<h2 align="center">Automatic Identification and Counting of Blood Cells<h2>

[![GitHub stars](https://img.shields.io/github/stars/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells)](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells/stargazers)

[![GitHub forks](https://img.shields.io/github/forks/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells)](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells/network)

[![GitHub issues](https://img.shields.io/github/issues/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells)](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells/issues)

[![Version](https://img.shields.io/badge/version-2.1-green.svg?longCache=true&style=flat)](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells)

[![GitHub license](https://img.shields.io/github/license/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells)](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells/blob/master/LICENSE)

<!--<img src="https://user-images.githubusercontent.com/37298971/123714340-f8d70800-d82a-11eb-9742-042a5d9334a1.png" width="28">-->

## Dataset

The [```Complete Blood Count (CBC) Dataset```](https://github.com/MahmudulAlam/Complete-Blood-Cell-Count-Dataset) has

been used for automatic identification and counting of blood cells. Download the dataset, unzip and put

the ```Training```, ```Testing```, and ```Validation```folders in the working directory.

## Requirements

![requirements](https://img.shields.io/badge/Python-3.6-3480eb.svg?longCache=true&style=flat&logo=python)

![requirements](https://img.shields.io/badge/Python-3.7-3480eb.svg?longCache=true&style=flat&logo=python)

- Tensorflow-GPU==2.2.0 (tested on 2.1.0, 2.2.0, and 2.3.0) ```conda install tensorflow-gpu```

- TF-slim==1.1.0 ```pip install tf-slim==1.1.0```

- Weights: [```download```](https://1drv.ms/u/s!AlXVRhh1rUKThlxTievX0X1CpXd0?e=9cKxYb) the trained weights file for

  blood cell detection and put the ```weights``` folder in the working directory.

[![Download](https://img.shields.io/badge/download-weights-blue.svg?longCache=true&style=flat&logo=microsoft-onedrive)](https://1drv.ms/u/s!AlXVRhh1rUKThlxTievX0X1CpXd0?e=9cKxYb)

[![Download](https://img.shields.io/badge/download-weights-ff160a.svg?longCache=true&style=flat&logo=mega)](https://mega.nz/#F!2kVUnKjS!z15tM9WLfga3l1gCNSLNGw)

## Getting Started 

1. Build the cython extension in place 

```python setup.py build_ext --inplace```

2. Run detect.py 

```python detect.py```

## Update

The ```darkflow.cython_utils.cy_yolo_findboxes``` problem has been fixed. Make sure to build the cython extension in place before running the code. 

[![Paper](https://img.shields.io/badge/TensorFlow-2.x-f57418.svg?longCache=true&style=flat&logo=tensorflow)](https://www.tensorflow.org/install)

The code was originally written and developed with `TensorFlow v1.x`. The new updated version `v2.0`

included `TensorFlow v2.x` support, tested on both TensorFlow `v2.1.0` and `v2.2.0`. You can download the previous

version

from **[`here`](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells/releases/tag/v1.0)**

.

## How to Run the Code  :runner:

To detect the blood cells, simply run the `detect.py` file in the terminal or use an IDE. A step-by-step guideline of

how to run the blood cell detection code in your computer is provided in

this **[`wiki`](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells/wiki/A-Step-by-Step-Guide-of-How-to-Run-the-Code)**

.

If you have any trouble running the code and facing any errors please feel free to create

an **[`issue`](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells/issues)**

or **[`contact me`](https://mahmudulalam.github.io/#contact)**.

## How to Train on Your Dataset  :bullettrain_side:

A seven-step guideline of how to train on your own dataset is provided in

this **[`wiki`](https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells/wiki/How-to-Train-on-Your-Dataset)**

.

## Paper

[![Paper](https://img.shields.io/badge/paper-IETDigiLib-830ceb.svg?longCache=true&style=flat)](http://ietdl.org/t/kmgztb) [![Paper](https://img.shields.io/badge/paper-Wiley-282829.svg?longCache=true&style=flat)](https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/htl.2018.5098)

The code was developed for the following blood cell detection paper. For a more detailed explanation of the proposed

method, please go through the pdf of the [```paper```][1]. If you use this code or associated dataset, please cite this

paper as:

[***```Machine learning approach of automatic identification and counting of blood cells```***](https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/htl.2018.5098)

```bibtex

@article{alam2019machine,

  title={Machine learning approach of automatic identification and counting of blood cells},

  author={Alam, Mohammad Mahmudul and Islam, Mohammad Tariqul},

  journal={Healthcare Technology Letters},

  volume={6},

  number={4},

  pages={103--108},

  year={2019},

  publisher={IET}

}

```

## Blood Cell Detection Output

<p align="center">

  <img src="https://user-images.githubusercontent.com/37298971/44617785-17eb0980-a88b-11e8-9018-c84f8be5cefa.png" width="500">

</p>

## KNN and IOU Based Verification

In some cases, our model predicts the same platelet twice. To solve this problem we propose a k-nearest neighbor (KNN)

and intersection over union (IOU) based verification system where we find the nearest platelet of a selected platelet

and calculate their overlap. We are allowing only a 10% overlap between two platelets. If the overlap is more than that

then it will be a spurious prediction and we will ignore the prediction.

| Before Verification  | After Verification  |

|:-:|:-:|

| <p align="center"> <img src="https://user-images.githubusercontent.com/37298971/46260207-b27ede00-c504-11e8-9d00-7d7be151ee5d.jpg"> </p>  | <p align="center"> <img src="https://user-images.githubusercontent.com/37298971/46260504-a268fd80-c508-11e8-9ef0-5230d00f47a3.jpg">  |

## Prediction on High-Resolution Image (HRI)

We have used our model to detect and count blood cells from high-resolution blood cell smear images. These test images

are of the size of ```3872 x 2592``` way higher than the size of our trained images of ```640 x 480```. So, to match the

cell size of our trained images we divide those images into grid cells and run prediction in each grid cell and then

combine all the prediction results.

<h3 align="center">Dividing Image into Grid/Patch</h3>

<p align="center">

  <img src="https://user-images.githubusercontent.com/37298971/45962420-a39ab600-c042-11e8-975f-9b0a077f0e0f.jpg" width="700">

</p>

<h3 align="center">Combined Output</h3>

<p align="center">

  <img src="https://user-images.githubusercontent.com/37298971/45961699-055a2080-c041-11e8-95b0-1c8ac3c8875b.jpg" width="700">

</p>

[1]: https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/htl.2018.5098