The clinical value of applying diffusion-weighted imaging combined with T2-weighted imaging to assess diagnostic performance of muscularis propria invasion in mid-to-high rectal cancer.

Purpose: This study aimed to evaluate the diagnostic performance of combining diffusion-weighted imaging (DWI) with T2-weighted imaging (T2WI) for detecting muscularis propria invasion in rectal cancer.
Methods: We conducted a retrospective analysis of MR images from 76 patients with pathologically confirmed rectal cancer between January 2018 and June 2022. Patients were categorized into invasion and non-invasion groups. the present of muscularis propria invasion. The examination regimen included T2WI, dynamic enhanced scanning and DWI.The apparent diffusion coefficient (ADC) values from DWI were compared between the groups, and the diagnostic performance of combining ADC with T2WI was assessed.
Results: There were differences in ADCmean, ADCmsi, and ADCmin values between the non-invasion group and theinvasion group, and the t values were 3.949, 2.221 and 2.978, respectively. The P values were 0.000, 0.029 and 0.004, respectively. Using an ADC _mean threshold of 1.07 × 10 ^- 3 mm²/s, the sensitivity and specificity for detecting muscularis propria invasion were 82.22% and 61.29%, respectively, with an overall diagnostic accuracy of 89.5%. For ADC _msi , with a threshold of 0.996 × 10 ^- 3 mm²/s, the sensitivity and specificity were 62.75% and 73.33%, respectively, yielding a diagnostic accuracy of 82.9%. The ADC _min , with a threshold of 0.854 × 10 ^- 3 mm²/s, demonstrated sensitivity and specificity of 78.43% and 66.67%, respectively, with a diagnostic accuracy of 85.5%.
Conclusion: T2WI combined ADC value has a good predictive value for judging the degree of musculature invasion of medium-to-high rectal cancer, while ADC mean has a higher comprehensive diagnostic efficiency. Our finding may provide reference for increasing accuracy in mid-to-high rectal cancer diagnosis.
(© 2024. The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.)