# Generation of multimodal realistic computational phantoms as a test-bed for validating deep learning-based cross-modality synthesis techniques

## TL;DR

This paper introduces a new method to create realistic CT and MRI images using computational phantoms to better test AI models for medical image translation in radiotherapy.

## Contribution

The novelty is using generated phantoms as validation data for deep learning-based cross-modality synthesis techniques.

## Key findings

- Strong anatomical consistency was observed between generated and original phantoms.
- High histogram correlation with patient images was achieved for both MRI and CT.
- Dosimetric accuracy of the generated data was comparable to real data.

## Abstract

The validation of multimodal deep learning models for medical image translation is limited by the lack of high-quality, paired datasets. We propose a novel framework that leverages computational phantoms to generate realistic CT and MRI images, enabling reliable ground-truth datasets for robust validation of artificial intelligence (AI) methods that generate synthetic CT (sCT) from MRI, specifically for radiotherapy applications. Two CycleGANs (cycle-consistent generative adversarial networks) were trained to transfer the imaging style of real patients onto CT and MRI phantoms, producing synthetic data with realistic textures and continuous intensity distributions. These data were evaluated through paired assessments with original phantoms, unpaired comparisons with patient scans, and dosimetric analysis using patient-specific radiotherapy treatment plans. Additional external validation was performed on public CT datasets to assess the generalizability to unseen data. The resulting, paired CT/MRI phantoms were used to validate a GAN-based model for sCT generation from abdominal MRI in particle therapy, available in the literature. Results showed strong anatomical consistency with original phantoms, high histogram correlation with patient images (HistCC = 0.998 ± 0.001 for MRI, HistCC = 0.97 ± 0.04 for CT), and dosimetric accuracy comparable to real data. The novelty of this work lies in using generated phantoms as validation data for deep learning-based cross-modality synthesis techniques.

The online version contains supplementary material available at 10.1007/s11517-025-03437-4.