Vestibular schwannomas, also known as acoustic neuromas, are benign brain tumors located on the auditory-vestibular nerve. It is a slow-growing tumor that presses against the hearing and balance nerve causing dizziness, or one-sided ringing buzzing or hearing loss. It can also cause facial numbness or paralysis. When tumors grow large enough, they can even press against the brain stem, resulting in life-threatening symptoms. Vestibular schwannomas are usually diagnosed with magnetic resonance imaging (MRI) and there are two different treatment options possible: microsurgery and radiosurgery. Microsurgery involves open skull base surgery in which the tumor is partially or fully removed. With radiosurgery, like Gamma Knife radiosurgery, the tumor is subjected to high-precision radiation. Due to the difficult location of the tumor, microsurgery is not the preferred treatment modality. It is very invasive for the patient, involving hospitalization (5-8 days), including 1 or 2 days on the intensive care, a long recovery time (4 to 6 weeks) and high risk of complications. Radiosurgical treatment however is less invasive and is performed in a single day, without the need for hospitalization. The risk for complications is also significantly lower with respect to the risks involved in the microsurgical removal of the tumor. It is however necessary to have a life-long surveillance of the tumor after radiosurgery, because some tumors might not respond to the treatment and will continue to grow. Because of this, a secondary treatment may be necessary. It is therefore very valuable to predict whether the tumor will respond to the treatment or not. Such prediction is of crucial importance for each specific patient to choose the most suited treatment modality. Furthermore, an estimation of the likelihood of tumor control after Gamma Knife radiosurgery can potentially provide a basis for an individualized surveillance protocol, reducing the overall number of follow-up hospital visits and MRI scans.
Currently, it is not possible to predict the Gamma Knife treatment outcome of vestibular schwannomas. Some reported influential parameters are the tumor volume and the pretreatment growth speed of the tumor, although these are also disputed. Therefore, in this project, we aim to find discriminative parameters that can be used to predict the radiosurgical treatment outcome. This involves discovering possible MRI characteristics of the tumor that can be used in a machine learning environment in order to find a model that can accurately predict the Gamma Knife treatment response of a vestibular schwannoma.