ESTRO 38 Abstract book

S184 ESTRO 38

decision

making.

are

the

same.

Conclusion Incidence methods capture the first occurrence of an event; LAPERS in contrast excludes transient symptoms and identifies patients with persisting symptoms. When analyzing longitudinal morbidity data, a complementary approach combining the information of both methods improves the understanding of the morbidity profile of a treatment. LAPERS can be utilized to improve tools for informing patients about the burden and duration of toxicity (e.g. figure 1); thus improving the basis for shared

PV-0369 Radiomics in Magnetic Resonance Imaging for prediction of radiotherapy outcomes in cervical cancer N. Thanamitsomboon 1 , N. Kosaisawe 2 , K. Thephamongkhol 1 , P. Dankulchai 1 1 Siriraj hospital, Division of Radiation Oncology- Department of Radiology, Bangkok, Thailand ; 2 University of California- Davis, Department of Molecular and Cellular Biology, Davis, USA Purpose or Objective The aim of this study is to investigate the predictive value of radiomics features from pre-treatment T2-weighted MR images for clinical outcome of radiotherapy with or without concurrent chemotherapy in cervical cancer patients. The primary and secondary clinical outcomes in this study were loco-regional recurrence (LRR) and overall 90 cervical cancer patients with stage IB-IVA between January 2015 to June 2016 were retrospectively enrolled in this study. All patients received definitive radiotherapy with or without concurrent chemotherapy as a primary treatment after underwent T2-weighted MRI simulation before the first fraction of external beam radiotherapy. Gross tumor volume (GTV) was delineated by a radiation oncologist who was an expertise in the gynecological field. Radiomics features [First orders, Gray level co-occurrence matrix (GLCM), and Gray level run length matrix (GLRLM)] survival respectively Material and Methods