• ePoster
  • P199

MRT-gestützte Radiomics zur Unterscheidung zwischen perifokalem Ödem und nicht kontrastmittelaufnehmenden Glioblastomanteilen

MRI-based radiomics for differentiation between perifocal edema and non-contrast-enhancing glioblastoma foci

Termin

Datum:
Zeit:
Redezeit:
Diskussionszeit:
Ort / Stream:
ePoster Station 7

Thema

  • Neuroonkologie

Abstract

Various intracranial tumors can lead to the development of perifocal edema. Differentiating edema from tumor on MRI might be easy in metastases and meningiomas but can be challenging or near impossible in patients with glioblastoma. We set out to compare and characterize perifocal edema and non-contrast-enhancing glioblastoma foci by means of radiomic feature analysis to potentially guide further treatment planning and possibly supramarginal resection borders.

A retrospective study of 60 patients (glioblastoma, meningioma, metastases; n=20 each) was conducted using preoperative T2-weighted and contrast-enhanced T1-weighted MRI to identify T2 hyperintense and contrast-enhancing lesions. Brain extraction was performed with HD-BET, followed by manual segmentation and spatial normalization using ANTs. Radiomic features were extracted from T2 hyperintense lesions with PyRadiomics. Statistical analysis included Shapiro-Wilk for normality, followed by ANOVA for normal and Kruskal-Wallis for non-normal features (p < 0.01). Significant features were further analyzed using Tukey"s HSD to compare glioblastoma with other tumor types.

We analyzed 107 radiomic features analyzed across shape, texture, and intensity categories. Among these, 29 significant features (27% of total) were identified for differentiating the T2
hyperintense perifocal regions around glioblastoma from those around meningiomas and metastases. Significant features mainly comprised measures of shape, texture and intensity distribution, e.g., sphericity (p<0.01), surface/volume (p<0.01), entropy (p<0.01), skewness (p<0.01) and kurtosis (p<0.01), showed no significant difference between metastases and meningiomas but differed significantly from the T2-hyperintensity of glioblastoma. The textual complexity of glioblastoma was significantly higher (p<0.01).

Radiomic feature analysis of T2 hyperintense lesions effectively differentiates perifocal T2 hyperintense lesions associated with glioblastoma from the perifocal edema of meningiomas and metastases. These findings may aid in more precise tumor margin delineation, potentially guiding surgical planning for supramaximal resection and improving patient outcomes through more targeted therapeutic strategies.