ESTRO 2024 - Abstract Book

S4940

Physics - Quality assurance and auditing

ESTRO 2024

Despite progress within the community to improve thoracic treatment, more improvement is needed to consistently produce high-quality radiation therapy. To reach this end, improvement can come from several factors, chiefly improved accuracy from dose estimations i.e. better dose calculation algorithms such as Monte Carlo, advancements in irradiation techniques, and improved understanding in using various motion management techniques. Random forest modeling can allow the community to understand the underpinnings of why failures occur and how clinics can solve these failures and improve dose deliver. To facilitate this, as suggested by the important variables in random forest prediction, clinics should focus on how target coverage is created and delivered. This can be done by monitoring closely the (1) set-up of phantom, (2) treatment irradiation technique, (3) motion management technique, (4) usage of MLCs, and (5) dose homogeneity.

Keywords: Random Forest, phantoms, clinical trials

3100

Mini-Oral

Automated Initial Chart Checks in Radiotherapy: Enhancing Safety, Plan Quality, & Efficiency

Maximilian Grohmann, Cordula Petersen, Andrea Baehr, Manuel Todorovic, Sebastian Schäfer

University Medical Center Hamburg-Eppendorf, Department of Radiotherapy and Radiation Oncology, Hamburg, Germany

Purpose/Objective:

Physics Chart Review (PCR) is one of the most effective quality assurance methods in radiotherapy for intercepting errors before they reach the patient [1]. However, the process is not without its constraints, with human oversight being a notable limiting factor [2]. The integration of automation into the PCR process, a strategy supported by monocentric studies [3–5] and recommended by the AAPM Task Group report 275 [6], aims to enhance error detection. This advancement permits medical physicists to engage more thoroughly in comprehensive plan quality analysis, addressing traditional PCR limitations and potentially elevating the standard of care in radiotherapy. Our study aims to validate these claims utilizing a proprietary script, named PlanCheck.

Material/Methods:

PlanCheck integrates with the Treatment Planning System (TPS) Eclipse v. 16.1 through its Application Programming Interface (API) and accesses the ARIA SQL database to perform comprehensive checks. It meticulously evaluates 122 distinct criteria that encompass a wide array of plan- and patient-specific factors. These considerations range from nomenclature, scheduling, patient demographics (e.g., presence of a pacemaker, anatomical constraints), field geometry, dose, calculation and optimization settings, planning CT, and technique. Enhancing its functionality, PlanCheck incorporates a graphical user interface (GUI) within the TPS, facilitating the direct visualization of findings (Fig.1). This feature promptly highlights and flags any anomalies or errors, enabling immediate review and action.