Current state evaluation of challenges and opportunities in standardized nomenclature and artificial intelligence adoption in Canadian radiation oncology practice

Abstract

PURPOSE: Variations in the implementation of emerging data nomenclature standards and usage of artificial intelligence (AI) tools in Canadian radiation therapy (RT) centres have not yet been fully characterized. To address this, a current state analysis was conducted to serve as a baseline assessment and to identify gaps and opportunities for harmonized pan-Canadian data practices and the adoption of AI in clinical settings within radiation oncology. METHODS AND MATERIALS: A survey was distributed to all Canadian RT centres with the aim to describe the perceived status and characteristics of implementation of standardized nomenclature, usage of AI tools, and relevant gaps and opportunities in this field. RESULTS: Thirty three of 51 (64.7%) Canadian RT centres responded. Responses characterized variation in standardized nomenclature implementation and usage of AI tools across centres, with some trends between regions. Approximately two-thirds of RT centres were using TG-263 guidance of the American Association of Physicists in Medicine (70.0%, n=23). whereas only a third of centres reported awareness of next steps for O3. The most commonly-reported barriers to data standardization included a lack of resources and forcing functions. Automation and quality improvement were recognized as facilitators, with (81.8%, n=27) are using automation tools to support standardization. CONCLUSIONS: This current state analysis informs and directs future initiatives to improve standardized nomenclature implementation and support informed AI adoption within RT, with the goal of ultimately improving RT quality and safety. Canada is well-positioned to lead data standardization efforts and serve as a case study to potentially provide guidance at an international level to equal partners.

Ähnliche Arbeiten