Application value of histogram analysis from ultra-high b-value diffusion-weighted imaging in assessing the tumor microenvironment of rectal cancer
LI Wanqing
HAO Yongfei
ZHAO Wanting
QIAO Xiao'ai
WANG Hong
ZHANG Guangwen
ZHANG Jinsong
Abstract：Objective To explore the value of functional parameter histogram features derived from ultra-high b-value diffusion-weighted imaging(UHBV-DWI)in assessing tumor microenvironment(TME)characteristics in stageⅡ-Ⅲ rectal cancer.Methods A retrospective analysis was conducted on 54 patients with pathologically confirmed rectal cancer between January and December 2024,who had postoperative HE-stained sections and UHBV-DWI images(b=0,1 700-3 500 s/mm2).Two pathologists evaluated TME features,including tumor stroma percentage(TSP)and tumor inflammatory cell infiltration(KM score),with inter-observer consistency assessed.UHBV-DWI data were processed using monoexponential and diffusion kurtosis models to generate parameter maps of ultra-high apparent diffusion coefficient(ADCuh),ultra-high mean diffusivity(MDuh),and ultra-high mean kurtosis(MKuh).Histogram features were extracted and analyzed for correlation with TSP and KM scores using independent samples t-test or Wilcoxon rank-sum test.The diagnostic performance of these features in discriminating TSP and KM score groups was evaluated using receiver operating characteristic(ROC)curve analysis.Results Multiple histogram features of ADCuh and MDuh showed statistically significant differences between TSP groups(P<0.05),including maximum value,variance,and entropy.In KM score groups,significant differences were observed in ADCuh mean,75th and 90th percentiles,as well as MDuh maximum,mean,25th,75th,and 90th percentiles(P<0.05).No significant differences were found in MKuh histogram features between either group comparison.ROC analysis demonstrated that MDuh maximum provided good discrimination for TSP grouping(AUC=0.717),while MDuh 75th percentile showed the highest diagnostic efficacy for KM score grouping(AUC=0.723).Conclusion The histogram features of ADCuh and MDuh derived from UHBV-DWI may serve as non-invasive imaging biomarkers for evaluating TME characteristics in stage Ⅱ-Ⅲ rectal cancer.
Keywords：rectal cancertumor microenvironmentmagnetic resonance imagingultra-high b-value diffusion-weighted imagingtumor stroma percentagetumor infiltrating cellular leukocytessingle exponential modeldiffusion kurtosis imaging
Publication Date：2026-04-30
Online Publishing Date：2026-08-26(First online date of this platform, not the publication date of the document)
Pages：8( 581-588 )