-11-12) had 56% sensitivity/90% specificity (E-score cut-off, ≥14.5), and the best 4-card combination (cards, 3-4-8-11) had 63% sensitivity/94% specificity (cut-off, ≥16.75). The full DCT had 70% sensitivity/90% specificity (cut-off, ≥16.00). Logistic regression revealed 95% classification accuracy when 6-card or 4-card “quick” combinations were combined with TOMMT1, with the DCT combinations and TOMMT1 both emerging as significant predictors. Conclusions Abbreviated DCT versions utilizing 6- and 4-card combinations yielded comparable sensitivity/specificity as the full DCT. When these “quick” DCT combinations were further combined with an abbreviated memory-based performance validity test (i.e., TOMMT1), overall classification accuracy for identifying invalid performance was 95%.

Ratio of height to thyromental distance: a comparison with Mallampati and Upper lip Bite test for predicting difficult intubation in Pakistani population

2021 ◽  
pp. 1-15
Author(s):  
Saman Tauheed Ali ◽  
Khalid Samad ◽  
Syed Amir Raza ◽  
Muhammad Qamarul Hoda
Keyword(s):  
Predictive Value ◽  
Characteristic Curve ◽  
Cross Sectional Study ◽  
Difficult Laryngoscopy ◽  
University Hospital ◽  
Upper Lip ◽  
Cross Sectional ◽  
Elective Surgical Procedures ◽  
Lehane Grade ◽  
Sensitivity Specificity

Objectives: We conducted this study to compare the accuracy of three diagnostic tests; ratio of height to thyromental distance (RHTMD), Modified Mallampati Test (MMT) and Upper Lip Bite Test (ULBT) in predicting difficult laryngoscopy using Cormack and Lehane grade as gold standard.Methods: This study was conducted in Aga Khan University Hospital, Karachi. Based on calculated sample size, 383 patients who required endotracheal intubation for elective surgical procedures were enrolled with consecutive sampling techniques during August 2014 to August 2015 for this cross-sectional study. Primary investigator used RHTMD, ULBT, and MMT for assessing the airway and correlated with laryngoscopic view.Results: A total of 383 patients were incorporated in this research, out of which 59(15.4%) classified as difficult laryngoscopy based on Cormack and Lehane (CL) grading. Sensitivity, specificity, positive predictive value, negative predictive value and accuracy of RHTMD (84.7%, 90.1%, 60.9%, 97%, 89.3%) and ULBT (83.1%, 89.2%, 58.3%, 96.7%, 88.3%) values were highest as compared to MMT (30.5%, 84.3%, 26.1%, 86.9%, 79.9%). The area under a receiver-operating characteristic curve (AUC of ROC curve) for ULBT and RHTMD was significantly more than for MMT (P<0.01). RHTMD and ULBT both are acceptable alternatives for prediction of difficult laryngoscopy as a simple, single bed-side test. Continuous...

Comparing the accuracy of lay diagnosis of childhood malaria and pneumonia with that of the revised IMCI guidelines in Nigeria

2020 ◽  
Author(s):  
K O Elimian ◽  
P R Myles ◽  
R Phalkey ◽  
A Sadoh ◽  
C Pritchard
Keyword(s):  
Health Workers ◽  
Characteristic Curve ◽  
Cross Sectional Study ◽  
Childhood Diseases ◽  
Cross Sectional ◽  
Predictive Values ◽  
Benin City ◽  
Sensitivity Specificity ◽  
Childhood Malaria ◽  
Pneumonia Diagnosis

Abstract Background Improving caregivers’ recognition of childhood malaria and pneumonia is crucial to early treatment and improving outcomes. The objective of this study was to assess the accuracy and reliability of caregivers’ recognition of malaria and pneumonia (lay diagnosis) as compared to the revised IMCI guidelines. Methods A cross-sectional study design was used to recruit 903 children aged 2–59 months who were assessed for malaria and pneumonia by health workers at five primary healthcare centres in Benin City, Nigeria. Accuracy of lay diagnosis as compared to the revised IMCI guidelines was assessed using sensitivity, specificity, positive and negative predictive values and area under the receiver operating characteristic curve (AUROC) values. Results The accuracy of caregivers’ ability to recognise malaria (AUROC: 0.60; 95% CI: 0.57–0.64) and pneumonia (AUROC: 0.54; 95% CI: 0.50–0.58) was, respectively, moderate and poor as compared to the IMCI guidelines. Caregivers were better able to identify children without than those with malaria and pneumonia. Agreement between caregivers and the IMCI guidelines for malaria and pneumonia diagnosis was poor (k = 0.14, 95% CI: 0.09–0.19; P = 0.0001). Conclusion Caregivers’ ability to recognise these childhood diseases as compared to the IMCI guidelines was poor overall, which was partly due to the approach used to ascertain lay diagnosis.

Sign in / Sign up

Export Citation Format

Share Document