Glioblastomas are an aggressive form of brain tumours. The survival time is about 5 months, 2 if left untreated. Standard procedure for the treatment of this cancer is a combination of surgery and radiation-therapy. In addition, chemotherapy with temozolomide (TMZ) is used. After this treatment it is necessary to examine whether the tumour progresses or the treatment is effective. A distinction between tumour progress and inflammation (pseudo-progress) is possible with biopsy. However, surgical interventions are associated with risks and high costs for little benefit either than this assertation. A reliable method of differentiation is necessary to protect patients who are not benefiting from another TMZ treatment due to toxic effects, and to avoid the unnecessary costs.
In the conventional T1 and T2-weighted MR images it cannot be easily distinguished between a tumour progress and Pseudo-Progress in the region of the removed tumour. However, through a combination of T1 and perfusion MR images it might be possible to distinguish both cases a bit better, since the regional cerebral blood volume is increased in recurrent gliomas (rZBV) by the increased vascularity. So far, no tool or software is available on the market that can automatically distinguish between Progress and Pseudo progress. Such classification software is the development goal of this project.
The market demand is immense, as the decision has been done so far by radiologists without technical support. Our software can prevent interventions associated with the risks of brain surgery. In addition, in patients who don’t show progress, the toxic effects of TMZ is avoided and costs are reduced by saving costly interventions and unnecessary chemotherapy.
Start: October 2017
Conclusion: August 2019
Funding Program: AiF – ZIM framework