Enhance MONAI Label tutorial on using VISTA3D

README.md

@@ -83,6 +83,7 @@ In addition, you can find a table of the basic supported fields, modalities, vie
       <li>Segmentation</li>
       <li>DeepGrow</li>
       <li>DeepEdit</li>
+      <li>VISTA3D (Universal)</li>
       <li>SAM2 (2D/3D)</li>
     </ul>
   </td>
@@ -345,6 +346,11 @@ monailabel start_server --app apps/radiology --studies datasets/Task09_Spleen/im
     - [MONAILabel: Multi-organ Segmentation with 3D Slicer](https://github.com/Project-MONAI/tutorials/blob/main/monailabel/monailabel_monaibundle_3dslicer_multiorgan_seg.ipynb): Multi-organ segmentation with CT scans in 3D Slicer.
     - [MONAILabel: Whole Body CT Segmentation with 3D Slicer](https://github.com/Project-MONAI/tutorials/blob/main/monailabel/monailabel_wholebody_totalSegmentator_3dslicer.ipynb): Whole body (104 structures) segmentation with CT scans.
     - [MONAILabel: Lung nodule CT Detection with 3D Slicer](https://github.com/Project-MONAI/tutorials/blob/main/monailabel/monailabel_monaibundle_3dslicer_lung_nodule_detection.ipynb): Lung nodule detection task with CT scans.
+- **VISTA3D App**:
+  - Viewer: [OHIF](https://ohif.org/) | [3D Slicer](https://www.slicer.org/) | Datastore: Local/DICOM Web | Task: Multi-organ Segmentation
+    - [MONAILabel: VISTA3D Quick Start](docs/source/tutorials/vista3d_quickstart.md): Fast setup guide for VISTA3D universal segmentation model.
+    - [MONAILabel: VISTA3D Complete Tutorial](docs/source/tutorials/vista3d_tutorial.md): Comprehensive tutorial for VISTA3D model setup, interactive prompting, and multi-organ segmentation workflows with 117+ anatomical structures.
+    - [MONAILabel: VISTA3D Configuration Examples](docs/source/tutorials/vista3d_configurations.md): Sample configurations for different VISTA3D deployment scenarios.
 - **Pathology App**:
   - Viewer: [QuPath](https://qupath.github.io/) | Datastore: Local | Task: Segmentation
     - [MONAILabel: Nuclei Segmentation with QuPath](https://github.com/Project-MONAI/tutorials/blob/main/monailabel/monailabel_pathology_nuclei_segmentation_QuPath.ipynb) Nuclei segmentation with QuPath setup and Nuclick models.

docs/source/index.rst

@@ -80,6 +80,7 @@ Table of Contents
    whatsnew
    installation
    quickstart
+   tutorials
    modules
    appdeployment
    apis

docs/source/tutorials.rst

@@ -0,0 +1,95 @@
+.. comment
+    Copyright (c) MONAI Consortium
+    Licensed under the Apache License, Version 2.0 (the "License");
+    you may not use this file except in compliance with the License.
+    You may obtain a copy of the License at
+        http://www.apache.org/licenses/LICENSE-2.0
+    Unless required by applicable law or agreed to in writing, software
+    distributed under the License is distributed on an "AS IS" BASIS,
+    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+    See the License for the specific language governing permissions and
+    limitations under the License.
+
+=========
+Tutorials
+=========
+
+This section provides comprehensive tutorials for using MONAI Label with advanced models and viewers.
+
+VISTA3D Universal Segmentation
+===============================
+
+VISTA3D is a universal foundation model for 3D CT segmentation that can segment 117+ anatomical structures 
+including organs, bones, vessels, and soft tissues. It supports both automatic and interactive segmentation workflows.
+
+**Quick Start**
+
+For immediate setup and basic usage, see the `VISTA3D Quick Start Guide`_.
+
+**Complete Tutorial** 
+
+For comprehensive instructions including advanced configuration, troubleshooting, and clinical workflows, 
+see the `VISTA3D Complete Tutorial`_.
+
+**Configuration Examples**
+
+For sample configurations covering different deployment scenarios (memory optimization, PACS integration, 
+multi-model setup, etc.), see the `VISTA3D Configuration Examples`_.
+
+Key Features
+------------
+
+- **Universal Segmentation**: Supports 117+ anatomical classes
+- **Interactive Prompting**: Point-based and class-based interactive segmentation  
+- **Multi-Viewer Support**: Works with OHIF and 3D Slicer
+- **High Performance**: Optimized for clinical workflows
+- **Flexible Deployment**: Local, cloud, and PACS integration options
+
+Supported Viewers
+-----------------
+
+VISTA3D integration is available through:
+
+- **OHIF Viewer**: Web-based DICOM viewer with full VISTA3D support
+- **3D Slicer**: Desktop application with VISTA3D capabilities
+
+Prerequisites
+-------------
+
+- Python 3.8+ with MONAI Label installed
+- NVIDIA GPU with 8GB+ VRAM (16GB+ recommended)  
+- CT data in DICOM or NIfTI format
+- Compatible viewer (OHIF or 3D Slicer)
+
+Basic Usage
+-----------
+
+1. **Setup**::
+
+    # Download bundle app
+    monailabel apps --download --name monaibundle --output apps
+
+    # Start VISTA3D server  
+    monailabel start_server \\
+      --app apps/monaibundle \\
+      --studies datasets/ct_volumes \\
+      --conf models vista3d \\
+      --conf preload true
+
+2. **Automatic Segmentation**: Load CT volume and run auto-segmentation for 117+ structures
+
+3. **Interactive Prompting**: Use point prompts to refine segmentation of specific anatomy  
+
+4. **Class Selection**: Choose anatomical categories (organs, bones, vessels) for targeted segmentation
+
+For detailed instructions, examples, and troubleshooting, refer to the tutorial links above.
+
+Additional Resources
+====================
+
+For more MONAI Label tutorials covering other models and use cases, visit the 
+`MONAI Tutorials Repository <https://github.com/Project-MONAI/tutorials/tree/main/monailabel>`_.
+
+.. _VISTA3D Quick Start Guide: tutorials/vista3d_quickstart.md
+.. _VISTA3D Complete Tutorial: tutorials/vista3d_tutorial.md  
+.. _VISTA3D Configuration Examples: tutorials/vista3d_configurations.md

docs/source/tutorials/README.md

@@ -0,0 +1,23 @@
+# MONAI Label Tutorials
+
+This directory contains comprehensive tutorials for using MONAI Label with various models and viewers.
+
+## Available Tutorials
+
+### VISTA3D Universal Segmentation
+- **[VISTA3D Quick Start](vista3d_quickstart.md)** - Get started with VISTA3D in minutes
+- **[VISTA3D Complete Tutorial](vista3d_tutorial.md)** - Comprehensive guide for VISTA3D setup and usage
+- **[VISTA3D Configuration Examples](vista3d_configurations.md)** - Sample configurations for different deployment scenarios
+
+### Getting Started
+1. Start with the [Quick Start Guide](vista3d_quickstart.md) for immediate setup
+2. Read the [Complete Tutorial](vista3d_tutorial.md) for detailed instructions
+3. Use [Configuration Examples](vista3d_configurations.md) for specific deployment needs
+
+### Prerequisites
+- MONAI Label installed
+- Compatible viewer (OHIF, 3D Slicer)
+- CT data for segmentation
+- GPU recommended for optimal performance
+
+For general MONAI Label tutorials, visit the [MONAI Tutorials Repository](https://github.com/Project-MONAI/tutorials/tree/main/monailabel).

docs/source/tutorials/vista3d_configurations.md

@@ -0,0 +1,187 @@
+# VISTA3D Configuration Examples
+
+This document provides sample configurations for integrating VISTA3D with MONAI Label.
+
+## Basic VISTA3D Server Configuration
+
+### 1. Standard Setup
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models vista3d \
+  --conf preload true \
+  --conf skip_trainers true
+```
+
+### 2. Memory-Optimized Setup
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models vista3d \
+  --conf vista3d_patch_size "[64,64,64]" \
+  --conf vista3d_overlap 0.125 \
+  --conf vista3d_cpu_fallback true
+```
+
+### 3. High-Performance Setup
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models vista3d \
+  --conf vista3d_patch_size "[128,128,128]" \
+  --conf vista3d_overlap 0.5 \
+  --conf preload true \
+  --conf workers 4
+```
+
+## VISTA3D with DICOM Web
+
+### 4. PACS Integration
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies "http://pacs.hospital.com/dicom-web" \
+  --studies_user "username" \
+  --studies_password "password" \
+  --conf models vista3d \
+  --conf preload true
+```
+
+## Custom VISTA3D Bundle Configuration
+
+### 5. Custom Bundle Path
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models "vista3d:/path/to/custom/vista3d/bundle" \
+  --conf vista3d_prompting true
+```
+
+## Multi-Model Setup
+
+### 6. VISTA3D with Other Models
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models "vista3d,spleen_ct_segmentation,swin_unetr_btcv_segmentation" \
+  --conf preload true
+```
+
+## VISTA3D with Interactive Features
+
+### 7. Full Interactive Setup
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models vista3d \
+  --conf vista3d_prompting true \
+  --conf vista3d_point_prompts true \
+  --conf vista3d_class_prompts true \
+  --conf preload true
+```
+
+## Development and Testing
+
+### 8. Debug Configuration
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models vista3d \
+  --conf debug true \
+  --conf vista3d_cache_transforms false \
+  --log_level DEBUG
+```
+
+### 9. Performance Profiling
+```bash
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models vista3d \
+  --conf profile true \
+  --conf vista3d_profile_inference true
+```
+
+## Docker Deployment
+
+### 10. Docker with VISTA3D
+```bash
+docker run --gpus all --rm -ti \
+  -p 8000:8000 \
+  -v /data/ct_volumes:/data/studies \
+  -v /data/models:/data/models \
+  projectmonai/monailabel:latest \
+  monailabel start_server \
+    --app /opt/monailabel/sample-apps/monaibundle \
+    --studies /data/studies \
+    --conf models vista3d \
+    --conf preload true \
+    --host 0.0.0.0
+```
+
+## Configuration Parameters
+
+| Parameter | Description | Default | Example |
+|-----------|-------------|---------|---------|
+| `vista3d_patch_size` | Input patch size for inference | `[96,96,96]` | `"[64,64,64]"` |
+| `vista3d_overlap` | Sliding window overlap ratio | `0.25` | `0.5` |
+| `vista3d_prompting` | Enable interactive prompting | `false` | `true` |
+| `vista3d_point_prompts` | Enable point-based prompting | `true` | `false` |
+| `vista3d_class_prompts` | Enable class-based prompting | `true` | `false` |
+| `vista3d_cpu_fallback` | Use CPU when GPU memory insufficient | `false` | `true` |
+| `vista3d_cache_transforms` | Cache preprocessing transforms | `true` | `false` |
+| `vista3d_profile_inference` | Profile inference performance | `false` | `true` |
+
+## Environment Variables
+
+```bash
+# Set GPU memory fraction
+export CUDA_VISIBLE_DEVICES=0
+export PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:512
+
+# Set cache directories
+export MONAI_CACHE_DIR=/data/cache
+export VISTA3D_CACHE_DIR=/data/vista3d_cache
+
+# Start server with environment
+monailabel start_server \
+  --app apps/monaibundle \
+  --studies datasets/ct_volumes \
+  --conf models vista3d
+```
+
+## Troubleshooting Common Configurations
+
+### GPU Memory Issues
+```bash
+# Reduce memory usage
+--conf vista3d_patch_size "[48,48,48]"
+--conf vista3d_overlap 0.125
+--conf vista3d_cpu_fallback true
+```
+
+### Network/Download Issues
+```bash
+# Use local model cache
+--conf zoo_source local
+--conf model_path /local/path/to/vista3d
+
+# Use alternative zoo repo
+--conf zoo_repo "custom-repo/model-zoo"
+```
+
+### Performance Optimization
+```bash
+# Optimize for speed
+--conf vista3d_patch_size "[128,128,128]"
+--conf vista3d_overlap 0.5
+--conf workers 8
+--conf preload true
+```