Brain MRI FLAIR Segmentation

A Complete MLOPS project on Brain MRI FLAIR Segmentation. A MobileNet v3 based segmentation project to perform instance segmentation on FLAIR (Fluid-Attenuated Inversion Recovery) abnormality in brain MRI images. This model is trained using Brain MRI segmentation from kaggle and is deployed using Streamlit.

Demo

Try it yourself here

Project Organization

---

├── LICENSE
├── README.md               <- Documentation to get more information about the project.
├── data
│   |
│   ├── processed           <- The final, canonical data sets for modeling.
│   └── raw                 <- The original, immutable data dump.
│
├── saved_models            <- Trained and serialized models.
│
├── inference               <- Code for inference.
│
├── report                  <- metrics and logs from training.
├── images                  <- Generated graphics and figures to be used in reporting
│
├── full-requirements.txt   <- The requirements file for reproducing the analysis environment.
|
├── requirements.txt        <- The requirements file for reproducing the deployment environment.
│
├── dvc.yaml                <- dvc.yaml is used to define dvc pipelines.
│
├── config.yaml             <- Contains all the parameters for training.
│
├── src                     <- Source code for use in this project.
│   ├── __init__.py         <- Makes src a Python module.
│   │
|   ├── utils.py            <- Script that contains utility functions.
|   |
|   ├── earl_stopping.py    <- Script that contains early stopping callback.
|   |
│   ├── data                <- Scripts to download or generate data.
│   │   └── preprocess_data.py
|   |   |
│   │   └── dataset.py
│   │
│   ├── models              <- Scripts to train models and optimize graphs.
│       ├── train.py
│       └── optimize-graph.py
│
├── tests                   <- Unit test code.
│   ├── __init__.py         <- Makes tests a Python module.
│   ├── config_test.py      <- tests the config.yaml file.
|
└── Procfile                <- Procfile is used for deployment in Heroku.
│
└── setup.sh                <- setup.sh  is used for deployment in Heroku.
│
└── runtime.txt             <- runtime.txt is used to specify the python runtime version in Heroku.
│
└── tox.ini                 <- tox file with settings for running tox; see tox.readthedocs.io.

Running on native machine

dependencies

• python3

pip packages

pip install -r requirements.txt

Steps to train your own model

Scripts

src/train.py - is used to train the model
inference/engine.py - is used to perform inference
inference/ui.py - is used to build the streamlit web application