# NEATmap documentation

### 1. Abstract & Availability Quantitative analysis of activated neurons in mice brains by a specific stimulation is usually a primary step to locate the responsive neurons throughout the brain. However, it’s challenging to comprehensively and consistently analyze the neuronal activity trace in whole brains of large cohort of mice from many Terabytes of volumetric imaging data. Here, we introduce **NEATmap**, a deep-learning based high-efficiency, high-precision, and user-friendly software for whole brain **NE**uronal **A**ctivity **T**race **map**ping by automated segmentation and quantitative analysis of immunofluorescence labeled c-Fos+ neurons. We applied NEATmap to study the brain-wide differentiated neuronal activation in response to physical and psychological stressors in cohorts of mice.

Availability and implementation: The source code for all modules of NEATmap is implemented in Python. Code source, tutorial, documentation are available at: https://github.com/mesobrain/NEATmap_code. Whole-brain test dataset is available at: https://zenodo.org/record/8133486. The trained neural network parameters can be downloaded from this [link](https://drive.google.com/drive/u/0/folders/1lPJaprmVCawnKinYXb5Wa_orkfEniNY4).
### 2. Installation $ Running The list of names of python libraries required by NEATmap can be found in the environment file `NEATmap_software.yaml`. $ conda env create -f NEATmap_software.yaml Similarly, you can install other python dependencies via pip and use $ pip install name Running: After completing the above installation, execute the command to enter NEATmap. $ python NEATmap_ui.py
### 3. Settings Click the **Settings** button to display the parameter setting interface. If the user needs to modify the parameters, make the changes and click the **Save** button.

### 4. All-in-one NEATmap pipeline Click the **All-in-one NEATmap pipeline** button to effortlessly complete the entire process, from data preprocessing to cell counting.

### 5. Transfer learning Click the **Transfer learning** button to fine-tune the 3D-HSFormer model. This operation enables the application of NEATmap to whole-brain volumertric datasets of various cell types.

### 6. Data preprocessing Clicking the **Data preprocessing** button enables the generation of both volumetric images and patch images (sub-volumes).

#### Diagram representation:

### 7. Whole brain segmentation Clicking on **Whole brain segmentation** enables automatic segmentation of immunolabeled signal (c-Fos) across the whole brain.

#### 3D-HSFormer architecture:

### 8. Whole brain reassemble & Post processing Clicking on **Whole brain reassemble** followed by **Post processing** results in the segmented map of immunolabeled signal (c-Fos) across the whole brain.

#### Diagram representation:

### 9. Registration & Cell counting Clicking on **Registration** followed by **Cell counting** enables the registration of the whole brain to the [Allen Common Coordinate Framework atlas](https://atlas.brain-map.org/) and provides cell counts results at all levels of whole-brain.

#### Results:

## Supplementary instruction Detailed instructions for specific operations can be found in the **User_guide.pdf** file located in the **doc** directory. The usage tutorial for the NEATmap software can be found in the **Tutorial_video** folder under the file name **neatmap_tutorial.mp4**. Additionally, the **seg-display.mp4** file demonstrates the segmentation and cell counting results of NEATmap. ### Update The video titled **User_friendly_tutorial** in the **Tutorial_video** folder demonstrates All-in-one NEATmap pipeline. The **transfer_learning_tutorial** video demonstrates how to fine-tune a deep learning model. This operation can also be applied to whole-brain datasets of other cell types using NEATmap. ## Reference Cite us if you use the software in any form: @article{zheng2024neatmap, title={NEATmap: a high-efficiency deep learning approach for whole mouse brain neuronal activity trace mapping}, author={Zheng, Weijie and Mu, Huawei and Chen, Zhiyi and Liu, Jiajun and Xia, Debin and Cheng, Yuxiao and Jing, Qi and Lau, Pak-Ming and Tang, Jin and Bi, Guo-Qiang and Wu, Feng and Wang, Hao}, journal={National Science Review}, pages={nwae109}, year={2024}, publisher={Oxford University Press} }