# joseberlin7/3d-mri-segmentation architecture

> Architecture diagram of the GitHub repository joseberlin7/3d-mri-segmentation, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-01). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/joseberlin7/3d-mri-segmentation
- Repository: https://github.com/joseberlin7/3d-mri-segmentation

## Overview

This repository implements a BraTS 3D medical-image segmentation workflow: subject volumes are loaded and transformed, then a SwinUNETR model is trained and evaluated, with metrics and checkpoints produced. Azure ML submission is an optional execution path; the README also identifies notebooks for metric and visualization review. The main call relationships are supported by the training and submission excerpts; notebook internals were not sampled.

Main entry points: [`src/train.py`](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py) (Training entry), [`src/data_loaders.py`](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/data_loaders.py) (Data loaders), [`src/model_code.py`](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/model_code.py) (SwinUNETR), [`src/submit_job.py`](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/submit_job.py) (Azure submission)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_data["Data preparation"]
  node_dataset["BraTS dataset<br/>[data_loaders.py]"]
  node_transforms["Transforms<br/>[data_loaders.py]"]
  node_loaders["Data loaders<br/>[data_loaders.py]"]
end

subgraph group_learning["Model training"]
  node_entry["Training entry<br/>[train.py]"]
  node_model["SwinUNETR<br/>[model_code.py]"]
  node_loop["Train loop<br/>[train.py]"]
  node_evaluation["Validation metrics<br/>[train.py]"]
end

subgraph group_operations["Monitoring and execution"]
  node_logging["TensorBoard logs<br/>[train.py]"]
  node_checkpoints["Model checkpoints<br/>[train.py]"]
  node_submission["Azure submission<br/>[submit_job.py]"]
end

subgraph group_review["Review"]
  node_notebooks["Metric visualization<br/>[dev_note.ipynb]"]
end

node_user(("Researcher"))
node_brats[("BraTS volumes")]
node_azure["Azure ML"]

node_user -->|"starts training"| node_entry
node_brats -->|"provides volumes"| node_dataset
node_dataset -->|"transforms samples"| node_transforms
node_transforms -->|"feeds batches"| node_loaders
node_entry -->|"requests loaders"| node_loaders
node_entry -->|"creates model"| node_model
node_entry -->|"calls training"| node_loop
node_loaders -->|"supplies batches"| node_loop
node_model -->|"trains model"| node_loop
node_loop -->|"validates periodically"| node_evaluation
node_loop -->|"writes scalars"| node_logging
node_evaluation -->|"saves weights"| node_checkpoints
node_user -->|"requests job"| node_submission
node_submission -->|"submits job"| node_azure
node_azure -->|"runs training command"| node_entry
node_logging -.->|"metrics reviewed"| node_notebooks

click node_dataset "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/data_loaders.py"
click node_transforms "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/data_loaders.py"
click node_loaders "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/data_loaders.py"
click node_entry "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py"
click node_model "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/model_code.py"
click node_loop "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py"
click node_evaluation "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py"
click node_logging "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py"
click node_checkpoints "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py"
click node_submission "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/submit_job.py"
click node_notebooks "https://github.com/joseberlin7/3d-mri-segmentation/blob/main/Notebooks/dev_note.ipynb"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_dataset,node_transforms,node_loaders toneBlue
class node_entry,node_model,node_loop,node_evaluation,node_brats toneAmber
class node_logging,node_checkpoints,node_submission toneMint
class node_notebooks toneRose
class node_user,node_azure toneIndigo
```

## Components

### Data preparation

- [BraTS dataset](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/data_loaders.py): `src/data_loaders.py`
- [Transforms](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/data_loaders.py): `src/data_loaders.py`
- [Data loaders](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/data_loaders.py): `src/data_loaders.py`

### Model training

- [Training entry](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py): `src/train.py`
- [SwinUNETR](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/model_code.py): `src/model_code.py`
- [Train loop](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py): `src/train.py`
- [Validation metrics](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py): `src/train.py`

### Monitoring and execution

- [TensorBoard logs](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py): `src/train.py`
- [Model checkpoints](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/train.py): `src/train.py`
- [Azure submission](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/src/submit_job.py): `src/submit_job.py`

### Review

- [Metric visualization](https://github.com/joseberlin7/3d-mri-segmentation/blob/main/Notebooks/dev_note.ipynb): `Notebooks/dev_note.ipynb`

### Other components

- **Researcher**
- **BraTS volumes**
- **Azure ML**

## Connections

- Researcher → Training entry: starts training (evidence: `README.md`)
- BraTS volumes → BraTS dataset: provides volumes (evidence: `src/data_loaders.py`)
- BraTS dataset → Transforms: transforms samples (evidence: `src/data_loaders.py`)
- Transforms → Data loaders: feeds batches (evidence: `src/data_loaders.py`)
- Training entry → Data loaders: requests loaders (evidence: `src/train.py`)
- Training entry → SwinUNETR: creates model (evidence: `src/train.py`)
- Training entry → Train loop: calls training (evidence: `src/train.py`)
- Data loaders → Train loop: supplies batches (evidence: `src/train.py`)
- SwinUNETR → Train loop: trains model (evidence: `src/train.py`)
- Train loop → Validation metrics: validates periodically (evidence: `src/train.py`)
- Train loop → TensorBoard logs: writes scalars (evidence: `src/train.py`)
- Validation metrics → Model checkpoints: saves weights (evidence: `src/train.py`)
- Researcher → Azure submission: requests job (evidence: `README.md`)
- Azure submission → Azure ML: submits job (evidence: `src/submit_job.py`)
- Azure ML → Training entry: runs training command
- TensorBoard logs → Metric visualization: metrics reviewed (evidence: `README.md`)
