I developed a computational pipeline for Accurate Mouse Brain Image Analysis.

This project included computer vision tasks such as cell detection, and image registration. The pipeline is deep learning based including a regression CNN model, segmentation model, and Image registration, and involved developing a GUI using PyQt to encapsulate all these.

Code on my Github: https://github.com/mrymsadeghi/AMBIA

Publications:

• Sadeghi, M., Neto, P., Ramos-Prats, A., Castaldi, F., Paradiso, E., Mahmoodian, N., ... & Goebel, G. (2022, April). Automatic 2D to 3D localization of histological mouse brain sections in the reference atlas using deep learning. In Medical Imaging 2022: Image Processing (Vol. 12032, pp. 718-724). SPIE.

I developed pipeline to locatalize a 2D histological mouse brain image in the 3D reference atlas of the brain using regression CNN models.

Github: https://github.com/mrymsadeghi/AMBIA

Publications:

• Sadeghi, M., Neto, P., Ramos-Prats, A., Castaldi, F., Paradiso, E., Mahmoodian, N., ... & Goebel, G. (2022, April). Automatic 2D to 3D localization of histological mouse brain sections in the reference atlas using deep learning. In Medical Imaging 2022: Image Processing (Vol. 12032, pp. 718-724). SPIE.

Visualizing image embeddings using Tensorboard in classification of the birds dataset

Dimentionality reduction using t-SNE and UMAP

Code on my Github: https://github.com/mrymsadeghi/Colab_notebooks/blob/main/Embeddings_Img_similarity_birds.ipynb

This project we created a pipeline to analys whole slide images of H&E stained renal slides. 

The pipelien inclded DL algorithms to detect and segment Glomeruli structures, and classify them into four categories. We also used XAI techniques such as GradCAM to shed light on the decision making criteria of our algorithm

This project started as an effort to develope a classification algorithm for breast cancer metastasis in histopathological wole slide images. In the second phase of the project I decided to make the algorithm learn actively from the patholgists by getting their feedback. The GUI was developed in Python and PyQt4. This was my Master thesis project

Publication:

Sadeghi, M., Maldonado, I., Abele, N., Haybaeck, J., Boese, A., Poudel, P., & Friebe, M. (2019, July). Feedback-based self-improving CNN algorithm for breast cancer lymph node metastasis detection in real clinical environment. In 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 7212-7215). IEEE. 

Stable diffusion integrated in a tkinter Gui to evaluate the guidance score parameter

Code on my Github: https://github.com/mrymsadeghi/Stable_diffusion_with_gui