Machine Learning and Statistical Analysis of Biochemical Remission Predictors in Somatotroph Adenoma Resections.

This article discusses the predictors of biochemical remission after surgical resection of somatotroph adenomas, which are tumors that produce excessive growth hormone. The study analyzed data from 97 patients who underwent surgery for somatotroph adenomas and found that larger tumor size, cavernous sinus invasion, and higher levels of preoperative growth hormone and IGF-1 were associated with a higher likelihood of remission failure. A machine learning model identified 3D tumor volume and preoperative IGF-1 as important predictors of remission status. The study concludes that patients who achieve postoperative remission typically have preoperative 3D tumor volumes less than 1.51 cm3. [Extracted from the article]