A Stable Matching Assignment for Cancer Treatment Centers using Survival Analysis
Navid Seidi
TL;DR
The work addresses allocating high-risk cancer patients to National Cancer Institute-designated centers by combining Survival Analysis with a stable matching framework. It computes patient risk scores $rs_i$ to select candidates and then solves a center assignment problem on a bipartite graph defined by location and affordability thresholds, with the goal of fully occupying staffed beds. A two-loop algorithm manages risk-based admission and a center-aware stable matching that respects capacity and accessibility constraints, demonstrated on SEER-derived data and a fixed center set. The approach offers a patient-centric, occupancy-maximizing mechanism for cancer care delivery, with practical relevance for resource planning and policy, and points to extensions for multi-treatment settings and dynamic bed flows.
Abstract
The treatment of cancer is one of the most discussed issues in the realm of contemporary public health research. One of the primary concerns of both the general public and the government is the development of the most effective cancer treatment at the most affordable price. This is due to the fact that the number of persons diagnosed with cancer increases on an annual basis. Within the scope of this project, we propose the development of a system for the recommendation of treatment centers. This system would initially select patients who posed a higher risk value, and then it would recommend the most appropriate cancer treatment center for those patients based on their income and the location where they lived using a stable matching algorithm.