mrsegmentator-konfai 1.5.3

Fast and lightweight MRSegmentator CLI powered by the KonfAI framework.

MRSegmentator-KonfAI

Fast and lightweight CLI for whole-body MRI segmentation using MRSegmentator models within the KonfAI framework.

---

🧩 Overview

MRSegmentator-KonfAI is a lightweight command-line interface (CLI) for running MRSegmentator models through the KonfAI deep learning framework.

It provides fast and efficient inference for whole-body MRI segmentation, including on low-resource hardware.

Pretrained models are automatically downloaded from Hugging Face Hub.

⭐ Key Advantages

📦 Lightweight model distribution

• ~128 MB per model, with up to 5 folds available
• Download only the folds you need
• Total size with 5 folds: ~640 MB
• 🔁 Compared to ~1.07 GB for the original full MRSegmentator model distribution

➡️ Faster setup, smaller disk footprint

---

⚡ Efficient inference

🔬 Performance comparison (single CT volume)

Experimental setup

• Input volume size: 512 × 512 × 366
• GPU: NVIDIA RTX 6000
• CPU: Intel® Xeon® w5-3425

---

Original MRSegmentator

Configuration	Time	Peak RAM	Peak VRAM
1 fold	160.3 s	82.3 GB	~3.5 GB
5 folds	166.4 s	82.8 GB	~5.1 GB

---

MRSegmentator-KonfAI

Configuration	Time	Peak RAM	Peak VRAM
1 fold	42.6 s	29.7 GB	~2.2 GB
5 folds (ensemble)	49.0 s	29.7 GB	~3.7 GB

---

📈 Key observations

• ~3–4× faster inference compared to the original MRSegmentator
• ~2.8× lower RAM usage (≈ 30 GB vs ≈ 83 GB)

---

🧠 Features

• ⚡ Fast inference powered by KonfAI
• 🤗 Automatic model download from Hugging Face
• 🧩 Multi-model ensembling
• 🧠 Supports evaluation workflows with reference data, and uncertainty estimation without reference
• 🧾 Multi-format compatibility: supports all major medical image formats handled by ITK

---

🚀 Installation

From PyPI:

python -m pip install mrsegmentator-konfai

From source:

git clone https://github.com/vboussot/KonfAI.git
python -m pip install -e apps/mrsegmentator

---

⚙️ Usage

Run inference on an MRI scan:

mrsegmentator-konfai -i path/to/input.nii.gz -o ./Output/

Optional arguments

Flag	Description	Default
-i, --input	Path to the input MRI volume	required
-o, --output	Path to save the segmentation	./Output/
--gt	Path to reference segmentation (ground truth), if available (enables evaluation workflows)	unset
--mask	Path to region-of-interest mask used for evaluation and uncertainty analysis	unset
-f, --folds	Number of model folds to ensemble (1–5)	2
-uncertainty	Save uncertainty maps	False
--gpu	GPU list (e.g. 0 or 0,1)	CPU if unset
--cpu	Number of CPU cores (if no GPU)	1
-q, --quiet	Suppress console output	False

Example

mrsegmentator-konfai -i path/to/input.nii.gz -o ./Output/ --gt path/to/reference.nii.gz --mask path/to/mask.nii.gz --gpu 0 -f 3 -uncertainty

---

📖 Reference

If you use MRSegmentator-KonfAI in your work, please cite the original MRSegmentator work in addition to this CLI tool.

• Häntze, H. et al. (2025).
  Segmenting Whole-Body MRI and CT for Multiorgan Anatomic Structure Delineation.
  Radiology: Artificial Intelligence, 7(6). https://doi.org/10.1148/ryai.240777

• Boussot, V., & Dillenseger, J.-L. (2025).
  KonfAI: A Modular and Fully Configurable Framework for Deep Learning in Medical Imaging.
  arXiv preprint arXiv:2508.09823

---

🔗 Links

• 🧠 Original MRSegmentator: github.com/hhaentze/MRSegmentator
• 🤗 Model Hub: huggingface.co/VBoussot/MRSegmentator-KonfAI
• 📦 PyPI Package: pypi.org/project/mrsegmentator-konfai