Welcome to sinthlab - the Sensorimotor Integration and NeuroTheory Laboratory at the University of Montreal and Mila.

We fuse AI and computational neuroscience with experimental neurophysiology and neural engineering to study how biological brains coordinate behavior.

By linking fundamental neuroscientific insights with innovations in AI, we aim to guide the development of next-generation neural interfaces.

Neural population dynamics underlying behavior

The lab aims to understand how the concerted activity of large populations of neurons control the intricate behaviors we produce. We combine neurophysiology experiments with computational neuroscience and artificial intelligence methods to identify and characterize dynamics underlying the neural activity observed throughout the brain. We ultimately seek to uncover principles of the neural control of behavior that are conserved over time, across individuals, and even across species in order to provide robust, interpretable view of brain activity to apply to neuroprosthetic devices.

More reading: Nature 2023, Nature Neuroscience 2020, Neuron 2018, Neuron 2017

Computational methods to parse multi-region interactions

The neural control of behavior is distributed across many functionally and anatomically distinct brain regions. The lab aims to develop computational tools that can uncover, in an unsupervised manner, the distributed, multi-region interactions that guide behavior. Our approaches unveil how disparate sensory, cognitive, and motor systems of the brain coordinate to produce flexible and adaptable behavior. We employ data-driven Recurrent Neural Network (RNN) models to create in silico models of neural recordings that we can reverse engineer.

More reading: bioRxiv 2020a, CONB 2020

Development of closed-loop interfaces between artificial and biological circuits

The lab is working to understand how disparate source of inputs (both internal to the brain and from the external environment) shape the neural dynamics that we record from the brain. We draw on cutting-edge machine learning and AI to develop new classes of neural "decoders" that integrate naturally in closed-loop with ongoing dynamics in the brain. These innovations will be critical for the widespread clinical adoption of technologies such as brain-controlled spinal stimulation that promise to restore movement lost due to paralysis or movement disorders such as Parkinson's.

More reading: bioRxiv 2020b, Nature Neuroscience 2022, Nature Medicine 2023


Matthew G. Perich

I am an Assistant Professor in the Department of Neuroscience at the Université de Montréal and an Associate Member of Mila (Quebec Artificial Intelligence Institute). My research fuses AI and computational neuroscience with experimental neurophysiology and neural engineering to study how biological brains throughout the evolutionary tree coordinate behavior.

Website | Scholar | Code | Bluesky | Twitter

Olivier Codol
(co-supervisor: Guillaume Lajoie)
I am a post-doctoral researcher in computational neuroscience interested in unraveling the algorithmic basis of learning in neural control of movements. Particularly, I focus on fundamental motion such as reaching movements and tabula rasa learning that occurs in infancy, as opposed to sequential and compositional skills that require re-arrangement of pre-acquired movements. I currently work on disentangling which learning rules best match neural dynamics of learning, and what are the most efficient and stable ways to improve performance in complex control problems, including non-linear and over-determined (redundant) control. To that end, I leverage biomechanical modeling, deep recurrent neural networks, and reinforcement learning一particularly policy gradient methods, which are best suited for continuous control problems.

Website | Scholar | Code | Bluesky | Twitter

Reza Asri
Grad Student
I'm a Ph.D. candidate navigating the wondrous realms of Neuroscience at Université de Montréal and Mila. In this exciting journey, I'm exploring the fascinating blend of artificial intelligence and neuroscience. My mission involves using powerful AI tools to unravel the secrets hidden within neurons and physiological data. I'm particularly captivated by AI models that mimic brain circuits involved in sensory-motor loops to decode the brain's mysterious language, uncovering how it processes information and choreographs our movements. We unearth the fundamental neural mechanisms behind sensation and motor control.

Scholar | Code | Twitter

Avery Ryoo
Grad Student
(co-supervisor: Guillaume Lajoie)
I am a graduate student in computer science working in the intersection of artificial intelligence and neuroscience. In particular, I am intrigued by how the brain can learn adaptable representations from minimal amounts of data to solve a variety of complex tasks, and how these insights can lead to more robust and interpretable AI algorithms and learning schemes. To investigate this, I aim to use tools from deep learning theory, dynamical systems, and cognitive science. Some other interests include foundation models, probabilistic inference, and brain-computer interfaces.

Outside of research, I enjoy daytime napping, singing Taylor Swift songs (pre-1989) in the shower, and delaying the inevitable.

Website | Scholar | Code | Twitter

Anirudh Jamkhandi
Grad Student
Blurb incoming...

Website | Code | Twitter

Ali Korojy
Grad Student
Blurb incoming...

Soraya Rahimi
Grad Student
(co-supervisor: Numa Dancause)
Blurb incoming...

Marlene Boutet
Blurb incoming...


Kelty Antilus | Intern

Join our team!
We are recruiting a postdoc and a research technician to help build our experimental research program.
Contact us if you're interested in multi-region neurophysiology during naturalistic reaching and grasping behaviors!


Paper Highlights

For a complete publication list, see Matt's Google Scholar
Safaie M, Chang J, Park J, Miller LE, Dudman JT, Perich MG*, Gallego JA*. (2023) Preserved neural population dynamics across animals performing similar behaviour. Nature. (*: co-supervising and co-corresonding authors)

Gallego JA*, Perich MG*, Chowdhury RH, Solla SA, Miller LE. (2020) Long-term stability of cortical population dynamics underlying consistent behavior. Nature Neuroscience. (*: co-first authors)

Perich MG, Rajan K. (2020) Rethinking brain-wide interactions through multi-region “network of networks” models. Current Opinion in Neurobiology.

Perich MG, Arit C, Soares S, Andalman A, Benster T, Young ME, Mosher CP, Minxha J, Carter E, Rutishauser U, Rudebeck PH, Harvey CD, Deisseroth K, Rajan K. (2021) Inferring brain-wide interactions using data-constrained recurrent neural network models. bioRxiv.

Perich MG, Conti S, Badi M, Bogaard A, Barra B, Wurth S, Bloch J, Courtine G, Micera S, Capogrosso M, Milekovic T. (2020) Motor cortical dynamics are shaped by multiple distinct subspaces during naturalistic behavior. bioRxiv.

Perich MG, Gallego JA, Miller LE. (2018) A neural population mechanism for rapid learning. Neuron.

Gallego JA, Perich MG, Miller LE, Solla SA. (2017) Neural manifolds for the control of movement. Neuron.

Barra B*, Conti S*, Perich MG, Zhuang K, Schiavone G, Fallager F, Galan K, James N, Barraud Q, Delacombaz M, Kaeser M, Rouiller EM, Milekovic T, Lacour S, Bloch J, Courtine G, Capogrosso M. (2022) Electrical stimulation of the cervical dorsal roots enables functional arm and hand movements in monkeys with spinal cord injury. Nature Neuroscience. (*: co-first authors)


For information about research and job opportunities:

matthew dot perich at umontreal dot ca.

Follow us on twitter: @mattperich

We are looking for a postdoc and/or technician to join our team!

Join us to help explore the interface between neuroscience, AI, and neuroprosthetics. Montréal has a thriving and vibrant neuroscience and AI community. All lab members enjoy access to numerous institutes including Mila and the Institut TransMedTech.

We are committed to supporting scientists from diverse backgrounds, including under-represented groups, and take great effort to ensure the lab is an inclusive and supportive environment for professional growth.

Our lab is part of the Départment de neurosciences in the Faculté de médecine at the Université de Montréal.

We are located at:
Pavillon Paul G. Desmerais
2960 Chemin de la Tour
Montréal QC H3T 1T9