The value of whole - tumor ADC histogram parameters combined with imaging biomarkers in predicting PNI/LVI in rectal adenocarcinoma

Motivation: The study investigates the value of combining whole-tumor ADC histogram parameters and mrEMVI in predicting perineural invasion (PNI) and lymphovascular invasion (LVI) in rectal adenocarcinoma. Goal(s): PNI/LVI prediction using ADC and mrEMVI shows improved accuracy. Approach: A retrospective analysis of 102 patients was conducted, using FireVoxel software for ROI delineation and histogram analysis. ADC parameters and mrEMVI were combined in a multivariate logistic regression model. Results: Significant differences were found between the PNI/LVI-positive and negative groups in ADC mean, standard deviation, ADC-1%, ADC-75%, ADC-95%, ADC-99%, and mrEMVI. The combined model had an AUC of 0.918, surpassing the histogram model and individual parameters. Impact: Combining whole-tumor ADC histogram parameters with mrEMVI can significantly improve the accuracy of preoperative PNI/LVI prediction in rectal adenocarcinoma. This approach has the potential to become an important tool in clinical diagnosis, advancing the application of imaging in tumor detection.