Combines clinical neurosurgery, intracortical signal recording, and AI-supported neural decoding. In summer 2025, as part of the program, the first chronic microelectrode BCI in Europe was implanted at the Klinikum der Technischen Universität München (TUM University Hospital) in a patient with traumatic spinal cord injury.

Indication: traumatic spinal cord injury

Methods: chronic intracortical recording, decoding, AI

Funding: BMFTR

AID website

Publications and further information

Combines clinically indicated procedures with scientific microelectrode recordings from different cortical regions. The approach opens access to human single-cell and population activity while behavior is monitored at the same time.

Methods: acute microelectrode arrays, single-cell recording

Funding: DFG

Publications and further information

First in the world for clinical research program on the restauration of communication afer stroke, running since 2022. Chronic neural recordings and patient-specific decoding gradually translate brain signals into outputs that can be used for communicatoin. Investigates how linguistic and communicative information is represented in cortical networks. Particular weight is given to hthe contribution of the right hemisphere and to the relationship between single-cell activity, natwork organization, and behavior.

Indication: severe language impairment after stroke

Methods: chronic BCI recording, speech decoding

Funding: ERC

Publications and further information

Builds on the clinical and technological experience gained in AID. Three pillars are planned: neuromorphic and data-driven decoding, a brain-spine interface architecture, and hand neuroprosthetics with exoskeleton, EMG and VR integration. A soft robotics branch complements the project.

Status: concept and application phase

Components: neuromorphic AI, digital bridge, neuroprosthetics, soft robotics

Publications and further information

Soft, wearable assistance systems for gait, arm, and hand function. Soft exosuits and active orthoses support a movement without guiding it completely through rigid structures, and they can be coupled with neural and myoelectric decoding.

Status: Ongoing development line

Technologies: soft exosuits, active orthoses, sensor systems

References: Nature Machine Intelligence 2024; Soft Robotics 2024

Publications and further information

From the talk by Prof. Lorenzo Masia at IROS 2025.