Improving lymph node characterization in staging malignant lymphoma using first-order ADC texture analysis from...

Area of Science:

• Radiology and Imaging Science
• Oncology and Hematology
• Medical Physics

Background:

• Accurate lymph node characterization is crucial for staging and treatment initiation in malignant lymphoma.
• Conventional and diffusion-weighted MRI (DWI) present challenges in nodal assessment, especially for smaller lymph nodes.

Purpose of the Study:

• To compare the diagnostic performance of first-order apparent diffusion coefficient (ADC) texture parameters against mean ADC for characterizing non-Hodgkin lymphoma (NHL) lymph nodes.
• To evaluate the utility of whole-body DWI (WB-DWI) in lymph node assessment using ADC texture analysis.

Main Methods:

• Retrospective analysis of 3T whole-body DWI data from 28 patients with NHL.
• Regions of interest were placed on hyperintense lymph nodes on b1000 images.
• Comparison of mean ADC (ADC_mean) with ADC texture parameters (standard deviation, kurtosis, skewness) for differentiating benign and malignant nodes, with subanalyses based on node volume and region.

Main Results:

• ADC skewness (ADC_skew) and kurtosis (ADC_kurt) significantly differed between benign and malignant nodes (P < 0.001), unlike ADC standard deviation (ADC_stdev; P = 0.21).
• ADC_skew exhibited the highest diagnostic performance: 79% sensitivity, 86% specificity, 83% accuracy, 85% positive predictive value, and 81% negative predictive value.
• ADC_skew demonstrated the highest accuracy across all lymph node volume categories and showed the least interregional variation in accuracy compared to other ADC parameters.

Conclusions:

• First-order ADC texture analysis combined with WB-DWI significantly improves lymph node characterization in NHL compared to mean ADC alone.
• ADC_skew emerges as the most accurate and robust parameter for differentiating benign from malignant lymph nodes, irrespective of node volume or body region.