S1342
Interdisciplinary - Education in radiation oncology
ESTRO 206
practice could transform communication into a form of precision care, in which technology and empathy converge to enhance the patient experience. References: - Charon R. Narrative Medicine: A model for empathy, reflection, profession and trust. JAMA. 2001. - Greenhalgh T. Cultural contexts of health: the use of narrative research in the health sector. World Health Organization. 2016. - Consensus Conference. Linee di indirizzo per l’utilizzo della medicina narrativa in ambito clinico-assistenziale. 2014. - Marini M.G. Languages of Care in Narrative Medicine. Springer. 2019. - Fioretti C. et al. Research studies on patients' illness experience using the Narrative Medicine approach: a systematic review. BMJ Open. 2016. Keywords: Narrative Medicine, Patient Experience, Simulation Proffered Paper 4488 Know your model – Has complacency set in with Autosegmentation? Jill T Nicholson 1,2 , Guhan Rangaswamy 1 , Maura O'Connell 1 , Palak Sharma 1 , Akvile Kovanaite 1 , Niall O'Sullivan 1 , Pierre Thirion 1,3 , Ciaran Malone 1,4 1 Radiation Oncology, St. Luke's Radiation Oncology Network, Dublin, Ireland. 2 Applied Radiation Therapy Trinity, Discipline of Radiation Therapy, Trinity College Dublin, Ireland. 3 Trinity St. James' Cancer Institute, St. James' Hospital, Dublin, Ireland. 4 Erasmus MC Cancer Institute, University Medical Center,, Rotterdam, Netherlands Purpose/Objective: While autosegmentation has demonstrably improved contouring efficiency, the research focus remains heavily on initial implementation effects, often highlighting the need for user awareness of model limitations. Less is known on how to maintain this critical awareness over time. The aim of this study is to assess user proficiency and identify potential complacency regarding the known limitations and weaknesses of our institutional autosegmentation model three years post-implementation Material/Methods: To assess ongoing awareness of the limitations and weakness of our autosegmenation model we developed a 2-step process. Step 1- The local AI working group conducted a comprehensive audit comparing pre and post AI implementation contouring across three anatomical sites involving 4,291 organ-at- risk (OAR) contours. Overall systematic changes were detected using mean structure volume & pairwise differences significant at p<0.05, however, random gross outlier structures were identified using dosimetric differences exceeding 10Gy. Using these
recurring topics within the texts. Semantic frame analysis also served as a proxy for metaphor detection, which was subsequently explored through deeper qualitative analysis. In addition, emotion detection models for Italian were applied to assess emotional expression across different phases of the therapy, enabling the identification of periods with heightened emotional distress and providing insights into the evolution of patients’ emotional states. Each narrative was then classified according to Kleinman’s tripartite model (illness, disease, sickness) to capture how patients construct meaning around the therapeutic process. Results: The corpus contained over 7.000 tokens and approximately 1.100 unique lemmas. Word count distribution and semantic frame analysis revealed a focus on affective states and emotions, desires, and perceptual experiences. Emotion detection revealed a temporal shift: negative emotions predominated before and during the procedure, while positive emotions emerged mainly afterwards (Figure 1). According to Kleinman’s framework (Figure 2), most narratives predominantly emphasized the subjective, personal dimension (illness), with fewer account reflecting the biomedical perspective (disease). Only limited references to the social dimension (sickness) were identified.
Conclusion: The shift from fear to relief highlights emotional adaptation as a key component of therapy. Understanding this transition through the analysis of patients’ narratives might help clinicians anticipate anxiety and communicate more effectively. Embedding linguistic and narrative analysis into HT
Made with FlippingBook - Share PDF online