Medical SpecialtyOpen Access

Comparison of adenohypophyse hormone levels and texture analysis findings of pituitary MRI images

2022
0 views
0 downloads
Advisor: Doç. Dr. Murat Baykara

Abstract (EN)

It is aimed to compare the texture analysis findings of adenohypophysial hormone levels in different sequences of images obtained by pituitary magnetic resonance imaging (MRI), and to investigate the detectability of adenohypophysial hormones and related hormones in the blood with texture analysis findings. The study was planned retrospectively and 156 cases who underwent pituitary examination with 3T MRI device were evaluated. Pre-contrast T1-weighted (W), T2W, and post-contrast T1W pituitary MRI images were processed with a specially written in MATLAB© program to obtain texture analysis findings. Relationships between adenohypophysis hormones and blood values of related hormones were evaluated statistically by correlation and regression analysis. There were weak correlations between hormone values and many tissue analysis parameters from MRI images. It was found that the detectability of DHEAS, free T4, ACTH and TSH blood values especially, with regression analysis tests was determined by texture analysis findings (38.5%, 45%, 27.2% and 25.7%, respectively). In conclusion, the relationship between adenohypophysis hormones and texture analysis findings obtained from pituitary MRI images and the detectability of hormone values with texture analysis were revealed in this study, which was not sufficient when the literature was reviewed. Texture analysis can be used in the evaluation of the pituitary gland and is a useful test. New studies with a strong infrastructure can contribute to the formation of more knowledge in this field. Key Words: Pituitary Gland; Pituitary Hormones; Magnetic Resonance Imaging; Image Processing, Computer-Assisted; Statistics

Author

Selva Güneş

How to Cite

Selva Güneş (Medical Specialty Thesis). Comparison of adenohypophyse hormone levels and texture analysis findings of pituitary MRI images, 2022, Fırat University.

License

Tüm Hakları Saklıdır

This work is shared under the specified license terms.

More theses from Fırat University