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 Dec 01 2021
Journal Name
Baghdad Science Journal
A Prevalence study of Entamoeba spp. in Basrah Province using Different Detection Methods
...Show More Authors

This study aims to determine the prevalence of Entamoeba histolytica, Entamoeba dispar and
Entamoeba moshkovskii by three methods of diagnosis (microscopic examination, cultivation and PCR) that
were compared to obtain an accurate diagnosis of Entamoeba spp. during amoebiasis. Total (n=150) stool
samples related to patients were (n = 100) and healthy controls (n= 50). Clinically diagnosed stool samples
(n=100) were collected from patients attending the consultant clinics of different hospitals in Basrah during
the period from January 2018 to January 2019. The results showed that 60% of collected samples were
positive in a direct microscopic examination. All samples were cultivated on different media; the Bra

... Show More
View Publication Preview PDF
Scopus (5)
Crossref (4)
Scopus Clarivate Crossref
Publication Date
Fri Jan 01 2016
Journal Name
Iraqi Journal Of Science
Land cover change detection of Baghdad city using multi-spectral remote sensing imagery
...Show More Authors

Publication Date
Fri Feb 01 2019
Journal Name
Journal Of Physics: Conference Series
Calculation of Concentrations and Transfer Factors of Uranium from Soil to Plants Using Nuclear Track Detector CR-39
...Show More Authors

In this study, the concentrations of uranium for four species of plants; Spinacia, Brassica Oleracea, BEASSICA Oleracea Var Capitata and Beta Vulgaris were measured in addition to the measurement of uranium concentrations in the selected soil by calculating the number of significant traces of alpha in CR-39. The 2.455 Bq/kg in Spinacia plant were the highest concentration while the lowest concentration of uranium were 1.91 Bq/kg in BEASSICA Oleracea Var Capitata plant. As for the transfer factor, the highest value 0.416 were found in Spinacia plant and the lowest value 0.323 were found in BEASSICA Oleracea Var Capitata plant. The uranium in the models studied in it did not exceed the international limit, according to the International Atomi

... Show More
View Publication
Scopus (3)
Crossref (2)
Scopus Clarivate Crossref
Publication Date
Sat Jun 01 2024
Journal Name
Journal Of Ecological Engineering
Using Machine Learning Algorithms to Predict the Sweetness of Bananas at Different Drying Times
...Show More Authors

View Publication
Scopus (4)
Crossref (6)
Scopus Clarivate Crossref
Publication Date
Mon May 06 2024
Journal Name
Journal Of Ecological Engineering
Using Machine Learning Algorithms to Predict the Sweetness of Bananas at Different Drying Times
...Show More Authors

The consumption of dried bananas has increased because they contain essential nutrients. In order to preserve bananas for a longer period, a drying process is carried out, which makes them a light snack that does not spoil quickly. On the other hand, machine learning algorithms can be used to predict the sweetness of dried bananas. The article aimed to study the effect of different drying times (6, 8, and 10 hours) using an air dryer on some physical and chemical characteristics of bananas, including CIE-L*a*b, water content, carbohydrates, and sweetness. Also predicting the sweetness of dried bananas based on the CIE-L*a*b ratios using machine learn- ing algorithms RF, SVM, LDA, KNN, and CART. The results showed that increasing the drying

... Show More
Preview PDF
Scopus (4)
Crossref (6)
Scopus Clarivate Crossref
Publication Date
Tue Apr 30 2024
Journal Name
Iraqi Geological Journal
Wellbore Instability Analysis to Determine the Safe Mud Weight Window for Deep Well, Halfaya Oilfield
...Show More Authors

Wellbore instability is one of the most common issues encountered during drilling operations. This problem becomes enormous when drilling deep wells that are passing through many different formations. The purpose of this study is to evaluate wellbore failure criteria by constructing a one-dimensional mechanical earth model (1D-MEM) that will help to predict a safe mud-weight window for deep wells. An integrated log measurement has been used to compute MEM components for nine formations along the studied well. Repeated formation pressure and laboratory core testing are used to validate the calculated results. The prediction of mud weight along the nine studied formations shows that for Ahmadi, Nahr Umr, Shuaiba, and Zubair formations

... Show More
View Publication
Scopus Crossref
Publication Date
Wed May 10 2023
Journal Name
Biomass Conversion And Biorefinery
Lactic acid-based deep eutectic solvents and activated carbon for soap removal from crude biodiesel
...Show More Authors

View Publication
Scopus (9)
Crossref (4)
Scopus Clarivate Crossref
Publication Date
Thu Jan 12 2017
Journal Name
Journal Of Agricultural And Food Chemistry
Using Standing Gold Nanorod Arrays as Surface-Enhanced Raman Spectroscopy (SERS) Substrates for Detection of Carbaryl Residues in Fruit Juice and Milk
...Show More Authors

View Publication
Scopus (116)
Crossref (106)
Scopus Clarivate Crossref
Publication Date
Sat Oct 29 2022
Journal Name
International Ophthalmology
Comparison of corneal flap thickness predictability and architecture between femtosecond laser and sub-Bowman keratomileusis microkeratome in laser in situ keratomileusis
...Show More Authors

View Publication
Scopus (4)
Crossref (2)
Scopus Clarivate Crossref
Publication Date
Sat Oct 01 2011
Journal Name
Journal Of Engineering
IMPROVED IMAGE COMPRESSION BASED WAVELET TRANSFORM AND THRESHOLD ENTROPY
...Show More Authors

In this paper, a method is proposed to increase the compression ratio for the color images by
dividing the image into non-overlapping blocks and applying different compression ratio for these
blocks depending on the importance information of the block. In the region that contain important
information the compression ratio is reduced to prevent loss of the information, while in the
smoothness region which has not important information, high compression ratio is used .The
proposed method shows better results when compared with classical methods(wavelet and DCT).

View Publication Preview PDF
Crossref