Biyomedikal Mühendisliği Bölümü Bildiri & Sunum Koleksiyonu
Biyomedikal Mühendisliği Bölümüne ait bildiri ve sunumlar bu koleksiyonda listelenir.
https://hdl.handle.net/20.500.12294/429
2024-03-28T18:39:55Z
2024-03-28T18:39:55Z
Development of a new algorithm for the estimation of the step percentage in compound muscle action potential scan
Goker, Imran
Baslo, Mehmet Baris
Oge, Ali Emre
https://hdl.handle.net/20.500.12294/3263
2023-02-08T15:32:19Z
2019-01-01T00:00:00Z
Development of a new algorithm for the estimation of the step percentage in compound muscle action potential scan
Goker, Imran; Baslo, Mehmet Baris; Oge, Ali Emre
Compound Muscle Action Potential (CMAP) Scan used in the diagnosis and in the monitoring of Neuromuscular Diseases is based on recording the responses obtained from the gradual stimulation of a peripheral nerve. Clearly visible differences between CMAPs are referred as step. The maximum CMAP percentage of all detected steps are defined as Step Percentage and this is one of the parameters providing evidence for Motor Unit (MU) loss and for reinnervation. The aim of this study is to develop an algorithm which enables to estimate the step percentage more quantitatively. Motor neuron groups which were created through a simulator software were stimulated gradually. The step sizes were computed by utilizing the obtained responses as CMAPs in computing step sizes. Then, these step sizes were sorted in descending order. The step percentage was estimated by taking the cumulative sum of the values greater than two standard deviations (SD) of these sorted step sizes by running the aforementioned algorithm. It was seen that there are larger in fluctuations Step% values in motor neuron groups with lower number of axons, however, a more regular decreasing trend with smaller fluctuations was observed in higher number of axons. In the future studies, the inclusion of further motor neurons with greater axons numbers will be established. Moreover, the application of the algorithm for the patient data will be also considered. © 2019 IEEE.
2019-01-01T00:00:00Z
Comparison of machine learning techniques on MS lesion segmentation
Dogan, Ahsen Feyza
Duru, Dilek Goksel
https://hdl.handle.net/20.500.12294/3262
2023-02-08T15:32:13Z
2019-01-01T00:00:00Z
Comparison of machine learning techniques on MS lesion segmentation
Dogan, Ahsen Feyza; Duru, Dilek Goksel
Multiple sclerosis arises with conformational change in myelin sheath. Magnetic resonance imaging is frequently used in detection of MS. In this study, to figure out MS lesion, machine learning techniques, namely k means and support vector machine are used. K means is an unsupervised technique used to cluster data into k groups. Support vector machine is a supervised machine learning technique used as classifier. Since dataset does not contain label of images, labels are generated by pixel values adopted from original MR image. Classification results were achieved as 70.24% and 91.04% for k means and SVM respectively. According to the promising results, future research will focus on the automatization of this segmentation process via deep learning leading to medical decision support system. © 2019 IEEE.
2019-01-01T00:00:00Z
Dynamic time warping based connectivity classification of event-related potentials
Al-Rubaye, Kadhum Kareem
Bayat, Oǧuz
Ucan, Osman Nuri
Duru, Dilek Göksel
Duru, Adil Deniz
https://hdl.handle.net/20.500.12294/3259
2023-02-08T15:32:16Z
2019-01-01T00:00:00Z
Dynamic time warping based connectivity classification of event-related potentials
Al-Rubaye, Kadhum Kareem; Bayat, Oǧuz; Ucan, Osman Nuri; Duru, Dilek Göksel; Duru, Adil Deniz
Human brain electrical responses measured as Electroencephalogram epochs have different characteristics by means of amplitude and frequency content depending on the conditions and stimuli. Event-related potentials are the responses given to the stimuli and can be measured using the EEG. The average of these epochs are computed to remove the background activity and helps to exhibit the response to stimuli solely. In the concept of this study, dynamic time warping based connectivity features are used to classify the single-trial ERP epochs. Color Stroop test was implemented and ERP data are collected from 10 subjects. Support vector machine and K-NN classifiers are used and accurate classification results are achieved with the use of DTW metrics. © 2019 IEEE.
2019-01-01T00:00:00Z
Synthesis and characterization of molecularly imprinted polymer from renewable sources
Yucel, Necla
Koten, Hasan
Hatir, Pinar Cakir
https://hdl.handle.net/20.500.12294/3235
2023-02-08T15:32:09Z
2021-01-01T00:00:00Z
Synthesis and characterization of molecularly imprinted polymer from renewable sources
Yucel, Necla; Koten, Hasan; Hatir, Pinar Cakir
In this study, molecular imprinted polymers targeting the potent antibiotic ciprofloxacin were prepared for the first time using renewable resources. Lactic acid maleic acid ester, methylene bisacrylamide, potassium persulfate and N, N, N', N' tetramethylenediamine were used as functional monomer, crosslinker and initiator, respectively. The chemical structure of the synthesized polymers was characterized by Fourier Transform Infrared spectroscopy, and the thermal behavior was characterized by Thermogravimetry. Developed molecularly imprinted polymers can be used in separation, catalysis, and sensor applications, and as drug delivery systems. © 2021 IEEE.
2021-01-01T00:00:00Z