Paul Friedrich

4 Papers Accepted at MICCAI 2025 Workshops

I co-authored 4 papers accepted at different MICCAI workshops. Make sure to check out:

• fastWDM3D: Fast and Accurate 3D Healthy Tissue Inpainting accepted at DGM4MICCAI
• VidFuncta: Towards Generalizable Neural Representations for Ultrasound Videos accepted at ASMUS
• Towards Diagnostic Quality Flat-Panel Detector CT Imaging Using Diffusion Models accepted at SWITCH
• Towards MR-Based Trochleoplasty Planning accepted at COLAS

Journal Paper Published

Our paper Generating 3D Pseudo-Healthy Knee MR Images to Support Trochleoplasty Planning which has previously been early-accepted to IPCAI has been published in the International Journal of Computer Assisted Radiology and Surgery.

New Preprint

Our paper MedFuncta: Modality-Agnostic Representations Based on Efficient Neural Fields is available on arXiv. We introduce a modality agnostic continuous data representation based on meta-learned neural fields (NFs) and show how NFs can be generalized from single instances to large datasets. We further demonstrate that we can solve relevant downstream tasks by applying neural networks to this functional data representation.

Book Chapter Published

Our review chapter Deep Generative Models for 3D Medical Image Synthesis has been published in Generative Machine Learning Models in Medical Image Computing. We review deep generative models (VAEs, GANs, and DDMs) for 3D medical image synthesis, covering basic principles, advances, and application for relevant downstream tasks. The chapter also addresses evaluation metrics for image fidelity, diversity, utility, and privacy.

Conference Paper Published

Our paper WDM: 3D Wavelet Diffusion Models for High-Resolution Medical Image Synthesis has been published. We propose a simple yet effective way of scaling 3D diffusion models to high resolutions (256 x 256 x 256 on a single 40 GB GPU) by applying Discrete Wavelet Transform for spatial dimensionality reduction.