The paper introduces N4ITK, an improved version of the popular nonparametric nonuniform intensity normalization (N3) algorithm for bias field correction in medical imaging. The authors propose a fast and robust B-spline approximation routine and a modified hierarchical optimization scheme to enhance the performance of the original N3 algorithm. The new method, denoted as "N4ITK," is made publicly available through the Insight Toolkit of the National Institutes of Health. The performance of N4ITK is evaluated using simulated data from the BrainWeb database, hyperpolarized 3 He lung image data, and 9.4T postmortem hippocampus data. The results show that N4ITK consistently outperforms the original N3 algorithm, particularly in noisy images and at higher field strengths. The paper also discusses the advantages of the new B-spline approximation algorithm, including its robustness to outliers and the ability to handle data with confidence values. The experimental comparisons demonstrate the superior performance of N4ITK in various scenarios, highlighting its potential for practical applications in medical image analysis.The paper introduces N4ITK, an improved version of the popular nonparametric nonuniform intensity normalization (N3) algorithm for bias field correction in medical imaging. The authors propose a fast and robust B-spline approximation routine and a modified hierarchical optimization scheme to enhance the performance of the original N3 algorithm. The new method, denoted as "N4ITK," is made publicly available through the Insight Toolkit of the National Institutes of Health. The performance of N4ITK is evaluated using simulated data from the BrainWeb database, hyperpolarized 3 He lung image data, and 9.4T postmortem hippocampus data. The results show that N4ITK consistently outperforms the original N3 algorithm, particularly in noisy images and at higher field strengths. The paper also discusses the advantages of the new B-spline approximation algorithm, including its robustness to outliers and the ability to handle data with confidence values. The experimental comparisons demonstrate the superior performance of N4ITK in various scenarios, highlighting its potential for practical applications in medical image analysis.