Target audience

Scientists, neurosurgeons and radiologists interested in application of study on machine vision

Purpose

The aim of this study was to evaluate Haar-like and MB-LBP features in the classification of preoperative temporal lobe glioma-related epilepsy.

Materials and Methods

Every patient under resting state underwent T2WI and conventional sequence at 3.0 T MR scanner (Prisima Siemens Healthcare, Erlangen, Germany). 64 channel of head-neck coil was used for signal reception. Graphcut app (Department of mathematics, Zhe Jiang University) was used to calculate the tumor volume, Haar-like and MB-LBP features. Independent samples T Test or Mann Whitney U Test was used. A value of P < 0.05 was considered statistically significant. Statistical analyses were performed using SPSS statistics software (version 21.0, IBM).

Results

A total of 33 patients with histologically proven glioma, which include in 8 patients with simplex partial epilepsy and 6 with generalized tonic-clonic epilepsy, 19 patients without epilepsy. The glioma grade has no significant among the groups. When group divide into epilepsy (n=14) or without epilepsy (n=19), patients’ age (P=0.007) and tumor volume (P=0.021) has significant between them, but the Haar-like and MB-LBP features. When group divide into simplex partial epilepsy, generalized tonic-clonic epilepsy and patients without epilepsy, there was significant between generalized tonic-clonic epilepsy and patients without epilepsy in LBP 95 percentiles (P=0.011) ,Haar horizontal properties kurtosis (P=0.01) and patients’ age (P=0.003).

Conclusion

T2WI maybe a useful imaging for predicted glioma-related epilepsy; patients’ age and tumor volume are influence the happen of glioma-related epilepsy.

Acknowledgements

No acknowledgement found.

References

No reference found.