Preferred Language
Articles
/
fBb2j4oBVTCNdQwCuJ_G
Deep Transfer Learning for Improved Detection of Keratoconus using Corneal Topographic Maps
...Show More Authors
Abstract <p>Clinical keratoconus (KCN) detection is a challenging and time-consuming task. In the diagnosis process, ophthalmologists must revise demographic and clinical ophthalmic examinations. The latter include slit-lamb, corneal topographic maps, and Pentacam indices (PI). We propose an Ensemble of Deep Transfer Learning (EDTL) based on corneal topographic maps. We consider four pretrained networks, SqueezeNet (SqN), AlexNet (AN), ShuffleNet (SfN), and MobileNet-v2 (MN), and fine-tune them on a dataset of KCN and normal cases, each including four topographic maps. We also consider a PI classifier. Then, our EDTL method combines the output probabilities of each of the five classifiers to obtain a decision based on the fusion of probabilities. Individually, the classifier based on PI achieved 93.1% accuracy, whereas the deep classifiers reached classification accuracies over 90% only in isolated cases. Overall, the average accuracy of the deep networks over the four corneal maps ranged from 86% (SfN) to 89.9% (AN). The classifier ensemble increased the accuracy of the deep classifiers based on corneal maps to values ranging (92.2% to 93.1%) for SqN and (93.1% to 94.8%) for AN. Including in the ensemble-specific combinations of corneal maps’ classifiers and PI increased the accuracy to 98.3%. Moreover, visualization of first learner filters in the networks and Grad-CAMs confirmed that the networks had learned relevant clinical features. This study shows the potential of creating ensembles of deep classifiers fine-tuned with a transfer learning strategy as it resulted in an improved accuracy while showing learnable filters and Grad-CAMs that agree with clinical knowledge. This is a step further towards the potential clinical deployment of an improved computer-assisted diagnosis system for KCN detection to help ophthalmologists to confirm the clinical decision and to perform fast and accurate KCN treatment.</p>
Scopus Clarivate Crossref
View Publication
Publication Date
Mon Mar 27 2023
Journal Name
8th Engineering And 2nd International Conference For College Of Engineering – University Of Baghdad: Coec8-2021 Proceedings
Experimental study of natural convection heat transfer on an enclosure partially filled porous medium heated from below by constant heat flux
...Show More Authors

This study reports on natural convection heat transfer in a square enclosure of length (L=20 cm) with a saturated porous medium (solid glass beads) having same fluid (air) at lower horizontal layer and free air fill in the rest of the cavity's space. The experimental work has been performed under the effects of heating from bottom by constant heat flux q=150,300,450,600 W/m2 for four porous layers thickness Hp (2.5,5,7.5,1) cm and three heaters length δ(20,14,7) cm. The top enclosure wall was good insulated and the two side walls were symmetrically cooled at constant temperature. Four layers of porous media with small porosity, Rayleigh number range (60.354 - 241.41) and (Da) 3.025x10-8 has been investigated. The obtained data of temperatu

... Show More
View Publication
Scopus (5)
Crossref (5)
Scopus Crossref
Publication Date
Tue Dec 25 2018
Journal Name
Summaries Of Working Papers, Research And Experiments
E-learning at the College of Mass Communication, subject: public relations campaigns as a model
...Show More Authors

Publication Date
Thu Aug 20 2020
Journal Name
Arab World English Journal
Investigating the Effectiveness of YouTube as a Learning Tool among EFL Students at Baghdad University
...Show More Authors

YouTube is not just a platform that individuals share, upload, comment on videos; teachers and educators can utilize it to the best maximum so that students can have benefits. This study aims at investigating how active and influential YouTube can be in the educational process and how it is beneficial for language teachers to enhance the skills of students. The study demonstrates different theoretical frameworks that tackle the employment of technology to enhance the learning/teaching process. It relies on the strategies of Berk (2009) for using multimedia media, video clips in particular to develop the abilities of teachers for using technology in classrooms. To achieve the objective of the study, the researchers develop a questionnair

... Show More
View Publication
Crossref (2)
Crossref
Publication Date
Fri Aug 19 2022
Journal Name
Mathematical Statistician And Engineering Applications (msea)
Application of TOPSIS method for Multi-Criteria Decision Making for poultry fields By using GIS in Hilla district
...Show More Authors

Statistical methods and statistical decisions making were used to arrange and analyze the primary data to get norms which are used with Geographic Information Systems (GIS) and spatial analysis programs to identify the animals production and poultry units in strategic nutrition channels, also the priorities of food insecurity through the local production and import when there is no capacity for production. The poultry production is one of the most important commodities that satisfy human body protein requirements, also the most important criteria to measure the development and prosperity of nations. The poultry fields of Babylon Governorate are located in Abi Ghareg and Al_Kifil centers according to many criteria or factors such as the popu

... Show More
Publication Date
Sat Apr 01 2023
Journal Name
Journal Of Engineering
Proposed Face Detection Classification Model Based on Amazon Web Services Cloud (AWS)
...Show More Authors

One of the most important features of the Amazon Web Services (AWS) cloud is that the program can be run and accessed from any location. You can access and monitor the result of the program from any location, saving many images and allowing for faster computation. This work proposes a face detection classification model based on AWS cloud aiming to classify the faces into two classes: a non-permission class, and a permission class, by training the real data set collected from our cameras. The proposed Convolutional Neural Network (CNN) cloud-based system was used to share computational resources for Artificial Neural Networks (ANN) to reduce redundant computation. The test system uses Internet of Things (IoT) services through our ca

... Show More
View Publication
Crossref (6)
Crossref
Publication Date
Thu May 30 2024
Journal Name
Iraqi Journal Of Science
A Review Study on Forgery and Tamper Detection Techniques in Digital Images
...Show More Authors

Digital tampering identification, which detects picture modification, is a significant area of image analysis studies. This area has grown with time with exceptional precision employing machine learning and deep learning-based strategies during the last five years. Synthesis and reinforcement-based learning techniques must now evolve to keep with the research. However, before doing any experimentation, a scientist must first comprehend the current state of the art in that domain. Diverse paths, associated outcomes, and analysis lay the groundwork for successful experimentation and superior results. Before starting with experiments, universal image forensics approaches must be thoroughly researched. As a result, this review of variou

... Show More
View Publication
Scopus Crossref
Publication Date
Wed Sep 07 2022
Journal Name
2022 Iraqi International Conference On Communication And Information Technologies (iiccit)
Construct an Efficient DDoS Attack Detection System Based on RF-C4.5-GridSearchCV
...Show More Authors

View Publication
Scopus (5)
Crossref (3)
Scopus Crossref
Publication Date
Sat Apr 01 2023
Journal Name
Journal Of Engineering
Proposed Face Detection Classification Model Based on Amazon Web Services Cloud (AWS)
...Show More Authors

One of the most important features of the Amazon Web Services (AWS) cloud is that the program can be run and accessed from any location. You can access and monitor the result of the program from any location, saving many images and allowing for faster computation. This work proposes a face detection classification model based on AWS cloud aiming to classify the faces into two classes: a non-permission class, and a permission class, by training the real data set collected from our cameras. The proposed Convolutional Neural Network (CNN) cloud-based system was used to share computational resources for Artificial Neural Networks (ANN) to reduce redundant computation. The test system uses Internet of Things (IoT) services th

... Show More
View Publication Preview PDF
Crossref (6)
Crossref
Publication Date
Fri Apr 30 2021
Journal Name
International Journal Of Intelligent Engineering And Systems
SMS Spam Detection Based on Fuzzy Rules and Binary Particle Swarm Optimization
...Show More Authors

View Publication
Scopus (12)
Crossref (6)
Scopus Crossref
Publication Date
Fri Jan 01 2021
Journal Name
Ieee Access
Fast Shot Boundary Detection Based on Separable Moments and Support Vector Machine
...Show More Authors

View Publication
Scopus (29)
Crossref (27)
Scopus Clarivate Crossref