Who We Are

Brain Co-Processors is a moonshot research initiative housed within the Pratiksha Trust–IISc Brain, Computation and Data Science (BCD) Initiative at the Indian Institute of Science, Bengaluru. We are a multidisciplinary team of cognitive neuroscientists, electrical engineers, bioelectronics researchers, AI scientists, and clinical partners united by a shared mission: to restore function and dignity to those who have lost it through stroke and neurological disease. Our team spans six research groups at IISc and IIT Kanpur, working in close collaboration with leading neurosurgery and neurology centres across India.

Our Mission

To develop intelligent, adaptive neural co-processors, both implantable and non-invasive, that restore goal-directed reaching in stroke survivors, by combining fundamental neuroscience with cutting-edge AI and indigenous hardware.

The Grand Challenge

Stroke is one of India’s most pressing health crises. India accounts for nearly 69% of global stroke incidences, 71% of stroke-related deaths, and 78% of stroke-related disability-adjusted life years lost. Yet fewer than 1 in 1,000 eligible patients in India currently receive epilepsy surgery; analogous gaps exist for stroke rehabilitation technology.

The problem is not a lack of will — it is a lack of tools that are simultaneously effective, affordable, scalable, and grounded in how the brain actually works. That is the gap we exist to close.

Our Vision

We envision a future in which stroke-induced deficits are not merely managed but fundamentally reversed through intelligent neural co-processors that work with the brain, not around it. We also envision a future in which these technologies are Made in India, affordable, clinically validated, and accessible to the patients who need them most, not just in urban hospitals but across the country.

Beyond stroke, the platforms, datasets, algorithms, and hardware we develop will have broad applications: for epilepsy, Parkinson’s disease, dementia, traumatic brain injury, and the fundamental science of human cognition.

Motivation: Why Now?

Three convergences make this the right moment for this project. 

First, neuroscience has reached the point where we understand enough about the brain’s distributed network computations to design interventions that respect and leverage them. Second, artificial intelligence, particularly deep learning and neuromorphic computing, has matured to the point where real-time, low-power, adaptive decoding of neural signals is tractable.
Third, Indian engineering and clinical infrastructure has grown to the point where a fully indigenous, globally competitive effort is not only possible but necessary.

Project Timeline

Phase I

Technology Development and Non-Invasive Co-Processor (Years 1–5)

01
Year 1
Establish human intracranial recording setup; recruit initial stroke cohort; design early neuromorphic chip
02
Year 2
Integrate EEG decoding with real-time feedback loops; begin clinical deployments
03
Year 3
Large-scale data collection; cross-population validation; electrical stimulation protocols
04
Year 4
Indigenous implantable probes validated in animal models; seizure localization tool deployed
05
Year 5
Full non-invasive system prototype; feasibility studies in stroke patients; public database release

Phase II

Embedded Co-Processor for Restorative Intervention (Years 6–10)

06
Year 6
Co-processor algorithms tested in real-world implant context with sEEG patients
07
Year 7
Limited cortical implants in stroke survivors in attention and premotor areas
08
Year 8
Neuromorphic processor embedded in an indigenous implant system
09
Year 9
Longitudinal recovery trials; quality-of-life and neural plasticity assessment
10
Year 10
Regulatory readiness; scale-up; open-source release of AI and datasets

Approach

We decompose the complex behavior of goal-directed action into its component cognitive and sensorimotor processes: visual perception, spatial attention, decision-making, motor planning, execution, and feedback. For each component, we build a modular decoder and stimulation interface. These modules are then integrated into a unified, adaptive co-processor that can interface with the brain across multiple cortical regions simultaneously.

Unlike conventional brain-computer interfaces that optimize for a single narrow function, our co-processor is designed to support naturalistic, contextually meaningful behaviors — and to learn and adapt as the patient’s brain reorganizes during recovery.

Ethical and Clinical Framework

Patient welfare is at the center of everything we do. All research activities follow national (ICMR) and international (NIH, Helsinki Declaration) ethical standards. Informed consent, data privacy, and long-term monitoring protocols are embedded in every study. We involve neurologists, therapists, patients, and caregivers in the design of our research and technologies throughout the project lifecycle.