Fractional order calculus model diffusion weighted imaging for evaluating pathological classification and differentiation degree of cervical cancer
Objective To explore the value of fractional order calculus(FROC)model diffusion weighted imaging(DWI)for evaluating pathological classification and differentiation degree of cervical cancer(CCA).Methods Totally 74 CCA patients were enrolled and divided into squamous cell carcinoma(SCC)group(n=54)and adenocarcinoma(ACA)group(n=20)based on pathological classification,also low differentiation group(n=33)and medium-high differentiation group(n=41)based on differentiation degree.Conventional MR examination and DWI with 12 b-values were performed,FROC model parameters(D,β,and p value)and the apparent diffusion coefficient(ADC)of mono-exponential model were obtained via software analysis.The parameters were compared between groups,and receiver operating characteristic curve of those being significantly different between groups were drawn,the area under the curves(AUC)were calculated to evaluate the diagnostic efficacy.Results Significant differences of ADC,D,and β values were found between SCC group and ACA group(all P＜0.05),and D value had the highest AUC(0.726)for distinguishing pathological classification CCA.Meanwhile,significant differences of D,β,p values and ADC were observed between low differentiation group and medium-high differentiation group(all P＜0.05),D value also had the highest AUC(0.865).AUC of the combined model constructed based on significant variables β and p values in logistic regression was 0.926,higher than that of each parameter alone(all P＜0.05).Conclusion FROC model DWI could be used to evaluate pathological classification and differentiation degree of CCA.
uterine cervical neoplasmsmagnetic resonance imagingfractional order calculus
NETL
NSTL
万方数据
