Cesar A. TELLO DAVILA and Xiomara A. TEJADA TEJADA
Peruvian University of Applied Sciences, Lima, Peru
Cancer recurrence in pediatric patients after treatment represents a significant and urgent challenge to global public health, particularly in low- and middle-income countries, where health systems face structural limitations, fragmented care, and disparities in access to specialized services. These challenges are exacerbated in rural and marginalized regions by geographic isolation, limited connectivity, and the centralization of cancer services in urban centers, leading to late diagnoses, high rates of treatment abandonment, and poor long-term survival outcomes. Addressing this situation requires innovative, scalable, and context-specific solutions capable of transforming follow-up care in vulnerable populations. This article presents the design, development, and preliminary validation of an intelligent telemedicine platform designed for risk-based remote monitoring of pediatric cancer survivors. It is based on a Random Forest machine learning model, trained on anonymized clinical datasets, and implemented using a Python backend. The model classifies patients into three risk levels: low, medium, and high, based on structured symptom reports, laboratory indicators, and other relevant clinical variables. The user interface adheres to the “clinical traffic light” principle, facilitating intuitive risk representation for non-expert users. Furthermore, it enables predictions, thereby strengthening clinical confidence. The platform’s modular and open-source design allows for cost-effective implementation and future enhancement through integration with biometric sensors. The system enables post-treatment detection of potential relapses via automated alerts, allowing healthcare professionals to take preventative action, even in regions with limited infrastructure and intermittent internet access.