Eroğul, Osman
Loading...
Profile URL
Name Variants
Erogul, Osman
Erogul, O.
Erogul, O.
Job Title
Email Address
erogul@etu.edu.tr
Main Affiliation
02.2. Department of Biomedical Engineering
Status
Current Staff
ORCID ID
Scopus Author ID
Turkish CoHE Profile ID
Google Scholar ID
WoS Researcher ID
Sustainable Development Goals
3
GOOD HEALTH AND WELL-BEING
20
Research Products
10
REDUCED INEQUALITIES
1
Research Products
11
SUSTAINABLE CITIES AND COMMUNITIES
3
Research Products
12
RESPONSIBLE CONSUMPTION AND PRODUCTION
1
Research Products
Documents
95
Citations
1127
h-index
14
WoS data could not be loaded because of an error. Please refresh the page or try again later.
Scholarly Output
123
Articles
30
Views / Downloads
84972/8392
Supervised MSc Theses
20
Supervised PhD Theses
6
WoS Citation Count
147
Scopus Citation Count
219
WoS h-index
6
Scopus h-index
9
Patents
0
Projects
0
WoS Citations per Publication
1.20
Scopus Citations per Publication
1.78
Open Access Source
60
Supervised Theses
26
Google Analytics Visitor Traffic
| Journal | Count |
|---|---|
| Journal of Biotechnology | 13 |
| Diagnostics | 6 |
| 2020 Medical Technologies Congress (Tiptekno) | 5 |
| TIPTEKNO 2023 - Medical Technologies Congress, Proceedings | 4 |
| The EuroBiotech Journal | 3 |
Current Page: 1 / 11
Scopus Quartile Distribution
Competency Cloud
123 results
Scholarly Output Search Results
Now showing 1 - 10 of 123
Article An Early Warning Algorithm To Predict Obstructive Sleep Apnea (osa) Episodes(Avestia Publishing, 2016-11) Özdemir, Galip; Nasıfoğlu, Hüseyin; Eroğul, OsmanSleep apnea is a common respiratory disorder during sleep. It is characterized by shallow or no breathing during sleep for at least 10 seconds. Decrease in sleep quality may effect the next day daily routine unfavorably. In some cases apnea period (not breathing interval) can last more than 30 seconds causing fatal outcomes. 14% of men and 5% of women suffer from Obstructive Sleep Apnea (OSA) in United States. Patients may face apnea for more than 300 times in a single overnight sleep. Polysomnography (PSG) is a multi-parametric recording of biophysiological changes, having Snorring, SpO2, Nasal Airflow EEG, EMG, ECG signals, performed in sleep study laboratories. In this study, a fully automatic apnea detection algorithm is mentinoed and an early warning system is proposed to predict OSA episodes by extracting time-series features of pre-OSA periods and regular respiration using nasal airflow signal. Extracted features are then reduced by RANSAC and entropy based approaches to improve the performance of prediction algorithm. Support vector machines (SVM), one of the commonly used classification algorithms in medical applications, k-Nearest Neighbor and a modified Linear Regression are implemented for learning and classification of nasal airflow signal episodes. The results show that OSA episodes are predicted with 86.9% of accuracy and 91.5% of sensitivity, 30 seconds before patient faces apnea. By the use of predicting an apnea episode before happening, it is possible to prevent patient to face apnea by early warning which can minimize the possible health risks.Article Citation - Scopus: 11Classifying Dysmorphic Syndromes by Using Artificial Neural Network Based Hierarchical Decision Tree(Springer Netherlands, 2018) Özdemir, M.E.; Telatar, Z.; Eroğul, O.; Tunca, Y.Dysmorphic syndromes have different facial malformations. These malformations are significant to an early diagnosis of dysmorphic syndromes and contain distinctive information for face recognition. In this study we define the certain features of each syndrome by considering facial malformations and classify Fragile X, Hurler, Prader Willi, Down, Wolf Hirschhorn syndromes and healthy groups automatically. The reference points are marked on the face images and ratios between the points’ distances are taken into consideration as features. We suggest a neural network based hierarchical decision tree structure in order to classify the syndrome types. We also implement k-nearest neighbor (k-NN) and artificial neural network (ANN) classifiers to compare classification accuracy with our hierarchical decision tree. The classification accuracy is 50, 73 and 86.7% with k-NN, ANN and hierarchical decision tree methods, respectively. Then, the same images are shown to a clinical expert who achieve a recognition rate of 46.7%. We develop an efficient system to recognize different syndrome types automatically in a simple, non-invasive imaging data, which is independent from the patient’s age, sex and race at high accuracy. The promising results indicate that our method can be used for pre-diagnosis of the dysmorphic syndromes by clinical experts. © 2018, Australasian College of Physical Scientists and Engineers in Medicine.Conference Object Citation - WoS: 1Citation - Scopus: 3Classification of Brain Tumors Via Deep Learning Models(IEEE, 2020) Daglı, Kaya; Eroğul, OsmanBrain tumors threathen human health significantly. Misdiagnosis of these tumors decrease effectiveness of decisions for intervention and patient's state of health. The conventional method to differentiate brain tumors is by the inspection of magnetic resonance images by clinicians. Since there are various types of brain tumors and there are many images that clinicians should examine, this method is both prone to human errors and causes excessive time consumption. In this study, the most common brain tumor types; Glioma, Meningioma and Pituitary are classified using deep learning models. While the main objective of this study is to have a high rate of accuracy, the time spent is also examined. The aim of this study is to ease clinicians work load and have a time efficient classification system. The system which has been built has an accuracy up to 90 %.Conference Object Makine Öğrenmesi İle Kalp Hastalıklarının Tespiti(Asos Yayınevi, 2022) Akkur, Erkan; Türk, Fuat; Eroğul, OsmanKalp hastalığı, yaygınlığı ve yüksek ölüm oranları nedeniyle insan sağlığını tehdit etmektedir. Kalp hastalığını tahmin etmek, geleneksel yöntemler kullanarak karmaşık bir iştir. Son yıllarda, kalp hastalıklarını tahmin etmek için makine öğrenimi teknikleri kullanılmaktadır. Bu çalışma kapsamında StatLog Kalp Hastalığı veri seti üzerinde Karar Ağacı, Naive Bayes, Lojistik Regresyon, Destek Vektör Makineleri ve K-En Yakın Komşu makine öğrenme teknikleri kullanılarak kalp hastalıklarının tespitine ilişkin ilişkin karşılaştırmalı bir analiz sunulmaktadır. Veri setindeki etkin öznitelikleri seçmek için Mann Whitney U testi kullanılmıştır. Makine öğrenimi algoritmalarının sınıflandırma performansı, doğruluk, kesinlik, duyarlılık ve F1-skoru açısından değerlendirilmiştir. Destek Vektör Makineleri 96.3% doğruluk, 95.83% kesinlik, 95.83% duyarlılık ve 95.83% F1skoru ile çalışmanın en iyi tahmin oranına sahip algoritması olmuştur. Bu çalışmanın klinisyenlere kalp hastalığını erken evrede tespit etmede yardımcı olacağına inanmaktayız.Conference Object Citation - WoS: 2Brain Tumor Detection and Volume Estimation Via Mr Imaging(Elsevier Science Bv, 2015) Öğretmenoğlu, Cansel; Eroğul, Osman; Telatar, Ziya; Güler, Emine Rumeysa; Yıldırım, Fahri[No abstract available]Conference Object Citation - WoS: 3Citation - Scopus: 4Investigating Ballistic Gelatin Based Phantom Properties for Ultrasound Training(Springer Nature Singapore Pte Ltd., 2019) Özdemir, Mertcan; Özdemir, Galip; Eroğul, OsmanThe simulation has become an important tool for healthcare practitioners who have difficulty in accessing patients to learn ultrasound imaging modes. The ultrasound phantoms are specially designed objects that are used or imagined to evaluate, analyze and adjust the performance of test devices. These phantoms for ultrasonography devices are expensive, and low-cost alternatives have provided an educational experience that does not give the best result. Ballistic gelatin is a member of the 250-Bloom hydrogel family that resembles human muscle tissue in terms of its mechanical properties. The 250-Bloom Ballistic Gelatin (BG) is prepared with different mixing ratios to be made the mechanical tests such as gunshot, compression and electrical conductivity measurement. The results are compared with the mechanical results of human muscle tissue in order to measure the similarity of the 250-Bloom BG we prepared to human muscle tissue. It is showed that the 250-Bloom BG phantom model has very close mechanical properties to human muscle tissue at time-dependent characteristics of mechanical test results. It is also measured how long it can last without degradation with the time required to use it in the simulation and it is coated with the thermal insulation material needed to extend the degradation period. Based on these results, 250-Bloom BG phantom is recommended as a model for the creation of phantom limb model. Consequently, this model is a much more affordable alternative and easy to produce, it facilitates to work with any organ model in ultrasound imaging for healthcare practitioners.Conference Object Sıkıştırılmış Algılama Tabanlı Sayısal Almaç: İlk Performans Ölçümleri(IEEE, 2019-04) Korucu, Ali Buğra; Çakar, Onur; Alp, Yaşar Kemal; Gök, Gökhan; Arıkan, Orhan; Eroğul, OsmanIn this work, sensitivity, instantaneous dynamic range and bandwidth measurement results of the Compressive Digital Receiver (CDR) implemented on a real hardware, will be reported for the first time. Developed CDR is a compressive sensing based sub-Nyquist sampling receiver which can monitor 2.25GHz bandwith instantaneously by using four ADC's each of which is sampling at 250MHz and has 14 bits resolution. It is observed that the sensitivity and dynamic range of the CDR changes with respect to frequency. For 2.25GHz bandwidth, the best and worst sensitivity values of the CDR are reported as -62dBm and -41dBm, respectively. Single-signal dynamic range of CDR is measured as at least 60dB for the whole band. The best and worst values of the two-signal dynamic rage values are observed as 45dB and 20dB, respectively.Conference Object Citation - Scopus: 5Kamu Özel Ortaklığı Hastaneleri İçin Tıbbi Atık Yönetim Modeli(IEEE, 2015) Koçak, Onur; Kurtuldu, Hüseyin; Akpek, Ali; Koçoğlu, Arif; Eroğul, OsmanToday, with developing technologies and expanding health care system, medical waste has reached a fairly large volume. Particularly, the extensive use of disposable medical devices and supplies are among the factors that increase the production of medical waste. Monitoring the processes involving the separation, temporary storage, disposal, and transfer of medical waste is critical in terms of the environment and human health. In this study, the implementation of medical waste collection, separation and classification processes were surveyed in new city hospitals constructed with public-private partnership. The standards for temporarily holding wastes were also discussed. Furthermore, the cost analysis required for the handling and disposal of medical waste was provided. By means of studying the medical waste disposal methods, few suggestions regarding the most appropriate methods and models of offset technology investments for the city hospitals were proposed.Conference Object Citation - Scopus: 1Multi-Regional Adaptive Image Compression (aic) for Hip Fractures in Pelvis Radiography(Springer-Verlag Singapore Pte Ltd, 2017) Nasıfoğlu, Hüseyin; Eroğul, Osman; Ataç, Gökçe Kaan; Özdemir, GalipHigh resolution digital medical images are stored in DICOM (Digital Imaging and Communications in Medicine) format that requires high storage space in database. Therefore reducing the image size while maintaining diagnostic quality can increase the memory usage efficiency in PACS. In this study, diagnostic regions of interest (ROI) of pelvis radiographs marked by the radiologist are segmented and adaptively compressed by using image processing algorithms There are three ROIs marked by red, blue and green in every image. ROI contoured by red is defined as the most significant region in the image and compressed by lossless JPEG algorithm. Blue and green regions have less importance than the red region but still contain diagnostic data compared to the rest of the image. Therefore, these regions are compressed by lossy JPEG algorithm with higher quality factor than rest of the image. Non-contoured region is compressed by low quality factor which does not have any diagnostic information about the patient. Several compression ratios are used to determine sufficient quality and appropriate compression level. Compression ratio (CR), peak signal to noise ratio (PSNR), bits per pixel (BPP) and signal to noise ratio (SNR) values are calculated for objective evaluation of image quality. Experimental results show that original images can approximately be compressed six times without losing any diagnostic data. In pelvis radiographs marking multiple regions of interest and adaptive compression of more than one ROI is a new approach. It is believed that this method will improve database management efficiency of PACS while preserving diagnostic image content.Conference Object Determination of the Optimal Eeg-Based Features To Detect Adhd by Machine Learning Algorithms(Institute of Electrical and Electronics Engineers Inc., 2023) Can, Y.; Bigat, I.; Nassehi, F.; Eken, Aykut; Erogul, O.This study proposes a highly accurate and fast algorithm for the diagnosis of attention deficit hyperactivity disorder (ADHD), which will reduce reliance on time-consuming subjective assessments, the findings of which are likely to be mistaken with other neurodevelopmental diseases. Time, frequency and nonlinear features were extracted from electroencephalographic (EEG) signals which recording based on visual attention task obtained from 61 ADHD and 60 healthy participants. In this study, Least Absolute Shrinkage and Selection Operator (LASSO) was used to find reliable features; and four machine learning classifiers such as support vector machine (SVM), k-nearest neighbors (KNN), decision tree and ensemble learning were evaluated for classifying ADHD and healthy children. The results were indicated that using LASSO with SVM can be useful for classifying ADHD and the highest average accuracy was reached in this study was 96.3%. In addition, the features selected with LASSO had shown that signals from the temporal, parietal, and occipital lobes might have the possible biomarkers for ADHD, at least in tasks that require visual attention. © 2023 IEEE.
