Kurtosis Analysis for DWI Improves Prediction of Short-Term Response in Head & Neck Cancer

Jacobus F. A. Jansen¹, Yonggang Lu², Hilda E. Stambuk², Nancy Y. Lee², Jason A. Koutcher², Amita Shukla-Dave²

The study aims to assess the added value of non-Gaussian analysis of diffusion weighted MR imaging (DWI) for the prediction of the short-term response to chemoradiation therapy in patients with advanced head and neck squamous cell carcinoma (HNSCC). Twenty-three patients were included, who all underwent a DWI protocol with 7 b-values at 1.5T. Non-Gaussian fitting was applied using the kurtosis model. DWI parameters derived from standard mono-exponential fitting and kurtosis analysis were evaluated for their efficacy in predicting short-term response using logistic regression analysis. The std(Dapp) was a significant predictor for short-term response, suggesting that non-Gaussian analysis provided a predictive biomarker.