Current availability of radiotherapy devices in Peru and artificial intelligence-based analysis for constructing a nationwide implementation plan.

Purpose: We provide for the first time a comprehensive situational diagnosis and propose an artificial intelligence (AI)-assisted nationwide plan of implementation, attending the most urgent needs.

Methods: Baseline information was collected from open-source databases of the Peruvian Government. Data on cancer incidence from the Health Authorities and GLOBOCAN were collected and compared. The existing external-beam radiotherapy (EBRT) devices and brachytherapy (BT) units were identified and information on their obsolescence was additionally collected. The ten most common cancer entities with RT indication were considered for the analysis. Utilizing open-source softwares, population clusters based on density, cancer incidence, geographic distribution, existing facilities able to be implemented with radiotherapy and travel times for patients were defined. A coding for identifying the best possible locations with AI was developed, keeping the allocation of resources to the minimum possible. A projection until 2030 on required resources was additionally elaborated.

Results: As of 2023 eight additional EBRT and seven BT devices were needed to cover the existing demand. The artificial-intelligence algorithm yielded the regions where these resources should be primarily allocated. An increase in demand of approximately 22% is expected until 2030, which translates into additional 23 EBRT and 16 BT devices, considering the replacement of obsolete units until then.

Conclusion: Increased investment pace is required to cover the existing RT demand in Peru. This AI-assisted analysis might help prioritize allocation of resources. The code employed in this work will be made publicly available, so this method could be replicated in other developing economies.