Machine learning models were developed utilizing pretreatment magnetic resonance imaging (MRI) radiomics and clinical parameters to predict neoadjuvant chemoradiotherapy response in patients with locally advanced rectal carcinoma. The impact of radiomics dimensionality on predictive performance was assessed. Two models were constructed, revealing moderately advantageous effects of increased dimensionality. These models outperformed the clinical parameter-only model, demonstrating the potential for routine clinical prediction of chemoradiotherapy responders using pretreatment MRI radiomics and clinicopathological features.
Journal Article by Marinkovic M, Stojanovic-Rundic S (…) Radulovic M et 7 al. in J Clin Med