Publications by authors named "I H Dundar"
Congenital adrenal hyperplasia (CAH) is an autosomal recessive disease caused by the deficiency of one of the enzymes involved in cortisol synthesis. More than 95% of the cases occur as a result of defects in the gene encoding 21-hydroxylase (). 21-hydroxylase deficiency has been divided into classical and non-classical forms.
View Article and Find Full Text PDFCongenital adrenal hyperplasia (CAH) is an autosomal recessive disease caused by the deficiency of one of the enzymes involved in cortisol synthesis. Between 90% and 99% of cases of CAH are caused by 21-hydroxylase deficiency (21-OHD) caused by mutations in . Although 21-OHD has been historically divided into classical and non-classical forms, it is now thought to show a continuous phenotype.
View Article and Find Full Text PDFObjective: This study aimed to evaluate the impact of continuous glucose monitoring (CGM) assistance on glycemic control in children with type 1 diabetes (T1D) in earthquake-affected regions, comparing those who benefited from CGM with those who did not. Additionally, the study assessed changes in CGM metrics over nine months of CGM use.
Methods: A multicenter, cross-sectional study was conducted across 11 centers in Türkiye.
Objective: This study aimed to determine the response to growth hormone (GH) therapy of patients with growth hormone deficiency (GHD), idiopathic short stature (ISS), and small for gestational age (SGA).
Materials And Methods: The data of the 202 children who received GH treatment were analyzed retrospectively. Laboratory parameters, puberty stages, annual growth rates, body mass index (BMI), GH side effects, target height (TH), estimated adult height (EAH), and bone age (BA) were obtained during the GH treatment period.
Deep learning in distinguishing pulmonary nodules as benign and malignant.
Turk Gogus Kalp Damar Cerrahisi Derg
July 2024
Article Synopsis
- - The study aimed to improve early lung cancer diagnosis by developing convolutional neural network models to accurately differentiate between benign and malignant lung nodules using diagnostic imaging data from 2015 to 2020.
- - Patients were categorized into three groups—benign, malignant, and control—while deep neural networks were trained on 80% of their data and tested on the remaining 20%, showing an accuracy of up to 80% with AlexNET and 93.5% when using a support vector machine classifier with the VGG19 model.
- - The findings indicate that utilizing deep learning and feature extraction effectively enhances the ability to identify lung nodules, supporting early diagnosis and potentially improving radiology practices.