Life sciences and Health Seminar Leah Dickhoff; Evi Sijben

Automated optimization for cervix brachytherapy requires more then the EMBRACE-II planning aims; AI for decision support of paraganglioma treatment planning
  • What Evolutionary Intelligence English
  • When 19-04-2022 from 16:00 to 17:00 (Europe/Amsterdam / UTC200)
  • Where CWI, room L016
  • Contact Name
  • Web Visit external website
  • Add event to calendar iCal

Join Zoom Meeting
https://cwi-nl.zoom.us/j/81535249101?pwd=ZDh3Y1kyZmxBUml1Tm00Rkx0NXBwdz09

Meeting ID: 815 3524 9101
Passcode: 967201

Title: Automated optimization for cervix brachytherapy requires more then the EMBRACE-II planning aims
Speaker: Leah Dickhoff

Abstract: A bi-objective model was introduced for MO-RV-GOMEA-based optimization for prostate HDR brachytherapy, intuitively capturing the dose trade-off between tumour target coverage and healthy organ sparing. This approach entails direct optimization on dose volume indices specified in a clinical protocol. We studied its immediate extension to cervical cancer brachytherapy by optimizing on planning aims from the officially recommended EMBRACE-II protocol. Results were judged as clinically unsatisfactory by medical specialists because of undesirable properties in the dose distributions, indicating that additional aims are needed. A first approach optimizing on both the EMBRACE-II aims as well as additionally tuned aims has been developed and indicates promising results.


Title: AI for decision support of paraganglioma treatment planning
Speaker: Evi Sijben

Abstract: Paraganglioma are tumors that appear in the head and neck area. Although they are usually benign slow-growing tumors, they can cause severe complaints. These complaints are dependent on the growth of the tumor. being able to predict the future volume of the tumor would be helpful to give inevitable treatment early on but to avoid unnecessary and stressful surveillance. In this talk, methods for paraganglioma tumor measurement and knowledge about tumor growth are presented as well as possible ways to further support the decision making process with help of AI.