Radioprotection 60-4 | Seite 24

B. Amaoui et al.: Radioprotection 2025, 60( 4), 310 – 317 311

Table 1. Demographic data of participants to this study expressed as a percentage %. Items G1 Onco-radiotherapists G2 Residents P-value

N % N %

treatment techniques, accelerate the adoption of best practices in clinical routine for a greater number of patients( Warren et al., 2023), and address challenges linked to limited staffing resources( Poortmans et al., 2020). Moreover, AI is expected to impact the roles of radiation oncology professionals. Automation may replace manual tasks such as the delineation of organs at risk and even target volumes, manual treatment planning, verification of treatment position, and treatment administration. These are all procedures that, when replaced by AI, will increase efficiency and reduce the time spent on planning and treatment( Korreman et al., 2020). Therefore, the role of professionals will shift from manual tasks to the development, individualiSation, and evaluation of radiotherapy treatment( Korreman et al., 2020). However, even if AI can outperform radiologists in cancer detection, there is a need to validate these tools in real clinical settings( McKinney et al., 2020). In his book Deep Medicine, Eric Topol emphasised the limitations of AI in medicine and the importance of maintaining a central role for clinicians in decision-making( Eric Topol, 2019). Furthermore, Char et al., addressed the ethical and practical concerns of using AI in medicine, such as the risks of bias and the necessity of human oversight( Char et al., 2018).

In Morocco, several studies on the perceptions of oncoradiotherapists have been carried out, such as on radiological risks when prescribing CT scans( Amaoui et al., 2023) and also on the practices of current practices in the management of cervical cancer( Amaoui et al., 2024), but concerning the use of AI in this field remains limited. This study aimed to assess the knowledge and perceptions of Moroccan onco-radiotherapists regarding the contribution of artificial intelligence( AI) to clinical practice.

2 Materials and methods

2.1 Study population

This retrospective study was conducted between February and May 2024 that aimed to assess the knowledge and perceptions of Moroccan onco-radiotherapists regarding the contribution of artificial intelligence( AI) to clinical practice. The study population was divided into two groups: onco-radiotherapists( G1) and Onco-Radiotherapy Residents( G2).

2.2 Questionnaire

To assess the knowledge and perceptions of the two groups medical, a standardised anonymous questionnaire of 19 questions was developed with reference to the literature( Ryan et al., 2021; Hindocha et al., 2023). It was created on the platform( Google form) and then sent to the study population via their email address. The first 4 questions explored the demographic characteristics of the population studied( gender, seniority, sector and field of practice), The remaining questions covered the following areas: Areas of RT mastered by participants, participants’ knowledge of AI, use of AI in the acquisition and reconstruction of radiotherapy images, impact of AI on the productivity of radiotherapy practices in image acquisition and reconstruction, influence of AI on the quality of radiotherapy practices in image acquisition and reconstruction, and impact of AI in the field of radiotherapy.

2.3 Statistical analysis

Responses were compared using between the two groups of participants, Fisher’ s exact test of the statistical tool for the social sciences( SPSS version 21.0) was used. with statistical significance set at P < 0.05 the difference is statistically significant.

3 Results

A total of 115 participants completed the questionnaire. They were distributed as follows: 50 onco-radiotherapists and 65 onco-radiotherapy residents.