Mühendislik Fakültesi / Faculty of Engineering
Permanent URI for this collectionhttps://hdl.handle.net/11727/1401
Browse
12 results
Search Results
Item Automatic Glacoma Detection Using Whale Optimization and Support Vector Machines(2022) Ozen, Sinasi Kutay; Aksahin, Mehmet FeyziGlaucoma is among the most common causes of permanent blindness in humans. The mass screening will aid in early diagnosis in a large population, as the initial symptoms are not obvious. This type of mass screening requires an automated diagnostic technique. Our proposed automation extracts feature by obtaining disk-to-cup ratio by applying histogram equalization, median filter, otsu thresholding, and whale optimization algorithm, respectively, on the optic disc region obtained by preprocessing. In addition, the optic disc circumference, optic disc area, optic cup circumference, and optic cup area values obtained from the optic disc region are given to the support vector machine model together with the cup-disc ratio, and glaucoma detection is made automatically. The proposed system has been validated on a real ophthalmological images of both normal and glaucoma cases. The results show the effectiveness of the proposed method when compared with other existing systems.Item Automatic Brain Tissue Segmentation on TOF MRA Image(2020) Ozen, Sinasi Kutay; Aksahin, Mehmet FeyziFor the segmentation of brain vessels from MRA images, brain tissue is used in the head, eye, skull, etc. must be separated from the structures. For this reason, studies are carried out for the segmentation of brain tissue. In this study, the method that automatically segregates brain tissue from magnetic resonance angiography images taken with time of flight (TOF) technique is presented. The method in the study consists of live steps. First of all, the tip contrast values in the image are filtered by anisotropic diffusion filtering method. Parameters of anisotropic diffusion method are determined automatically by the natural image quality evaluator method. Sudden density transitions arc detected by applying LoG edge detection filter on the filtered image. It is made ready for image analysis by applying etching on the image with density transitions. According to the conditions determined in image analysis, brain tissue is obtained separated from other head structures. As a result of this study, an easy-to-apply, fast-delivering, high-accuracy automatic algorithm has been introduced.Item Idfatification Using FCG Signals(2020) Kilicer, Elif Cansu; Ay, Sevval; Aksahin, Mehmet FeyziSystems that determine identity with individual features are called biometric systems. Today, voice, fingerprint, retina/iris, and facial recognition systems are some of the biometric identification methods. These methods have become replicable with the advancement of technology. Accordingly, Electrocardiogram (ECG) signals are universal, unique, easy to measure, and can only be obtained from living people. For this reason, it can be accepted that ECG is an effective method that can he used to prevent counterfeiting among biometric identification methods. In this study, an algorithm that can make identification via ECG is proposed. Within the scope of the study, the time and time-frequency domain analyzes of the ECG signals obtained from the PhsiyoNet database are performed then various features are determined. These determined features were classified using machine learning methods. The performance of the developed algorithm has been calculated as 100% accuracy, 1000/0 specificity, and 100% sensitivity.Item Detection of Hypopnea Using Respiratory Signals(2019) Oktan, Aynur Didem; Aksahin, Mehmet FeyziHypopnea is a respiratory disorder that affects people's sleep quality and reduces their standard of living. Detection and treatment of sleep disorders are costly. It requires time and effort. Because patients have to spend their time with special systems in which their physiological signals are recorded and specialist personnel in their sleep laboratories. Polysomnograms should be analyzed by medical doctors every night. Reliable sleep stage scoring is done manually by experts. This means that each morning, a specialist visually analyzes the 960 period of an eight-hour polysomnogram to create a hypnogram. This requires a long time. In this study, a method for automatic detection of hypopnea by eliminating the effect of the doctor is proposed. In this method, epoxes were scored by using air flow, thorax and abdominal amplitude information obtained from the person. A training data was created using hypopnea and normal epochs and grading was performed using the determined attributes. Quadratic Support Vector Machines (SVM) gave the highest accuracy when determining the presence of hypopnea. The linear DVM method was trained in 90.6% accuracy and the system was then tested. It was found that hioped epochs can be detected with 90% sensitivity.Item Detection of Multiple Sclerosis Disease by EEG Coherence Analysis(2019) Karaca, Busra Kubra; Aksahin, Mehmet Feyzi; Ocal, Ruhsen; V-3553-2017Multiple sclerosis (MS) is a chronic and inflammatory disease affecting the brain and spinal cord. Although the exact cause of MS is not known, genetic, environmental and immunological factors are involved in the etiology of the disease. The lack of a single diagnostic test for early diagnosis of MS and the similarity of clinical features in MS to other diseases is a serious problem. Early detection of MS is important, and therefore a rapid and reliable pre-diagnosis of MS is important for the treatment and prognosis of the disease. Electroencephalography (EEG) signals provide important information about brain and nerve diseases. Therefore, in the proposed study, a decision support system has been developed which will contribute to the pre-diagnosis by using EEG signals. In this context, coherence analysis of bipolar channel pairs of EEG signals obtained from MS patients and healthy individuals was performed and feature extraction was performed from certain frequency bands. Using the obtained features, the "Subspace Discriminant" classifier was trained with 95.8% accuracy and then the system was tested. As a result, accuracy, sensitivity and specificity rates were 91.67%, 85.71% and 100%, respectively.Item Sleep Apnea Detection Using Blood Pressure Signal(2018) Aksahin, Mehmet Feyzi; Oltu, Burcu; Karaca, Busra KubraSleep apnea is a common respiratory disease. Apnea affects sleep quality, reduces people's life standards, and it can result in death at advanced stage. Therefore the ability to detect the apnea quickly and accurately is important for the treatment of this disease. Apnea is diagnosed by specialists however this is a long and exhausting process. Accordingly, a decision support system that automatically diagnoses apnea has been developed to facilitate this process and make it more objective. The developed decision support system in this study is based on patient's blood pressure signals instead of traditional Polysomnography (PSG) records, which requires various physiological signals measured from the patients. In the examined blood pressure signals, the change that results from each heart beat was determined and heart rate variability (HRV) was calculated based on these changes. At the same time, maximum and minimum amplitude values were found for each change period and amplitude variability vector was created. The features for each epoch were determined using the generated amplitude variability vector and HRV data. Presence of apnea in each epoch is classified with determined features and with the use of "Quadratic SVM" classifier. The Quadratic SVM classifier was trained with 87.5% accuracy and then the system is tested. As a result 75.4% sensitivity and 75% positive predictive values were obtained.Item Classification of Heart Sound Recordings With Continuous Wavelet Transform Based Algorithm(2018) Karaca, Busra Kubra; Oltu, Burcu; Kantar, Tugce; Kilic, Erkin; Aksahin, Mehmet Feyzi; Erdamar, AykutCardiovascular diseases are the major cause of death in the world. Early diagnosis of heart diseases provide an effective treatment. Heart diseases can be diagnosed using data obtained from heart sounds. Heart sounds are listened by a physician with auscultation method and the disease diagnosis can vary depending on the physician's experience and hearing ability. For this reason, automatic detection of anomalies in heart sounds can give more objective results. In this study, features were obtained by processing phonocardiogram signals taken from Physionet database. The heart sounds are classified as normal and abnormal using these features and the k - nearest neighbor method. As a result, sensitivity, specificity and accuracy were determined as 100%, 96.1% and 98.2%, respectively.Item Neurofeedback System Design Based On Discrete Wavelet Transform(2018) Ozen, Sinasi Kutay; Aksahin, Mehmet FeyziNeurofeedback is a learning strategy that helps to alter brain waves called electroencephalography (EEG). Neurofeedback technique is used in the treatment of various diseases such as epilepsy, sleep disorders, attention deficit and learning disability. In the first stage of this study, a device was developed in which the EEG signals are collected and transferred to the computer. In the second stage of the study, the obtained EEG signals were analyzed by the discrete wavelet transform method, and according to the analysis result, a system was designed by which the patients could be treated using visual feedback. The usability of the visual feedback part based on theta/ beta protocol with EEG subbands has been tested in real time.Item Detection of Epilepsy Disease From EEG Signals With Artificial Neural Networks(2016) Ozkan, Cansu; Dogan, Seda; Kantar, Tugce; Aksahin, Mehmet Feyzi; Erdamar, Aykut; 0000-0001-8588-480X; AAA-6844-2019The diagnosis of the epilepsy diseases are made by physicians with analyzing the electroencephalography (EEG) records. The epilepsy diseases can be determined with observing the main properties of before and on-time seizure signals in time and frequency domain. Physicians are evaluating the results after some necessary scoring on EEG records. However, this evaluation is specialistic, time consuming processes and also may subjective results. At this point, to allow detection of epilepsy diseases, a decision support system can give more objective results to the physicians for diagnosing. The subject of the study is automatically diagnosing the epilepsy diseases on EEG signals. In the proposed study, analyses of EEG signals in time and frequency domain were done and features of diseases were obtained. As a result, using artificial neural network (ANN) and obtained features, a decision support system is realized to diagnose the epilepsy. The specificity and the sensitivity of the algorithm are 94% and 66% respectively.Item Sleep Apnea Detection Using With EEG, ECG and Respiratory Signals(2017) Aksahin, Mehmet Feyzi; Erdamar, Aykut; Isik, Atakan; Karaduman, Asena; 0000-0001-8588-480X; AAA-6844-2019Sleep apnea is one of the major sleep disorders of today. The diagnosis of sleep apnea is performed by specialist physicians. This siruation extends the duration of the diagnosis. To shorten this period and at the same time to avoid the mistakes that may occur in diagnosis, an automated decision support system has been considered in the diagnosis and classification of sleep apnea. In this study, ECG signal was analyzed to obtain Heart Rate Variability (HRV) signal and the power spectral density (PSD) of this signal was examined. It has been observed that the low and high frequency energy ratios are different in the PSD of examined HRV. Parallel to this analysis, the energy of the respiratory signal is obtained and it is understood that there is a significant energy exchange in the apnea cases. However, the powers of the frequency bands in the EEG signal were found separately and the ratios of these bands to each other were calculated. In the analysis, it was observed that the ratios of and non-apnea periods. By using these differences, an artificial neural network (ANN) algorithm is constructed to diagnose and classify the sleep apnea. This algorithm was tested on two patient data; ANN was trained and tested separately for each patient. As a result, it was determined that the average accuracy rate of ANN is high.