About
I am a postdoctoral researcher in Wellcome Sanger Institute and University of Cambridge. I’m interested in machine learning for biomedical data, causal modelling and inference, and interpretability for biomedical discovery.
I was a PhD student at University of Alberta, where I was supervised by professor Nilanjan Ray from Computing Science department
and professor Gilbert Bigras from cross-cancer institue (CCI) and UofA department of Pathology.
Before that I recieved my B.Sc. and M.Sc. in computer engineering from Sharif University of Technology.
Email: ah8 at ualberta dot ca, aa36 at sanger dot ac dot uk
Office: 111 Athabasca Hall.
Research
In the past, cancer diagnosis was based on viewing tissues/cells under microscope. Recently, making primary diagnosis from scanned/digitized pathology images has been officially approved. Apart from its clinical benefits, digital pathology provided machine learning researchers with abundant of data.
My research is about developing/adopting machine learning methods for digitized pathology images.
Code
- GPEX is a tool that enables min-batch training of Gaussian processes thereby making it scalable to large, e.g., image datasets and with GPU acceleration. Moreover, given an artificial neural network it finds equivalent Gaussian processes whereby the artificial neural network can be unboxed.
GPEX documentation: https://gpex.readthedocs.io/en/latest/
- We’ve developed a tool we called PyDmed to facilitate developing machine learning methods for whole-slide images and medical images in general. If you work on ML + medical imaging, PyDmed is really a life changer (research changer :) ).
You can check out PyDmed’s repository on github.
Publications
-
N. Guruprasad, A. Akbarnejad, PJ Barnes, G. Bigras, and N. Ray, “A Closer Look at Weak Supervision’s Limitations in WSI Recurrence Score Prediction”, 2023 IEEE International Conference on Bioinformatics and Biomedicine (BIBM 2023).
-
A. Akbarnejad and N. Ray and P. J Barnes and G. Bigras, “Predicting Ki67, ER, PR, and HER2 Statuses from H&E-stained Breast Cancer Images,” arxiv preprint: https://arxiv.org/abs/2308.01982.
Link to the datasetihc4bc.githubio
- A. Akbarnejad and G. Bigras and N. Ray, “GPEX, A Framework For Interpreting Artificial Neural Networks,” Conference on Neural Information Processing Systems (NeurIPS) 2023.
Link to the github repo https://github.com/amirakbarnejad/gpex
Link to GPEX documentation https://gpex.readthedocs.io/en/latest/
-
Y. Yang and A. Akbarnejad and N. Ray and G. Bigras, “Double adversarial domain adaptation for whole-slide-imageclassification,” Medical Imaging with Deep Learning (MIDL), 2021.
-
A. Akbarnejad and N. Ray and G. Bigras, “Deep Fisher Vector Coding For Whole Slide Image Classification,” in International Symposium of Biomedical Imaging (ISBI), 2021.
Code on github:https://github.com/amirakbarnejad/code_submission_isbi2021
-
A. Akbarnejad and M. Soleymani Baghshah, “An Efficient Semi-supervised Multi-label Classi- fier Capable of Handling Missing Labels,” in IEEE Transactions on Knowledge and Data Engineering, 2018.
-
A. Akbarnejad and M. Soleymani Baghshah, “A Probabilistic Multi-label Classifier with Miss- ing and Noisy Labels Handling Capability.,” in Pattern Recognition Letters, 2017.