Amr Farahat

I believe that we reached the time when AI and neuroscience have to advance together. That is why after graduating medical school, I moved to Germany to do a master degree in integrative neuroscience with a special focus on computational neuroscience and data analysis, studied machine learning and taught myself deep learning. My goal is to develop novel machine learning-based data analysis techniques that are capable of handling the vast amount of neural data produced in neuroscience labs and help accelerate neuroscience research. Moreover, I concur with the philosophy of neuroscience-inspired AI research and believe that accelerating the neuroscience research will benefit the AI field itself. Additionally, because of my medical background, I am also interested in applying machine learning to healthcare problems. I have practiced medicine for a period of time before turning to research and I see how machine learning can empower healthcare specialists especially in ambiguous scenarios and scenarios that require fast decision making.


Experience

PhD Student

Vinck Lab, Ernst Strüngmann Institute for Neuroscience in Cooperation with Max Planck Society, Frankfurt, Germany.
May 2020 – Present

Neuroradiology Resident Doctor

Neuroradiology Department – Otto von Guericke University Hospital, Magdeburg, Germany.
April 2019 – October 2019

Reseach Assistant (fMRI data analysis)

Neurology Department – Otto von Guericke University Hospital, Magdeburg, Germany. Neurocybernetics and Rehabilitation group. Dr. Catherine Sweeney-Reed.
April 2017 – March 2019

General Practitioner

Ministry Of Health and Population, Egypt.
March 2015 – September 2015

Internships

Research Intern

Ernst Strüngmann Institute for Neuroscience in Cooperation with Max Planck Society, Frankfurt, Germany. Vinck Lab
February 2020 – April 2020

Research Intern

Frankfurt Institute for Advanced Studies (FIAS), Frankfurt, Germany. Research Group of Jochen Triesch
November 2019 – Present

Master thesis student

Leibniz Institute for Neurobiology (LIN), Magdeburg, Germany. Department of Behavioral Neurology. Dr.-Ing. Christoph Reichert.
January 2018 – August 2018

Research Intern (EEG data analysis)

Leibniz Institute for Neurobiology (LIN), Magdeburg, Germany. Visual Attention & Perceptual Learning research group. Prof. Dr. Jens-Max Hopf.
December 2017 – January 2018

Research Intern (Modeling neural networks)

German center for neurodegenerative diseases (DZNE), Magdeburg, Germany. Cognitive Neurophysiology research group. Dr. Motoharu Yoshida.
August 2016 – February 2017

Reseach Intern (LFP data analysis)

Leibniz Institute for Neurobiology (LIN), Magdeburg, Germany. Department of systems physiology. Dr. Max Happel.
March 2016 – April 2016

Intern Doctor

Mansoura University Hospitals, Egypt.
March 2014 – February 2015

Education

International Max Planck Research School for neural circuits, Frankfurt, Germany

PhD student
November 2019 - Present

Otto von Guericke University, Germany

Master of Science (Integrative Neuroscience)
October 2015 - September 2018

Mansoura University, Egypt

MBBCh (Bachelor of Medicine, Bachelor of Surgery)
September 2007 - February 2014

Further Qualifications

IBRO-Simons Computational Neuroscience Imbizo 2022

Cape Town, South Africa

FENS Summer School on ‘Artificial and natural computations for sensory perception: what is the link?’ 2022

Bertinoro, Italy

Eastern European Machine Learning Summer School (EEML) 2021

A deep learning and reinforcement learning summer school

Neuromatch Academy (Interactive Track) 2020

An online school for Computational Neuroscience.

International Startup School 2018

Transfer- und Gründerzentrum (TUGZ), Otto von Gureicke University, Germany.

Business Planning Course 2018

Chair of Entrepreneurship, Otto von Gureicke University, Germany.

5th Human Brain Project winter school: Future Medicine 2017

Obegugl, Austria

Influencing clinical diagnoses and treatments by data mining analysis- and modeling-driven neuroscience.

4th Human Brain Project summer school: Future Computing 2017

Obegugl, Austria

Brain Science and Artificial Intelligence.


Recent Projects

The role of spatial relations among CNN features of different granularities in object recognition

We develop a feature-scrambling approach in order to investigate the granularity of features used by CNNs for object recognition and whether CNNs are capable of encoding the spatial relations among features to build more complex diagnostic features for objects.
Abstract Poster Preprint

Diagnosing Epileptogenesis with Deep Anomaly Detection

We use data from a rodent epilepsy model to show the feasibility of using an unsupervised deep anomaly detection framework using adversarial autoencoders to detect subtle changes in brain electrical activity triggered by brain injury that lead eventually to an epileptic brain.
Abstract Paper Code Video

Deep learning for EEG decoding and brain dynamics discovery

I am using deep learning models for classifying P300 ERP component in EEG data for BCI speller applications. In addition to that, I am using different features visualization techniques e.g. saliency maps to extract the important spatial and temporal features that drive the model and should reflect where attention originates in the brain and compare it to what we already know from the neuroscience literature.
Abstract Paper Preprint Poster Master Thesis Code

Investigating motor learning circuitry in a 7T functional connectivity fMRI study

Motor/sequence learning is known to involve a cerebellar-thalamo-cortical circuit that especially includes VIM (Ventral Intermediate Nucleus Of The Thalamus). We are trying to look closely into VIM with 7T high spatial resolution to investigate if it is possible to subdivide it into functionally distinct subregions.
Poster 1 Poster 2 Talk Paper 1 Paper 2

YODO You Only Diagnose Once

I am using YOLOv2 for detcting pneumonic patches in chest X-ray images.
Code

Publications

Peer-reviewed papers Farahat, A., Lu, D., Bauer, S., Neubert, V., Costard, L.S., Rosenow, F. and Triesch, J., (2022). Diagnosing Epileptogenesis with Deep Anomaly Detection. Proceedings of the 7th Machine Learning for Healthcare Conference, PMLR 182:1-18
Voegtle, A., Terlutter, C., Nikola, K., Farahat, A., Hinrichs, H., Sweeney-Reed, C., (2022). Suppression of Motor Sequence Learning and Execution Through Anodal Cerebellar Transcranial Electrical Stimulation. The Cerebellum. doi:10.1007/s12311-022-01487-0
Terzic, L., Voegtle, A., Farahat, A., Hartong, N., Galazky, I., Nasuto, S.J., Andrade, A.D.O., Knight, R.T., Ivry, R.B., Voges, J. and Buentjen, L., Sweeney-Reed, C., (2022). Deep brain stimulation of the ventrointermediate nucleus of the thalamus to treat essential tremor improves motor sequence learning. Human Brain Mapping. doi:10.1002/hbm.25989
Farahat, A., Reichert, C., Sweeney-Reed, C. M., & Hinrichs, H. (2019). Convolutional neural networks for decoding of covert attention focus and saliency maps for EEG feature visualization. Journal of Neural Engineering. doi:10.1088/1741-2552/ab3bb4
Preprints Farahat A, Effenberger F, Vinck M (2022) A novel feature-scrambling approach reveals the capacity of convolutional neural networks to learn spatial relations. arXiv. doi:10.48550/ARXIV.2212.06021
Master thesis Deep Learning for EEG Decoding and Automatic Feature Discovery.
Abstracts and Posters Farahat A, Effenberger F, Vinck M (2022). The role of spatial relations among CNN features of different granularities in object recognition. Bernstein Conference 2022. doi:10.12751/nncn.bc2022.285
Farahat A, Lu D, Bauer S, Rosenow F, Triesch J (2021) P2. Unsupervised anomaly detection for diagnosing brain disorders from EEG Recordings – results from a rodent epilepsy model. Clinical Neurophysiology. doi:10.1016/j.clinph.2021.02.327
Farahat A, Effenberger F, Vinck M (2021) How do convolutional neural networks understand shape? Champalimaud Research Symposium.
Farahat A, Hinrichs H, Heinze H-J, Sweeney-Reed CM (2019) Implicit motor learning modulates functional connectivity of DBS therapeutic target for tremor. Clinical Neuroradiology, 29(1), 32–33. doi:10.1007/s00062-019-00826-9
Farahat A, Reichert C, Sweeney-Reed CM, Hinrichs H (2018) Convolutional neural networks for EEG decoding and exploration of brain dynamics. Bernstein Conference 2018. doi:10.12751/nncn.bc2018.0092
Farahat A, Hinrichs H, Heinze H-J, Sweeney-Reed CM (2017) Can The Ventral Intermediate Nucleus Of The Thalamus Be Divided Into Functionally Distinct Subregions Differentially Involved In Motor Learning And Essential Tremor? 5th HBP Winter School 2017.

Awards

  • Scholarship granted from Leibniz institute for Neurobiology in the context of the project “SFB 779 – MGK” for scoring the best grades in the integrative neuroscience master program.
  • Poster prize during the Neurorad Conference in Frankfurt 2019 awarded from the German and Austrian Societies of Neuroradiology.

Interests

Apart from my scientific interests, in my free time I like to read and watch documentries, movies and sci-fi TV shows.

I am also passionate about cooking and like to explore different kinds of food.

My favourite sports are biking and swimming.