How to Train your own AI model for segmenting medical imaging data with the use of MONAI on windows скачать в хорошем качестве

How to Train your own AI model for segmenting medical imaging data with the use of MONAI on windows

In this video I will show what the quickest way is to train your own MONAI model for AI image segmenting. Please do realise that all of the download sequences and computing sequences are heavily sped up. And that the time it might take to download or compute differs per computer and connection. good luck training your own model! Chapters: 00:00 Introduction and goal 00:12 Minimum requirements 00:32 Overview of software used 00:57 Dataset used (Decathlon Heart Segmentation from Kaggle) 01:29 Installing 3D Slicer 02:12 Installing MONAI extensions in 3D Slicer 04:05 Installing Docker 05:32 Installing WSL2 (Linux Subsystem for Windows) 06:17 Fixing WSL Internet issues 07:29 Installing NVIDIA Container Toolkit 08:10 installing MONAI 09:00 Testing GPU support in Docker 09:15 Cloning MONAI Label sample apps 09:50 Editing config for custom label (left atrium only) 10:57 Copying dataset into MONAI-readable directory 11:14 Launching MONAI server via Docker 11:37 Connecting Slicer and MONAI training 13:17 Editing segmentations 15:42 Locating the trained model 16:15 Final thoughts and next steps Links: https://www.kaggle.com/datasets/vivek... https://www.slicer.org/ https://www.docker.com/ https://learn.microsoft.com/en-us/win... https://docs.nvidia.com/datacenter/cl... https://docs.monai.io/en/stable/insta... Commands: wsl --install curl -fsSL https://nvidia.github.io/libnvidia-co... | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \ && curl -s -L https://nvidia.github.io/libnvidia-co... | \ sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \ sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list sudo apt-get update sudo apt-get install -y nvidia-container-toolkit ping google.com sudo nano /etc/resolv.conf docker run --gpus all nvcr.io/nvidia/k8s/cuda-sample:nbody nbody -gpu -benchmark mkdir ~/monai-server cd ~/monai-server git clone https://github.com/Project-MONAI/MONA... docker run -it --rm --gpus all -p 8001:8000 -v ~/monai-server/apps/sample-apps/radiology:/app -v ~/monai-server/data:/data -v ~/monai-server/train:/train -v ~/monai-server/train_cache:/root/.cache/monailabel/train projectmonai/monailabel:latest monailabel start_server --app /app --studies /data --conf models segmentation --conf train_path /train