3508

The optimal b value study of endogenous water molecule diffusion imaging in the diagnosis of brain tumors
Fen Xing1 and Guangyao Wu2

1RenMin hospital of Wuhan University, Wuhan, China, 2Zhongnan Hospital of Wuhan University, Wuhan, China

Synopsis

In this study, thirty-six cases of meningioma were divided into benign and malignant meningioma. ADC values fitted by 11 b values and any pairs of two b values were used to distinguish between benign and malignant meningioma. We found that, the ADC values of benign, malignant, and all meningiomas showed a decreasing trend with the increase of b value, and the smaller the b value, the greater the decrease. when the b value was ≥ 200 s/mm2, any combinations of b values could discriminate between benign and malignant meningioma. The ADC values obtained by fitting b = 700 and b = 1000 s/mm2 had the highest AUC value.

Purpose

To evaluate the differential diagnostic value of ADC values fitted with different b-value combinations in benign and malignant meningiomas, and to investigate the optimal b value of DWI imaging in differential diagnosis of meningiomas.

Methods

Thirty-six cases of meningioma were eventually included in the experiment. According to the pathological results, they were divided into two groups: benign and malignant meningioma. ADC values fitted by 11 b values (0, 50, 100, 150, 200, 300, 400, 500, 700, 900, 1000 s/mm2) and any pairs of two b values (56 pairs in total) were used to distinguish between benign and malignant meningioma. Independent sample t-test and Mann-Whitney U test were used to evaluate the differential diagnostic value of ADC values, and receiver operating characteristic (ROC) curves were used to determine the diagnostic performance, thresholds, sensitivity, and specificity of ADC values.

Results

The ADC values of benign, malignant, and all meningiomas showed a decreasing trend with the increase of b value, and the smaller the b value, the greater the decrease (as shown in Table 1 and Fig.1). We exclued the sensitivity and specificity of combinations less than 60%, and found that, when the b value was ≥ 200 s/mm2, any combinations of b values could discriminate between benign and malignant meningioma (all P values < 0.05, as shown in Table 2 and Table 3). The ADC values obtained by fitting b = 700 and b = 1000 s/mm2 had the highest AUC value (0.785), and the threshold, sensitivity and specificity for distinguishing between benign and malignant meningiomas were 0.755×10-3 mm2/s, 72.0%, and 72.7%, respectively (as shown in Table 3 and Fig.2). The difference between b = 900 s/mm2 and any other b-values (including 200, 300, 400, 500, 700, and 1000 s/mm2) in benign and malignant meningiomas was statistically significant (P = 0.022, 0.010, 0.015, 0.004, 0.018, and 0.026, respectively, as shown in Table 3).

Conclusions

ADC values of DWI obtained by fitting different combinations of b-values can effectively distinguish between benign and malignant meningiomas. The ADC values of benign, malignant, and all meningioma showed a decreasing trend with the increase of b value, and the smaller the b value, the greater the decrease. b = 700 and 1000 s/mm2 is considered to be the most significant b value.

Acknowledgements

No acknowledgement found.

References

No reference found.

Figures

Fig.1 The linear graph of ADC value among benign, malignant and all meningioma with different combinations of b values. ADC:10-3 mm2/s; b:s/mm2

Fig.2 The ROC curves of ADC (Sensitivity > 60%) fitted with two different b values between benign and malignant meningioma

Table 1 The ADC value fitted with two different b and 11 b values between Benign and Malignant Meningioma

a was used with Mann–Whitney U test, P < 0.05 has a significant difference


Table 2 The ROC analysis of ADC value fitted with two different b values between Benign and Malignant Meningioma

AUC: area under the curve


Table 3 The ROC analysis of ADC (Sensitivity > 60%) fitted with two different b values between benign and malignant meningioma

AUC: area under the curve


Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
3508