Kidney Transplant Allocation Simulation Framework

Overview

This project provides a comprehensive simulation framework for modeling and evaluating kidney transplantation allocation policies across a network of hospitals. The framework enables researchers and healthcare professionals to simulate complex organ donation scenarios, test matching strategies, and analyze transplantation outcomes.

Key Features

• 🏥 Multi-Hospital Network Simulation

  • Model organ exchanges across multiple hospitals
  • Configure inter-hospital travel times and connections

• 🧬 Detailed Domain Modeling

  • Sophisticated representations of donors, organs, and patients
  • Captures complex medical factors like blood type, HLA matching, and organ quality

• 🎲 Stochastic Simulation

  • Probabilistic patient and organ arrivals
  • Randomized transplant outcome generation
  • Configurable matching strategies

• 📊 Comprehensive Metrics

  • Calculates detailed transplantation metrics
  • Supports parameter sweeping for in-depth analysis

Example Usage

# Import necessary components.
from sim import (blood_type_compatible, hla_match_score, 
    default_matching_strategy, Simulation, 
    TransplantCenter, Hospital)

# Build a network of hospitals.
center = TransplantCenter()
hosp1 = Hospital("General_Hospital")
hosp2 = Hospital("City_Med")
hosp3 = Hospital("Regional_Care")
center.add_hospital(hosp1)
center.add_hospital(hosp2)
center.add_hospital(hosp3)
# Define connections between hospitals (travel time in hours).
center.add_edge("General_Hospital", "City_Med", travel_time=1.5)
center.add_edge("City_Med", "Regional_Care", travel_time=2.0)
center.add_edge("General_Hospital", "Regional_Care", travel_time=3.0)

# Optionally, plug in a custom matching strategy.
def custom_strategy(patient, organ):
    if not blood_type_compatible(organ.blood_type, patient.blood_type):
        return 0.0
    if not patient.crossmatch(organ):
        return 0.0
    return 1.0 + 0.3 * patient.wait_time + 1.2 * patient.urgency
center.matching_strategy = custom_strategy

# Run the simulation.
sim = Simulation(center, steps=365 * 1, organ_probability=0.7, patient_probability=0.3)
sim.run()
sim.save_logs()
metrics = sim.calculate_metrics()

# Run a parameter sweep.
param_ranges = {
    "organ_probability": [0.6, 0.7, 0.8],
    "patient_probability": [0.2, 0.3, 0.4]
}
Simulation.parameter_sweep(param_ranges, steps=365 * 1, runs_per_combination=3)

Authors

• Aniruth Ananthanarayanan
• Benjamin Zijan Hu
• Alex Sha

Citation

If you use KidneyBench, please cite us as follows:

Ananthananarayanan, Aniruth, Benjamin Hu, and Alex Sha. ‘Network-Based Kidney Allocation Simulation: Evaluating Organ Matching Strategies in Variable Hospital Networks’. bioRxiv, 2025. https://doi.org/10.1101/2025.04.22.650043.

BibTex:

@article {Ananthananarayanan2025.04.22.650043,
	author = {Ananthananarayanan, Aniruth and Hu, Benjamin and Sha, Alex},
	title = {Network-Based Kidney Allocation Simulation: Evaluating Organ Matching Strategies in Variable Hospital Networks},
	elocation-id = {2025.04.22.650043},
	year = {2025},
	doi = {10.1101/2025.04.22.650043},
	publisher = {Cold Spring Harbor Laboratory},
	URL = {https://www.biorxiv.org/content/early/2025/04/25/2025.04.22.650043},
	eprint = {https://www.biorxiv.org/content/early/2025/04/25/2025.04.22.650043.full.pdf},
	journal = {bioRxiv}
}