Whole-tumor histogram analysis of apparent diffusion coefficient for differentiating adenosquamous carcinoma and adenocarcinoma from squamous cell carcinoma in patients with cervical cancer

Background Differentiating adenosquamous carcinoma (ASC) and adenocarcinoma (AC) from squamous cell carcinoma (SCC) precisely is crucial for treatment strategy and prognosis prediction in patients with cervical cancer (CC). Purpose To differentiate ASC and AC from SCC in patients with CC using the apparent diffusion coefficient (ADC) histogram analysis. Material and Methods A total of 118 patients with histologically diagnosed ASC, AC, and SCC were included. The ADC histogram parameters were extracted from ADC maps. Receiver operating characteristic analysis was performed to evaluate the diagnostic performance of each ADC histogram parameter in differentiating the subtypes of CC. The predictors for histologic subtypes were further selected using univariate and multivariate logistic regression analyses. Results The ADC mean , ADC max , ADC P10 , ADC P25 , ADC P75 , ADC P90 , ADC median , and ADC mode of the ASC were significantly lower than those of the AC; and ADC kurtosis and ADC skewness of the ASC were lower than those of the SCC. The ADC mean , ADC max , ADC P10 , ADC P25 , ADC P75 , ADC P90 , ADC median , and ADC mode of AC were significantly higher than those of the SCC. The ADC P10 and ADC P10 + diameter yielded the AUCs of 0.753 and 0.778 in differentiating ASC from AC. The ADC median and ADC median + diameter yielded the AUCs of 0.807 and 0.838 in differentiating AC from SCC. The ADC skewness yielded the AUC of 0.713 in differentiating ASC from SCC. Conclusion The ADC P10 and ADC P10 + diameter, ADC median , and ADC median + diameter performed well in differentiating ASC from AC and AC from SCC, respectively. However, ADC skewness exhibited a limited ability in differentiating ASC from SCC.