Background/Objectives: Heart attacks are the leading cause of death in the world. There are two important classes of heart attack: ST-segment Elevation Myocardial Infarction (STEMI) and Non-ST-segment Elevation Myocardial Infarction (NSTEMI) patient groups. While the STEMI group has a higher mortality rate in the short term, the NSTEMI group is considered more dangerous and insidious in the long term. Blocked coronary arteries can be predicted from ECG signals in STEMI patients but not in NSTEMI patients. Therefore, coronary angiography (CAG) is inevitable for these patients. However, in the elderly and some patients with chronic diseases, if there is a single blockage, the CAG procedure poses a risk, so medication may be preferred. In this study, a novel deep learning-based approach is used to automatically detect the occluded main coronary artery or arteries in NSTEMI patients. For this purpose, a new seven-class dataset was created with expert cardiologists. Methods: A new Multi Input-Multi Scale (MI-MS) ConvMixer model was developed for automatic detection. The MI-MS ConvMixer model allows simultaneous training of 12-channel ECG data and highlights different regions of the data at different scales. In addition, the ConMixer structure provides high classification performance without increasing the complexity of the model. Moreover, to maximise the classifier performance, the WSSE algorithm was developed to adjust the classification prediction value according to the feature importance weights. Results: This algorithm improves the SVM classifier performance. The features extracted from this model were classified with the WSSE algorithm, and an accuracy of 88.72% was achieved. Conclusions: This study demonstrates the potential of the MI-MS ConvMixer model in advancing ECG signal classification for diagnosing coronary artery diseases, offering a promising tool for real-time, automated analysis in clinical settings. The findings highlight the model’s ability to achieve high sensitivity, specificity, and precision, which could significantly improve.

Objective: Abdominal involvement of COVID-19 is a current issue. We aimed to evaluate hepatic and pancreatic density alterations on computed tomography (CT) and to analyze whether these alterations had a relationship with chest CT score and laboratory findings. Methods: Patients with reverse transcription-polymerase chain reaction-confirmed COVID-19 from March 11, 2020, to February 6, 2021, were retrospectively analyzed. Patients were divided into nonprogressive and progressive groups according to their chest CT scores. Liver and pancreas density, and liver-to-spleen (L/S) ratio were calculated. Laboratory findings, medication, intensive care unit stay, and survival were noted. Results: There were 51 patients in the nonprogressive group and 123 patients in the progressive group. The median (minimum to maximum) L/S value of the nonprogressive group was 1 (0.28–1.53) at admission and 1.06 (0.33–1.83) at follow-up (P < 0.001). In the progressive group, the median L/S value was 1.08 (0.35–1.51) at admission and 0.92 (0.33–1.75) at follow-up (P < 0.001). A significant difference was found between the 2 groups at admission and follow-up (P = 0.010 and P < 0.001, respectively). Pancreatic density measured at follow-up was significantly lower in the progressive group (P = 0.045). In the progressive group, aspartate aminotransferase, total bilirubin, creatinine, urea, C-reactive protein, D-dimer, and white blood cell values were higher; albumin and lymphocyte values were lower (P < 0.05). Conclusions: Patients with COVID-19 with progressive CT scores may have a decrease in L/S values, and their pancreatic density is lower than nonprogressives. Aspartate aminotransferase, total bilirubin, creatinine, urea, C-reactive protein, D-dimer, and white blood cell values tend to be higher in patients with a high chest CT score.

Background/Objectives: Rib fracture detection holds critical importance in the field of medical image processing. Methods: In this study, two different data augmentation methods, traditional data augmentation (Albumentations) and focused data augmentation (focused augmentation), were compared using computed tomography (CT) images for the detection of rib fractures on YOLOv8n, YOLOv8s, and YOLOv8m models. While the traditional data augmentation method applies general transformations to the entire image, the focused data augmentation method performs specific transformations by targeting only the fracture regions. Results: The model performance was evaluated using the Precision, Recall, mAP@50, and mAP@50–95 metrics. The findings revealed that the focused data augmentation method achieved superior performance in certain metrics. Specifically, analysis on the YOLOv8s model showed that the focused data augmentation method increased the mAP@50 value by 2.18%, reaching 0.9412, and improved the recall value for fracture detection by 5.70%, reaching 0.8766. On the other hand, the traditional data augmentation method achieved better results in overall precision metrics with the YOLOv8m model and provided a slight advantage in the mAP@50 value. Conclusions: The study indicates that focused data augmentation can contribute to achieving more reliable and accurate results in medical imaging applications.

Brain MRI is a medical image obtained by MRI, which stands for "Magnetic Resonance Imaging". Brain MRI uses magnetic fields and radio waves to create detailed images of the brain and surrounding tissues. Today, deep learning algorithms are used to detect brain tumors or classify different brain regions. In this study, feature extraction has been performed with current deep learning models using a dataset consisting of 7023 open access images obtained from patients from various parts of the world, and the results were evaluated by training Support Vector Machine (SVM) and XGBoost models with the extracted features. In this study, 4 deep learning models, VGG16, VGG19, ResNet50 and MobileNetV2, have been used for feature extraction. In order to achieve higher performance, transfer learning method is used in this study, which allows the weights of models that are pre-trained with large data sets to be used in other models. The weights of the models trained with ImageNet were included in the study to improve performance and save time. Although the original layer structures of the models are fixed, the GlobalAveragePooling2D layer has been added to the CNN models to improve performance and generalize the features extracted from deep learning models. Brain MRI images divided into 4 classes as glioma tumor, meningioma tumor, pituitary tumor and no tumor. Auxiliary functions have been used to obtain optimum values for the parameters used for training the models. Accuracy, F1-score, precision and sensitivity metrics used to evaluate the training results. When the results are evaluated, the best performance with an F1-score of 97.87% is obtained by classifying the features extracted from the ResNet50 CNN model with Support Vector Machine (SVM).

Objective: To determine the maternal deaths and the factors affecting them in our city. Methods: The maternal deaths occurred in our city between January 2015 and June 2020 were reviewed retrospectively. The review was conducted by checking “Maternal Death Registry Forms” of the Provincial Directorate of Health. In cases where additional data related with the cause of death were required, the relatives of the cases, associated family practitioner, The Council of Forensic Medicine or local authorities were contacted. The data of the cases including age, gravida, parity, abortion, delivery type, week of gestation during delivery, period of death and maternal deaths due to direct, indirect and incidental causes were recorded. Based on total live births and maternal deaths within 6.5 years, maternal mortality rate was found as the maternal death number per 100,000 live births. Descriptive statistics were used for the statistical analysis of the data. Results: A total of 46.618 live births occurred between 2015 and 2020 in Elazığ. The number of maternal deaths due to direct and indirect causes is 7, and maternal mortality rate was found 15.01/100,000. Hypertensive diseases during pregnancy (n=3, 42.8%), pulmonary embolism (n=1, 14.3%) and cerebral thrombosis (n=1, 14.3%) were among the natural causes of maternal deaths. Indirect cause for maternal death was cardiac diseases (n=2, 28.6%). When they were categorized according to the Three Delays Model, there were 3 death cases in the first delay model and 2 death cases in the third delay model, but there was no maternal death in the second delay model. Conclusion: Maternal death is an significant public health issue which develops due to the generally preventable causes and maintains its importance. The factors contributing to death should be paid attention in order to decrease maternal death rates.