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 Jan 01 2024
Journal Name
Fifth International Conference On Applied Sciences: Icas2023
Facial deepfake performance evaluation based on three detection tools: MTCNN, Dlib, and MediaPipe
...Show More Authors

View Publication
Scopus (4)
Crossref (4)
Scopus Crossref
Publication Date
Tue Dec 20 2022
Journal Name
2022 4th International Conference On Current Research In Engineering And Science Applications (iccresa)
Noise Detection and Removing in Heart Sound Signals via Nuclear Norm Minimization Problems
...Show More Authors

Heart sound is an electric signal affected by some factors during the signal's recording process, which adds unwanted information to the signal. Recently, many studies have been interested in noise removal and signal recovery problems. The first step in signal processing is noise removal; many filters are used and proposed for treating this problem. Here, the Hankel matrix is implemented from a given signal and tries to clean the signal by overcoming unwanted information from the Hankel matrix. The first step is detecting unwanted information by defining a binary operator. This operator is defined under some threshold. The unwanted information replaces by zero, and the wanted information keeping in the estimated matrix. The resulting matrix

... Show More
View Publication
Scopus (2)
Crossref (2)
Scopus Crossref
Publication Date
Sat Apr 01 2023
Journal Name
Baghdad Science Journal
Interior Visual Intruders Detection Module Based on Multi-Connect Architecture MCA Associative Memory
...Show More Authors

Most recent studies have focused on using modern intelligent techniques spatially, such as those
developed in the Intruder Detection Module (IDS). Such techniques have been built based on modern
artificial intelligence-based modules. Those modules act like a human brain. Thus, they should have had the
ability to learn and recognize what they had learned. The importance of developing such systems came after
the requests of customers and establishments to preserve their properties and avoid intruders’ damage. This
would be provided by an intelligent module that ensures the correct alarm. Thus, an interior visual intruder
detection module depending on Multi-Connect Architecture Associative Memory (MCA)

... Show More
View Publication Preview PDF
Scopus (1)
Scopus Crossref
Publication Date
Mon Apr 01 2019
Journal Name
Journal Of Educational And Psychological Researches
The Teaching Practices of Faculty Members in Northern Border University According to the Brain-Based Learning Theory
...Show More Authors

The present study aims to identify the most and the least common teaching practices among faculty members in Northern Border University according to brain-based learning theory, as well as to identify the effect of sex, qualifications, faculty type, and years of experiences in teaching practices. The study sample consisted of (199) participants divided into 100 males and 99 females. The study results revealed that the most teaching practice among the study sample was ‘I am trying to create an Environment of encouragement and support within the classroom which found to be (4.4623). As for the least teaching practice was ‘I use a natural musical sounds to create student's mood to learn’ found to be (2.2965). The study results also in

... Show More
View Publication Preview PDF
Publication Date
Thu Oct 29 2020
Journal Name
Complexity
Training and Testing Data Division Influence on Hybrid Machine Learning Model Process: Application of River Flow Forecasting
...Show More Authors

The hydrological process has a dynamic nature characterised by randomness and complex phenomena. The application of machine learning (ML) models in forecasting river flow has grown rapidly. This is owing to their capacity to simulate the complex phenomena associated with hydrological and environmental processes. Four different ML models were developed for river flow forecasting located in semiarid region, Iraq. The effectiveness of data division influence on the ML models process was investigated. Three data division modeling scenarios were inspected including 70%–30%, 80%–20, and 90%–10%. Several statistical indicators are computed to verify the performance of the models. The results revealed the potential of the hybridized s

... Show More
View Publication
Scopus (65)
Crossref (36)
Scopus Clarivate Crossref
Publication Date
Fri Mar 01 2024
Journal Name
Baghdad Science Journal
Exploring the Challenges of Diagnosing Thyroid Disease with Imbalanced Data and Machine Learning: A Systematic Literature Review
...Show More Authors

Thyroid disease is a common disease affecting millions worldwide. Early diagnosis and treatment of thyroid disease can help prevent more serious complications and improve long-term health outcomes. However, thyroid disease diagnosis can be challenging due to its variable symptoms and limited diagnostic tests. By processing enormous amounts of data and seeing trends that may not be immediately evident to human doctors, Machine Learning (ML) algorithms may be capable of increasing the accuracy with which thyroid disease is diagnosed. This study seeks to discover the most recent ML-based and data-driven developments and strategies for diagnosing thyroid disease while considering the challenges associated with imbalanced data in thyroid dise

... Show More
View Publication Preview PDF
Scopus (15)
Crossref (13)
Scopus Crossref
Publication Date
Mon Dec 01 2014
Journal Name
Iraqi Journal Of Science
Detection of human leukocyte antigen and celiac disease auto antibodies in serum of patients with multiple sclerosis
...Show More Authors

To determine the important pathogenic role of celiac disease in triggering several autoimmune disease, thirty patients with Multiple Sclerosis of ages (22-55) years have been investigated and compared with 25 healthy individuals. All the studied groups were carried out to measure anti-tissue transglutaminase antibodies IgA IgG by ELISA test, anti-reticulin antibodies IgA and IgG, and anti-endomysial antibodies IgA and IgG by IFAT. There was a significant elevation in the concentration of anti-tissue transglutaminase antibodies IgA and IgG compared to control groups (P≤0.05), there was 4(13.33%) positive results for anti-reticulin antibodies IgA and IgG , 3(10%) positive results for anti-endomysial antibodies IgA and IgG . There were 4 pos

... Show More
Publication Date
Sat Feb 01 2014
Journal Name
World Journal Of Pharmaceutical Sciences
Detection of JAK2V617F Mutation and Serum Levels of Alkaline Phosphatase and Lactate Dehydrogenase in Chronic Myelogenous Leukemia.
...Show More Authors

Chronic myelogenous leukemia (CML) is a myeloproliferative neoplasm arises from Bcr-Abl gene translocation (called Ph chromosome) in hematopoietic stem cells (HSCs). This genetic abnormality results in constitutive activation of tyrosine kinase and subsequent uncontrol growth and multiplication of granulocytes. The cornerstone in treatment of CML are tyrosine kinase inhibitors, of which imatinib is the most effectively used. JAK2V617F mutation is an acquired single nucleotide polymorphism (SNP) occurs in JAK2 gene and is associated with many hematological malignancy other than CML. It was thought that the two genetic abnormalities (Bcr-Abl and JAK2V617F) occur mutually; however, growing body of evidences suggested the reverse. This study a

... Show More
Preview PDF
Publication Date
Mon Dec 01 2014
Journal Name
Iraqi Journal Of Science,
Detection of human leukocyte antigen and celiac disease auto antibodies in serum of patients with multiple sclerosis
...Show More Authors

To determine the important pathogenic role of celiac disease in triggering several autoimmune disease, thirty patients with Multiple Sclerosis of ages (22-55) years have been investigated and compared with 25 healthy individuals. All the studied groups were carried out to measure anti-tissue transglutaminase antibodies IgA IgG by ELISA test, anti-reticulin antibodies IgA and IgG, and anti-endomysial antibodies IgA and IgG by IFAT. There was a significant elevation in the concentration of anti-tissue transglutaminase antibodies IgA and IgG compared to control groups (P≤0.05), there was 4(13.33%) positive results for anti-reticulin antibodies IgA and IgG , 3(10%) positive results for anti-endomysial antibodies IgA and IgG . There were 4 pos

... Show More
Publication Date
Tue Apr 01 2025
Journal Name
Al-rafidain Journal Of Medical Sciences ( Issn 2789-3219 )
Serological and Molecular Detection of Prevalence of Human Parvovirus (B19) in Beta Thalassemia Major Patients in Baghdad
...Show More Authors

Background: Beta thalassemia major (β-TM) is an inheritable condition with many complications, especially in children. The blood-borne viral infection was proposed as a risk factor due to the recurrent blood transfusion regimen (hemotherapy) as human parvovirus B19 (B19V). Objective: This study investigated the B19V seroprevalence, DNA presence, B19V viral load, and B19V genotypes in β-TM patients and blood donors. Methods: This is a cross-sectional study incorporating 180 subjects, segregated into three distinct groups each of 60 patients, namely control, β-TM, and β-TM infected with Hepatitis C Virus (HCV).  For the B19V prevalence in the studied group, the ELISA technique and real-time PCR were used. The genotyping was follo

... Show More
View Publication Preview PDF
Scopus (5)
Crossref (1)
Scopus Crossref