# PRICE: A Personalized Recursive Intelligent Cost Estimation Framework for Rare Disease Diagnosis

## TL;DR

PRICE is a new AI-based framework that helps personalize and optimize diagnostic cost-effectiveness for rare diseases.

## Contribution

PRICE introduces a dynamic, tree-based model for individualized cost-effectiveness analysis in rare disease diagnosis.

## Key findings

- PRICE evaluates diagnostic strategies using back-propagation and utility-based effectiveness metrics.
- The framework adapts to AI performance changes, influencing optimal strategy selection.
- An interactive web tool visualizes diagnostic pathways, aiding clinician and patient decision-making.

## Abstract

Rare disease diagnosis often involves complex procedures that can be both costly and time-consuming. Traditional cost-effectiveness analyses typically employ static models, applying uniform diagnostic strategies across diverse patient populations. With advancements in artificial intelligence (AI) and a growing emphasis on personalized medicine, there is a pressing need for dynamic frameworks that assess diagnostic cost-effectiveness at the individual patient level.

We introduce the PRICE framework—a novel, tree-based analytical model designed to evaluate the cost-effectiveness of diagnostic strategies, accommodating both expert-alone and AI-delegation decision-making modes. The model computes the expected cost of a diagnostic process via a back-propagation algorithm and quantifies effectiveness through a utility-based approach. Parameters such as disease prevalence, test costs, test performance metrics, and turnaround time are incorporated, allowing for individualized assessments.

We conducted a case study applying the framework to four diagnostic strategies for developmental delay (DD) and multiple congenital anomalies (MCA). The results demonstrate how PRICE can support patient decision-making by modeling outcomes under varying parameters such as test cost and accuracy. Additionally, we show that changes in AI performance influence the selection of optimal cost-efficient strategies under AI delegation. To facilitate interpretation and engagement, we developed an interactive web-based tool that visualizes and simulates diagnostic pathways in real time, enhancing decision-making support for both clinicians and patients.

PRICE is a flexible cost-effective analysis framework that captures the sequential and recursive nature of real-world diagnostic workflows, with the ability to be extended to future AI-integrated clinical practice. It enables personalized evaluations of diagnostic strategies from both economic and clinical perspectives, promoting more informed and individualized decision-making, especially in rare disease diagnosis.