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
Wed Apr 28 2021
Journal Name
2021 1st Babylon International Conference On Information Technology And Science (bicits)
Enhanced Twitter Community Detection using Node Content and Attributes
...Show More Authors

View Publication
Scopus (3)
Crossref (2)
Scopus Crossref
Publication Date
Sun Sep 01 2013
Journal Name
International Journal Of Advanced Research In Computer Science And Software Engineering
Real Time Motion Detection in Surveillance Camera Using MATLAB
...Show More Authors

Surveillance cameras are video cameras used for the purpose of observing an area. They are often connected to a recording device or IP network, and may be watched by a security guard or law enforcement officer. In case of location have less percentage of movement (like home courtyard during night); then we need to check whole recorded video to show where and when that motion occur which are wasting in time. So this paper aims at processing the real time video captured by a Webcam to detect motion in the Scene using MATLAB 2012a, with keeping in mind that camera still recorded which means real time detection. The results show accuracy and efficiency in detecting motion

Preview PDF
Publication Date
Sun Mar 01 2020
Journal Name
Computer Networks
An improved multi-objective evolutionary algorithm for detecting communities in complex networks with graphlet measure
...Show More Authors

View Publication
Scopus (8)
Crossref (5)
Scopus Clarivate Crossref
Publication Date
Tue Feb 20 2024
Journal Name
Aip Conference Proceeding
Evaluation and production of predictive maps on the impact of climate factors on the the land cover for the Baghdad city for the period (1999-2021)
...Show More Authors

The general objective of the research is to better understand changes in land cover and their impact on climatic factors by measuring changes in land cover for the Baghdad city for the period 1999-2021 and evaluating changes in land cover and measuring changes in climatic factors (relative humidity and evaporation). This study from 1999 to 2021 and in two different seasons: the April of the growing season and August the dry season. When using the supervised classification method to determine the differences, the results showed remarkable changes, the study showed the spatial variations in LC from 1999 to 2021 as follows: increase in the vegetation and water bodies during April and decrease this in August while the soil and built up decreas

... Show More
Scopus (3)
Crossref (1)
Scopus Crossref
Publication Date
Sun Jun 06 2010
Journal Name
Baghdad Science Journal
Algebraic Coincidence Periods Of Self – Maps Of A Rational Exterior Space Of Rank 2
...Show More Authors

Let f and g be a self – maps of a rational exterior space . A natural number m is called a minimal coincidence period of maps f and g if f^m and g^m have a coincidence point which is not coincidence by any earlier iterates. This paper presents a complete description of the set of algebraic coincidence periods for self - maps of a rational exterior space which has rank 2 .

View Publication Preview PDF
Crossref
Publication Date
Sun Aug 01 2021
Journal Name
Sensors And Actuators B: Chemical
Sensitive and simple colorimetric methods for visual detection and quantitative determination of semicarbazide in flour products using colorimetric reagents
...Show More Authors

After baking the flour, azodicarbonamide, an approved food additive, can be converted into carcinogenic semicarbazide hydrochloride (SEM) and biurea in flour products. Thus, determine SEM in commercial bread products is become mandatory and need to be performed. Therefore, two accurate, precision, simple and economics colorimetric methods have been developed for the visual detection and quantitative determination of SEM in commercial flour products. The 1st method is based on the formation of a blue-coloured product with λmax at 690 nm as a result of a reaction between the SEM and potassium ferrocyanide in an acidic medium (pH 6.0). In the 2nd method, a brownish-green colored product is formed due to the reaction between the SEM and phosph

... Show More
View Publication
Crossref (27)
Crossref
Publication Date
Mon Jun 30 2025
Journal Name
Ingénierie Des Systèmes D Information
Comparative Analysis of Four Programming Languages for Machine Learning
...Show More Authors

View Publication
Scopus (2)
Scopus Crossref
Publication Date
Mon Mar 14 2022
Journal Name
Periodicals Of Engineering And Natural Sciences (pen)
Mathematical simulation of memristive for classification in machine learning
...Show More Authors

View Publication
Scopus Crossref
Publication Date
Mon Aug 01 2011
Journal Name
Journal Of Engineering
INVESTIGATION OF TWISTED TAPE TURBULATOR FOR FIRE TUBE BOILER Part I. Heat Transfer
...Show More Authors

The present work presents a new experimental study of the enhancement of turbulent
convection heat transfer inside tubes for combined thermal and hydrodynamic entry length of one
popular “turbulator” (twisted tape with width slightly less than internal tube diameter) inserted for
fire tube boilers. Cylindrical combustion chamber was used to burn (1.6 to 7kg/h) fuel oil #2 to
deliver hot gases with ranges of Reynolds number (10500 to 21700), and (11400 to 24150) for both
empty and inserted tube respectively.A uniform wall temperature technique was used by keeping
approximately constant water temperature difference (25ºC) between inlet and exit cooling water in
parallel flow shell and tube heat exchanger. The test

... Show More
View Publication Preview PDF
Crossref (1)
Crossref
Publication Date
Wed May 03 2023
Journal Name
Periodicals Of Engineering And Natural Sciences (pen)
Enhancing smart home energy efficiency through accurate load prediction using deep convolutional neural networks
...Show More Authors

The method of predicting the electricity load of a home using deep learning techniques is called intelligent home load prediction based on deep convolutional neural networks. This method uses convolutional neural networks to analyze data from various sources such as weather, time of day, and other factors to accurately predict the electricity load of a home. The purpose of this method is to help optimize energy usage and reduce energy costs. The article proposes a deep learning-based approach for nonpermanent residential electrical ener-gy load forecasting that employs temporal convolutional networks (TCN) to model historic load collection with timeseries traits and to study notably dynamic patterns of variants amongst attribute par

... Show More
View Publication
Crossref