# gregorator/medical-image-enhancement-system architecture

> Architecture diagram of the GitHub repository gregorator/medical-image-enhancement-system, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-21). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/gregorator/medical-image-enhancement-system
- Repository: https://github.com/gregorator/medical-image-enhancement-system

## Overview

This is a CT-image denoising product spanning NetCDF dataset preparation, Noise2Noise training, ONNX deployment, and an interactive Streamlit studio backed by a FastAPI service. The primary user workflow is to inspect or batch-submit images in the UI, which calls the API for denoising; users can also switch inference configuration and request heatmaps or unload the model. Training and dataset preparation are separate workflows. Some links rely on README descriptions or the source index because those implementation excerpts were not provided.

Main entry points: [`scripts/train.py`](https://github.com/gregorator/medical-image-enhancement-system/blob/master/scripts/train.py) (Training entry), [`onnx/onnx_inference.py`](https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/onnx_inference.py) (Inference manager), [`app/utils/api_client.py`](https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/utils/api_client.py) (API client), [`src/utils/factories.py`](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/utils/factories.py) (Training factories)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_studio["Streamlit Studio"]
  node_dashboard["Dashboard<br/>[0_Dashboard.py]"]
  node_inspector["Image Inspector"]
  node_batch["Batch Processor"]
  node_navigation["Page routing<br/>[routes.py]"]
  node_sidebar["Model controls<br/>[sidebar.py]"]
  node_api_client["API client<br/>[api_client.py]"]
end

subgraph group_service["Inference Service"]
  node_api_server["REST API<br/>[api_server.py]"]
end

subgraph group_data["Dataset Pipeline"]
  node_dataset_build["Dataset preparation<br/>[make_dataset.py]"]
  node_dataset_prep["NetCDF conversion<br/>[preparation.py]"]
  node_dataset["Image datasets<br/>[dataset.py]"]
end

subgraph group_training["Training"]
  node_train_entry["Training entry<br/>[train.py]"]
  node_config["Run configuration<br/>[config.py]"]
  node_factories["Training factories<br/>[factories.py]"]
  node_noise["Noise transforms<br/>[noise.py]"]
  node_model["Noise2Noise U-Net<br/>[noise2noise.py]"]
  node_trainer["N2N trainer<br/>[n2n.py]"]
  node_engine["Training engine<br/>[engine.py]"]
  node_metrics["Evaluation metrics<br/>[metrics.py]"]
  node_logger["Experiment logger<br/>[logger.py]"]
end

subgraph group_deploy["ONNX Deployment"]
  node_manager["Inference manager<br/>[onnx_inference.py]"]
  node_gpu["GPU tiling<br/>[inference_gpu.py]"]
  node_cpu["CPU inference<br/>[inference_cpu.py]"]
  node_onnx_wrap["ONNX wrapping<br/>[onnx_wrappers.py]"]
  node_export["ONNX export<br/>[export_onnx.py]"]
  node_onnx_models[("Bundled ONNX models")]
end

node_user(("Image user"))

node_user -->|"opens studio"| node_dashboard
node_user -->|"inspects images"| node_inspector
node_user -->|"submits batches"| node_batch
node_dashboard -->|"registered by"| node_navigation
node_inspector -->|"requests inference"| node_api_client
node_batch -->|"requests inference"| node_api_client
node_sidebar -->|"updates settings"| node_api_client
node_sidebar -->|"uses pages"| node_navigation
node_api_client -->|"calls REST endpoints"| node_api_server
node_api_server -->|"dispatches inference"| node_manager
node_manager -->|"selects CUDA backend"| node_gpu
node_manager -->|"selects CPU backend"| node_cpu
node_gpu -->|"uses wrapper"| node_onnx_wrap
node_dataset_build -->|"converts blocks"| node_dataset_prep
node_train_entry -->|"loads settings"| node_config
node_train_entry -->|"loads samples"| node_dataset
node_train_entry -->|"builds components"| node_factories
node_factories -->|"selects transforms"| node_noise
node_factories -->|"builds models"| node_model
node_train_entry -->|"starts training"| node_trainer
node_trainer -->|"extends engine"| node_engine
node_trainer -->|"computes evaluation"| node_metrics
node_engine -->|"logs progress"| node_logger
node_export -->|"exports network"| node_model
node_export -->|"writes model"| node_onnx_models
node_manager -->|"loads model"| node_onnx_models

click node_dashboard "https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/pages/0_Dashboard.py"
click node_inspector "https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/pages/1_Image_Inspector.py"
click node_batch "https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/pages/2_Batch_Processor.py"
click node_navigation "https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/routes.py"
click node_sidebar "https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/components/sidebar.py"
click node_api_client "https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/utils/api_client.py"
click node_api_server "https://github.com/gregorator/medical-image-enhancement-system/blob/master/api/api_server.py"
click node_manager "https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/onnx_inference.py"
click node_gpu "https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/inference_gpu.py"
click node_cpu "https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/inference_cpu.py"
click node_onnx_wrap "https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/onnx_wrappers.py"
click node_dataset_build "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/datasets/make_dataset.py"
click node_dataset_prep "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/datasets/preparation.py"
click node_dataset "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/datasets/dataset.py"
click node_train_entry "https://github.com/gregorator/medical-image-enhancement-system/blob/master/scripts/train.py"
click node_config "https://github.com/gregorator/medical-image-enhancement-system/blob/master/configs/config.py"
click node_factories "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/utils/factories.py"
click node_noise "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/transforms/noise.py"
click node_model "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/models/noise2noise.py"
click node_trainer "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/trainers/n2n.py"
click node_engine "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/trainers/engine.py"
click node_metrics "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/evaluation/metrics.py"
click node_logger "https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/utils/logger.py"
click node_export "https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/export_onnx.py"
click node_onnx_models "https://github.com/gregorator/medical-image-enhancement-system/tree/master/onnx/models"

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_dashboard,node_inspector,node_batch,node_navigation,node_sidebar,node_api_client,node_user toneBlue
class node_api_server toneAmber
class node_dataset_build,node_dataset_prep,node_dataset toneMint
class node_train_entry,node_config,node_factories,node_noise,node_model,node_trainer,node_engine,node_metrics,node_logger toneRose
class node_manager,node_gpu,node_cpu,node_onnx_wrap,node_export,node_onnx_models toneIndigo
```

## Components

### Streamlit Studio

- [Dashboard](https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/pages/0_Dashboard.py): `app/pages/0_Dashboard.py`
- [Image Inspector](https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/pages/1_Image_Inspector.py): `app/pages/1_Image_Inspector.py`
- [Batch Processor](https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/pages/2_Batch_Processor.py): `app/pages/2_Batch_Processor.py`
- [Page routing](https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/routes.py): `app/routes.py`
- [Model controls](https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/components/sidebar.py): `app/components/sidebar.py`
- [API client](https://github.com/gregorator/medical-image-enhancement-system/blob/master/app/utils/api_client.py): `app/utils/api_client.py`

### Inference Service

- [REST API](https://github.com/gregorator/medical-image-enhancement-system/blob/master/api/api_server.py): `api/api_server.py`

### Dataset Pipeline

- [Dataset preparation](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/datasets/make_dataset.py): `src/datasets/make_dataset.py`
- [NetCDF conversion](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/datasets/preparation.py): `src/datasets/preparation.py`
- [Image datasets](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/datasets/dataset.py): `src/datasets/dataset.py`

### Training

- [Training entry](https://github.com/gregorator/medical-image-enhancement-system/blob/master/scripts/train.py): `scripts/train.py`
- [Run configuration](https://github.com/gregorator/medical-image-enhancement-system/blob/master/configs/config.py): `configs/config.py`
- [Training factories](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/utils/factories.py): `src/utils/factories.py`
- [Noise transforms](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/transforms/noise.py): `src/transforms/noise.py`
- [Noise2Noise U-Net](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/models/noise2noise.py): `src/models/noise2noise.py`
- [N2N trainer](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/trainers/n2n.py): `src/trainers/n2n.py`
- [Training engine](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/trainers/engine.py): `src/trainers/engine.py`
- [Evaluation metrics](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/evaluation/metrics.py): `src/evaluation/metrics.py`
- [Experiment logger](https://github.com/gregorator/medical-image-enhancement-system/blob/master/src/utils/logger.py): `src/utils/logger.py`

### ONNX Deployment

- [Inference manager](https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/onnx_inference.py): `onnx/onnx_inference.py`
- [GPU tiling](https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/inference_gpu.py): `onnx/inference_gpu.py`
- [CPU inference](https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/inference_cpu.py): `onnx/inference_cpu.py`
- [ONNX wrapping](https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/onnx_wrappers.py): `onnx/onnx_wrappers.py`
- [ONNX export](https://github.com/gregorator/medical-image-enhancement-system/blob/master/onnx/export_onnx.py): `onnx/export_onnx.py`
- [Bundled ONNX models](https://github.com/gregorator/medical-image-enhancement-system/tree/master/onnx/models): `onnx/models`

### Other components

- **Image user**

## Connections

- Image user → Dashboard: opens studio (evidence: `README.md`)
- Image user → Image Inspector: inspects images (evidence: `README.md`)
- Image user → Batch Processor: submits batches (evidence: `README.md`)
- Dashboard → Page routing: registered by (evidence: `app/pages/0_Dashboard.py`)
- Image Inspector → API client: requests inference (evidence: `app/pages/1_Image_Inspector.py`)
- Batch Processor → API client: requests inference (evidence: `app/pages/2_Batch_Processor.py`)
- Model controls → API client: updates settings (evidence: `app/components/sidebar.py`)
- Model controls → Page routing: uses pages (evidence: `app/components/sidebar.py`)
- API client → REST API: calls REST endpoints (evidence: `README.md`)
- REST API → Inference manager: dispatches inference (evidence: `api/api_server.py`)
- Inference manager → GPU tiling: selects CUDA backend (evidence: `onnx/onnx_inference.py`)
- Inference manager → CPU inference: selects CPU backend (evidence: `onnx/onnx_inference.py`)
- GPU tiling → ONNX wrapping: uses wrapper (evidence: `onnx/inference_gpu.py`)
- Dataset preparation → NetCDF conversion: converts blocks (evidence: `src/datasets/make_dataset.py`)
- Training entry → Run configuration: loads settings (evidence: `scripts/train.py`)
- Training entry → Image datasets: loads samples (evidence: `scripts/train.py`)
- Training entry → Training factories: builds components (evidence: `scripts/train.py`)
- Training factories → Noise transforms: selects transforms (evidence: `src/utils/factories.py`)
- Training factories → Noise2Noise U-Net: builds models (evidence: `src/utils/factories.py`)
- Training entry → N2N trainer: starts training (evidence: `scripts/train.py`)
- N2N trainer → Training engine: extends engine (evidence: `src/trainers/n2n.py`)
- N2N trainer → Evaluation metrics: computes evaluation (evidence: `src/trainers/n2n.py`)
- Training engine → Experiment logger: logs progress (evidence: `src/trainers/engine.py`)
- ONNX export → Noise2Noise U-Net: exports network (evidence: `onnx/export_onnx.py`)
- ONNX export → Bundled ONNX models: writes model (evidence: `README.md`)
- Inference manager → Bundled ONNX models: loads model (evidence: `README.md`)
