Add k-space basics tutorial for MRI reconstruction

README.md

@@ -386,4 +386,11 @@ Example shows the use cases of using MONAI to evaluate the performance of a gene
 
 #### [VISTA2D](./vista_2d)
 This tutorial demonstrates how to train a cell segmentation model using the [MONAI](https://monai.io/) framework and the [Segment Anything Model (SAM)](https://github.com/facebookresearch/segment-anything) on the [Cellpose dataset](https://www.cellpose.org/).
-ECHO°¡ ¼³Á¤µÇ¾î ÀÖ½À´Ï´Ù.
+
+#### <ins>**Reconstruction**</ins>
+##### [K-Space Basics with fastMRI Knee Data](./reconstruction/MRI_reconstruction/tutorials/01_kspace_basics_fastmri_knee.ipynb)
+This tutorial introduces MRI reconstruction fundamentals: what k-space is, how the Fourier transform connects k-space to images, why undersampling causes aliasing, and how MONAI's reconstruction transforms process k-space data. Uses the fastMRI knee single-coil dataset.
+##### [U-Net MRI Reconstruction](./reconstruction/MRI_reconstruction/unet_demo)
+Training and inference for accelerated MRI reconstruction using BasicUNet on the fastMRI brain multi-coil dataset.
+##### [VarNet MRI Reconstruction](./reconstruction/MRI_reconstruction/varnet_demo)
+Training and inference for accelerated MRI reconstruction using e2e-VarNet on the fastMRI brain multi-coil dataset.

reconstruction/MRI_reconstruction/tutorials/README.md

@@ -0,0 +1,25 @@
+# MRI Reconstruction Tutorials
+
+This folder contains educational Jupyter notebooks that introduce the fundamentals of MRI reconstruction.
+
+## Tutorials
+
+### 01 - K-Space Basics with fastMRI Knee Data
+**Notebook:** [01_kspace_basics_fastmri_knee.ipynb](./01_kspace_basics_fastmri_knee.ipynb)
+
+An introductory tutorial covering:
+- What k-space is and its relationship to MRI images via the Fourier transform
+- How low and high spatial frequencies contribute to image content
+- Why undersampling k-space causes aliasing artifacts
+- How MONAI's reconstruction transforms (`RandomKspaceMaskd`, `EquispacedKspaceMaskd`, etc.) process k-space data
+- The zero-filled reconstruction problem that deep learning methods aim to solve
+
+**Dataset:** [fastMRI](https://fastmri.org/dataset) knee single-coil validation set (requires registration, non-commercial license). Only one `.h5` file is needed.
+
+**Prerequisites:** Basic Python and NumPy. No MRI experience required.
+
+## Related Production Tutorials
+
+For training-focused tutorials using the brain multi-coil dataset, see:
+- [U-Net Demo](../unet_demo/) - BasicUNet for MRI reconstruction
+- [VarNet Demo](../varnet_demo/) - End-to-end Variational Network for MRI reconstruction

runner.sh

@@ -83,6 +83,7 @@ doesnt_contain_max_epochs=("${doesnt_contain_max_epochs[@]}" maisi_inference_tut
 doesnt_contain_max_epochs=("${doesnt_contain_max_epochs[@]}" realism_diversity_metrics.ipynb)
 doesnt_contain_max_epochs=("${doesnt_contain_max_epochs[@]}" omniverse_integration.ipynb)
 doesnt_contain_max_epochs=("${doesnt_contain_max_epochs[@]}" hugging_face_pipeline_for_monai.ipynb)
+doesnt_contain_max_epochs=("${doesnt_contain_max_epochs[@]}" 01_kspace_basics_fastmri_knee.ipynb)
 
 # Execution of the notebook in these folders / with the filename cannot be automated
 skip_run_papermill=()