No abstract

Glaucoma is a disease that causes an abnormal increase in intraocular pressure and therefore causes permanent damage to the optic nerves. Early and accurate diagnosis of the disease, known as the most "insidious" disease among eye diseases, is important. In this study, glaucoma prediction application was performed from high-resolution fundus photographs taken from an open-source database. Correlation, energy, homogeneity, contrast and entropy features were extracted from the segmented photographs using the gray-level co-occurrence matrix. Extracted features were divided into 66% test and 33% training after taking their average values. A 3-fold cross-validation was applied to the data and a feedback artificial neural network, classification and regression trees algorithm and k nearest neighbor algorithm were trained using 66% of the data. Classification success was also tested with 33% of test data. As a result, glaucoma and healthy individuals were classified with an average of 86.7% accuracy with the k nearest neighbor algorithm, an average of 87.8% accuracy with the decision trees, and an average of 96.7% accuracy with the artificial neural network algorithm. According to the results obtained, it was seen that glaucoma disease could be detected with high accuracy with the gray-level co-occurrence matrix features of glaucoma disease.

Determining the blood glucose level is important for the prevention and treatment of diabetes mellitus. We developed a sensor system using Quartz Crystal Microbalance (QCM) to determine the blood glucose level from human blood serum. This study consists of two experimental stages: artificial glucose/pure water solution tests and human blood serum tests. In the first stage of the study, the QCM sensor with the highest performance was identified using artificial glucose solution concentrations. In the second stage of the study, human blood serum measurements were performed using QCM to determine blood glucose levels. QCM sensors were coated with phthalocyanines (Pcs) by jet spray method. The blood glucose values of 96 volunteers, which ranged from 71 mg/dL to 329 mg/dL, were recorded. As a result of the study, human glucose values were determined with an average error of 3.25%.

Alzheimer's disease is a neurodegenerative disorder that causes loss of cognitive function and cognitive decline in individuals. Detection of the disease at an early stage is important to slow down the devastating effects of the disease. The use of an autonomous computerized support system that can assist specialist physicians in the diagnostic process saves time and helps reduce human error. For this reason, a high-accuracy classification study was aimed at utilizing different machine learning algorithms for early diagnosis of Alzheimer's disease. Within the scope of this study, an open source data set created with Electroencephalogram (EEG) signals from 24 healthy and 24 Alzheimer's patient volunteers was used. 28 features, including spectral and statistical features, were extracted from each channel of the EEG signals. The extracted features were evaluated to the feature importance algorithm and the five most significant features that could distinguish between Alzheimer's individuals and healthy individuals were determined. Four machine learning algorithms are trained with the determined features. 70% of the data was used for training and the algorithms were trained with a 10-fold cross-validation method. When the four machine learning algorithms were tested with the data reserved for testing, which the algorithms had not seen before, the highest accuracy was obtained with the Gradient Boosting Classifier (GBC) algorithm with 96.43%.

No abstract