B. Amaoui et al.: Radioprotection 2025, 60( 4), 310 – 317 313
Table 3. Data concerning the impact of AI on the productivity of radiotherapy practices in image acquisition and reconstruction by participant group.
The impact of AI on the productivity of radiotherapy practices in image acquisition and reconstruction Lower productivity No change Improve productivity Not applicable G1 G2 G1 G2 G1 G2 G1 G2 P-value
Image intervention and reconstruction |
0 |
0 |
20 |
7.7 |
70 |
76.9 |
10 |
15.4 |
0.809 |
Image fusion |
0 |
0 |
20 |
7.7 |
70 |
69.2 |
10 |
23.1 |
0.559 |
Target volume contouring |
10 |
0 |
30 |
7.7 |
60 |
61.5 |
0 |
30.8 |
0.111 |
OAR contouring |
10 |
0 |
20 |
15.4 |
70 |
69.2 |
0 |
15.4 |
0.559 |
Setting up a plan |
10 |
0 |
10 |
15.4 |
80 |
61.5 |
0 |
23.1 |
0.405 |
Plan optimisation |
10 |
0 |
10 |
7.7 |
70 |
84.6 |
10 |
7.7 |
0.847 |
Plan evaluation |
0 |
0 |
10 |
15.4 |
70 |
69.2 |
20 |
15.4 |
1.000 |
Quality assurance |
0 |
0 |
10 |
23.1 |
60 |
46.2 |
30 |
30.8 |
0.855 |
Image analysis and matching |
0 |
0 |
10 |
15.4 |
90 |
61.5 |
0 |
23.1 |
0.405 |
Table 4. Data concerning the influence of AI on the quality of radiotherapy practices in terms of image acquisition and reconstruction stratified by groups.
The influence of AI on the quality of radiotherapy practices in terms of image acquisition and reconstruction Decreases quality No change Increases quality Not applicable P-value G1 G2 G1 G2 G1 G2 G1 G2
Image intervention and reconstruction |
0 |
0 |
10 |
7.7 |
90 |
76.9 |
0 |
15.4 |
0.736 |
Image fusion |
0 |
0 |
10 |
0 |
90 |
84.6 |
0 |
15.4 |
0.325 |
Target volume contouring |
10 |
15.4 |
10 |
0 |
70 |
53.8 |
10 |
30.8 |
0.516 |
OAR contouring |
20 |
7.7 |
10 |
15.4 |
60 |
69.2 |
10 |
7.7 |
0.899 |
Setting up a plan |
10 |
7.7 |
10 |
0 |
80 |
76.9 |
0 |
15.4 |
0.566 |
Plan optimisation |
10 |
7.7 |
10 |
0 |
70 |
76.9 |
10 |
15.4 |
0.873 |
Plan evaluation |
10 |
7.7 |
10 |
0 |
70 |
69.2 |
10 |
23.1 |
0.706 |
Quality assurance |
10 |
7.7 |
10 |
15.4 |
70 |
53.8 |
10 |
23.1 |
0.906 |
Image analysis and matching |
10 |
0 |
10 |
7.7 |
80 |
76.9 |
0 |
15.4 |
0.566 |
3.6 The influence of AI on the quality of radiotherapy practices in terms of image acquisition and reconstruction
According to Table 4, most G1s favour an increase in quality in the Image Intervention and Reconstruction functions( 90 %), image fusion( 90 %), target volume contouring( 70 %), organ at risk contouring( 60 %), plan implementation( 80 %), plan optimisation( 70 %), plan evaluation( 70 %), image analysis and matching( 80 %). The majority of G2s also supported of an increase in productivity in the functions of intervention and image reconstruction( 77 %), image fusion( 84.6 %), contouring of target volumes( 54 %), contouring of organs at risk( 69.2 %), implementation of a plan( 77 %), Optimisation of a plan( 77 %), evaluation of a plan( 69. 2 %), and image analysis and matching( 77 %).
3.7 Impact of AI on the field of radiotherapy
Regarding the perceived impact of AI on the field of radiotherapy, 70 % of G1s and 84.6 % of G2s are optimistic about the use of AI in radiotherapy, 70 % of G1s and 84.6 % of G2s think that AI will increase job satisfaction, 80 % of G1s and 92.3 % of G2s believe that AI will have a positive impact on the patient’ s treatment pathway, 40 % of G1s think AI may impact their current roles, while 53.8 of G2s think AI will have no impact on their current role. Additionally, 40 % of G1s and 31 % of G2s expressed a neutral stance on the introduction of AI into healthcare until the’ black box’ aspect becomes transparent.
4 Discussion
The results presented in this study underscore several important findings regarding the use and perception of artificial intelligence( AI) tools in radiotherapy, as well as the differences between experienced onco-radiotherapists( G1) and onco-radiotherapy residents( G2).
The results show that onco-radiotherapists( G1) and residents( G2) have different mastery of key steps in radiotherapy. While all G1 and G2 report proficiency in CT simulation, a significantly lower proportion of both groups