Electrical - Electronic Engineering / Elektrik - Elektronik Mühendisliği
Permanent URI for this collectionhttps://hdl.handle.net/11147/11
Browse
3 results
Search Results
Article Citation - WoS: 14Citation - Scopus: 18Separating Normosmic and Anosmic Patients Based on Entropy Evaluation of Olfactory Event-Related Potentials(Elsevier Ltd., 2019) Güdücü, Çağdaş; Olcay, Bilal Orkan; Schaefer, L.; Aziz, M.; Schriever, V. A.; Özgören, Murat; Hummel, T.Objective: Methods based on electroencephalography (EEG) are used to evaluate brain responses to odors which is challenging due to the relatively low signal-to-noise ratio. This is especially difficult in patients with olfactory loss. In the present study, we aim to establish a method to separate functionally anosmic and normosmic individuals by means of recordings of olfactory event-related potentials (OERP) using an automated tool. Therefore, Shannon entropy was adopted to examine the complexity of the averaged electrophysiological responses. Methods: A total of 102 participants received 60 rose-like odorous stimuli at an inter-stimulus interval of 10 s. Olfactory-related brain activity was investigated within three time-windows of equal length; pre-, during-, and post-stimulus. Results: Based on entropy analysis, patients were correctly diagnosed for anosmia with a 75% success rate. Conclusion: This novel approach can be expected to help clinicians to identify patients with anosmia or patients with early symptoms of neurodegenerative disorders. Significance: There is no automated diagnostic tool for anosmic and normosmic patients using OERP. However, detectability of OERP in patients with functional anosmia has been reported to be in the range of 50%.Article Citation - WoS: 9Citation - Scopus: 13Effect of Faraday Mirror Imperfections in a Fiber Optic Current Sensor Dedicated To Iter(Elsevier Ltd., 2019) Karabulut, Doğuş; Miazin, Anton; Gusarov, Andrei; Moreau, Philippe; Leysen, Willem; Megret, Patrice; Wuilpart, MarcPlasma current measurements in ITER are safety-related and must therefore satisfy a very demanding specification. In this paper, the use of the Fiber Optics Current Sensor (FOCS) operating in the reflection mode with a Faraday mirror to perform plasma current measurements is analyzed. Based on the Jones matrix formalism, we performed numerical simulations to investigate the impact of the Faraday mirror detuning on the measurement accuracy. We show that the use of standard commercial components does not allow to satisfy the ITER requirements for the whole plasma current range. A simple solution to the problem is proposed, which consists in taking into account a mirror calibration in the current estimator. We show that the achievable mirror calibration accuracy is sufficient to fulfill the ITER requirements.Article Citation - WoS: 16Citation - Scopus: 20Automated Labelling of Cancer Textures in Colorectal Histopathology Slides Using Quasi-Supervised Learning(Elsevier Ltd., 2013) Önder, Devrim; Sarıoğlu, Sülen; Karaçalı, BilgeQuasi-supervised learning is a statistical learning algorithm that contrasts two datasets by computing estimate for the posterior probability of each sample in either dataset. This method has not been applied to histopathological images before. The purpose of this study is to evaluate the performance of the method to identify colorectal tissues with or without adenocarcinoma. Light microscopic digital images from histopathological sections were obtained from 30 colorectal radical surgery materials including adenocarcinoma and non-neoplastic regions. The texture features were extracted by using local histograms and co-occurrence matrices. The quasi-supervised learning algorithm operates on two datasets, one containing samples of normal tissues labelled only indirectly, and the other containing an unlabeled collection of samples of both normal and cancer tissues. As such, the algorithm eliminates the need for manually labelled samples of normal and cancer tissues for conventional supervised learning and significantly reduces the expert intervention. Several texture feature vector datasets corresponding to different extraction parameters were tested within the proposed framework. The Independent Component Analysis dimensionality reduction approach was also identified as the one improving the labelling performance evaluated in this series. In this series, the proposed method was applied to the dataset of 22,080 vectors with reduced dimensionality 119 from 132. Regions containing cancer tissue could be identified accurately having false and true positive rates up to 19% and 88% respectively without using manually labelled ground-truth datasets in a quasi-supervised strategy. The resulting labelling performances were compared to that of a conventional powerful supervised classifier using manually labelled ground-truth data. The supervised classifier results were calculated as 3.5% and 95% for the same case. The results in this series in comparison with the benchmark classifier, suggest that quasi-supervised image texture labelling may be a useful method in the analysis and classification of pathological slides but further study is required to improve the results.