Systematic Review On the Use of Artificial Intelligence in Brachytherapy
J Theeler1, Y Kim2*, University Of Iowa, (1) Biomedical Engineering Department,(2) Radiation Oncology Department, Iowa City, Iowa, USA
Presentations
SU-H400-IePD-F7-2 (Sunday, 7/10/2022) 4:00 PM - 4:30 PM [Eastern Time (GMT-4)]
Exhibit Hall | Forum 7
Purpose: To systematically review literature on the use of artificial intelligence (AI), including deep learning (DL) and machine learning (ML) approaches, in brachytherapy (BT). Evaluation was performed based on application to BT, disease site, and image modality. AI methods were quantitively and/or qualitatively compared to ground truth for accuracy and timing.
Methods: Included studies were accepted peer-reviewed journal articles on AI in BT published from 01/01/80 – 08/01/21 on PubMed, Google Scholar, Cochrane Library, and University of Iowa Libraries databases. Articles were reviewed and either included or excluded due to inclusion criteria or scope. Studies were searched for application to BT, AI description, training and testing datasets, input and output of AI, treatment description, ground truth classification, accuracy compared to ground truth, and time for results. This review was performed based on the Preferred Reporting Items for Systemic reviews and Meta Analyses (PRISMA) guidelines.
Results: A total of 76 studies were identified as fitting inclusion criteria and scope after an initial yield of 7,488 results from database searches. Studies per application were 21, 17, 6, 7, 14, and 11 for applicator/needle reconstruction, target or organs-at-risk segmentation, imaging applications, dose calculation, outcome prediction, and other planning applications, respectively. Studies per disease site were 41, 24, 1, 1, and 9 for prostate, gynecological, breast, choroidal, and no specific site or multiple sites, respectively. Studies demonstrate that AI has the potential to improve the results and time required for all steps of BT process.
Conclusion: The use of AI contributes to improving the efficiency of time-consuming repetitive treatment planning tasks such as contouring and applicator/needle reconstructions, and is promising for improving patient outcomes through predicting clinical outcomes.
Funding Support, Disclosures, and Conflict of Interest: The study was supported by ICRU (Iowa Center for Research by Undergraduates) Fellowship Program of the University of Iowa.
Keywords
Taxonomy
TH- Brachytherapy: Treatment planning using machine learning/automation
Contact Email