BACKGROUND: In this study, we aimed to compare the success rates of computed tomography image-based artificial intelligence models and magnetic resonance imaging in the diagnosis of preoperative cholesteatoma. METHODS: The files of 75 patients who underwent tympanomastoid surgery with the diagnosis of chronic otitis media between January 2010 and January 2021 in our clinic were reviewed retrospectively. The patients were classified into the chronic otitis group without cholesteatoma (n=34) and the chronic otitis group with cholesteatoma (n=41) according to the presence of cholesteatoma at surgery. A dataset was created from the preoperative computed tomography images of the patients. In this dataset, the success rates of artificial intelligence in the diagnosis of cholesteatoma were determined by using the most frequently used artificial intelligence models in the literature. In addition, preoperative MRI were evaluated and the success rates were compared. RESULTS: Among the artificial intelligence architectures used in the paper, the lowest result was obtained in MobileNetV2 with an accuracy of 83.30%, while the highest result was obtained in DenseNet201 with an accuracy of 90.99%. In our paper, the specificity of preoperative magnetic resonance imaging in the diagnosis of cholesteatoma was 88.23% and the sensitivity was 87.80%. CONCLUSION: In this study, we showed that artificial intelligence can be used with similar reliability to magnetic resonance imaging in the diagnosis of cholesteatoma. This is the first study that, to our knowledge, compares magnetic resonance imaging with artificial intelligence models for the purpose of identifying preoperative cholesteatomas. Cite this article as: Eroğlu O, Eroğlu Y, Yıldırım M, et al. Comparison of computed tomography-based artificial intelligence modeling and magnetic resonance imaging in diagnosis of cholesteatoma. J Int Adv Otol. 2023;19(4):342-349

Data evaluating the presence and characteristics of mesenteric lymph nodes (LNs) in patients with ulcerative colitis (UC) are scarce. The aim of this study is to determine the presence and characteristics of LNs in UC. The LN characteristics in computed tomography (CT), including LN dimension and attenuation, were evaluated retrospectively in 100 patients with UC (61 active and 39 inactive cases). Clinical characteristics and laboratory parameters, including CBC, biochemical analysis, erythrocyte sedimentation rate (ESR), and C reactive protein (CRP) were also compared. Mesenteric LNs were evident in all patients with UC. The attenuation and dimension of mesenteric LNs did not differ between active and inactive patients with UC. No correlation was found among patients with UC in terms of LN dimension, attenuation, ESR, CRP, leucocyte, and albumin (all with p > 0.05). The current study suggested that inflammation results in the development of mesenteric LN in UC, similar to Crohn’s disease and other inflammatory disorders.

Brain tumors are common tumors arising from parenchymal cells in the brain and the membranes that surround the brain. The most common brain tumors are glioma and meningioma. They can be benign or malignant. Treatment modalities such as surgery and radiotherapy are applied in malignant tumors. Tumors may be very small in the early stages and may be missed by showing findings similar to normal brain parenchyma. The correct determination of the localization of the tumor and its neighborhood with the surrounding vital tissues contributes to the determination of the treatment algorithm. In this paper, we aim to determine the classification and localization of gliomas originating from the parenchymal cells of the brain and meningiomas originating from the membranes surrounding the brain in brain magnetic resonance images using artificial intelligence methods. At first, the two classes of meningioma and glioma tumors of interest are selected in a public dataset. Relevant tumors are then labeled with the object labeling tool. The resulting labeled data is passed through the EfficientNet for feature extraction. Then Path Aggregation Network (PANet) is examined to generate the feature pyramid. Finally, object detection is performed using the detection layer of the You Only Look Once (YOLO) algorithm. The performance of the suggested method is shown with precision, recall and mean Average Precision (mAP) performance metrics. The values obtained are 0.885, 1.0, and 0.856, respectively. In the presented study, meningioma, and glioma, are automatically detected. The results demonstrate that using the proposed method will benefit medical people.