Periodic Reporting for period 2 - INTRECOM (Intracranial Neuro Telemetry to Restore Communication)
Período documentado: 2023-12-01 hasta 2025-05-31
Not being able to communicate while still being conscious is a horrifying prospective for many patients worldwide. Patients with motor neuron disorders, trauma or stroke, risk losing complete muscle control leading to Locked-In Syndrome (LIS) which leaves them completely paralyzed and unable to communicate. This is a fearful situation with significant impact on quality of life and very high burden of care for patients, family, and care givers. Intracranial Neuro Telemetry to REstore COMmunication (INTRECOM) will provide a breakthrough for these patients by developing a novel, fully implantable Brain-Computer Interface (BCI) technology that allows for real-time brain-based computer control and communication in the home environment. This BCI system will significantly transcend current technology by providing a high-performance sustainable device, combining state-of-the-art hardware and software solutions based on Microelectronics Engineering and Artificial Intelligence (AI) to liberate LIS patients from their isolation.
INTRECOM’s ambition is to substantially advance BCI technology and validate its working principle in LIS patients in the home environment. The specific objectives of INTRECOM are to:
1. Develop a safe and fully functional high-performance BCI prototype device with 128 electrodes embedded in silicone sheets under the skull and dura for electrocorticography (ECoG grids), and innovative implantable connectors that allows safe and high-quality 24/7 brain signal recording.
2. Develop decoding algorithms based on AI that can translate brain signals to real-time computer control, as well as real-time computer speech (written or audio), with less than 1 second delay.
3. Implant the developed prototype device in 2 people with LIS and validate its working principle, including use in the home environment with initial basic functionality within weeks after implantation, and real-time intelligible speech decoding, thereby restoring communication by the end of the project.
4. Gather an unprecedented wealth of brain data through the in vivo research. Data will improve INTRECOM's hard- and software, neural signal decoding algorithms, and functionality and use of the device. Openly shared, they will advance scientific knowledge of human brain function, BCI, and therapeutic opportunities, and provide novel real-time decoding principles also for future closed-loop neurostimulation applications.
5. Show acceptability of the BCI technology by patients, caregivers, and health care professionals by the end of the project.
INTRECOM’s ambition is to substantially advance BCI technology and validate its working principle in LIS patients in the home environment. The specific objectives of INTRECOM are to:
1. Develop a safe and fully functional high-performance BCI prototype device with 128 electrodes embedded in silicone sheets under the skull and dura for electrocorticography (ECoG grids), and innovative implantable connectors that allows safe and high-quality 24/7 brain signal recording.
2. Develop decoding algorithms based on AI that can translate brain signals to real-time computer control, as well as real-time computer speech (written or audio), with less than 1 second delay.
3. Implant the developed prototype device in 2 people with LIS and validate its working principle, including use in the home environment with initial basic functionality within weeks after implantation, and real-time intelligible speech decoding, thereby restoring communication by the end of the project.
4. Gather an unprecedented wealth of brain data through the in vivo research. Data will improve INTRECOM's hard- and software, neural signal decoding algorithms, and functionality and use of the device. Openly shared, they will advance scientific knowledge of human brain function, BCI, and therapeutic opportunities, and provide novel real-time decoding principles also for future closed-loop neurostimulation applications.
5. Show acceptability of the BCI technology by patients, caregivers, and health care professionals by the end of the project.
Over the past 30 months, INTRECOM has advanced its brain–computer interface system, enhancing both hardware and software while achieving key scientific and clinical milestones. Collaboration remains central, with the BOX platform and regular (in-person) meetings supporting clear communication, coordination, and agile decision-making across partners.
Hardware and Manufacturing Progress
Significant achievements were made in the development and production of the ABILITY implant and associated components. These efforts were accompanied by process and yield improvements that have optimized manufacturing workflows. Fully functional prototypes have been manufactured and used to set up dedicated test equipment. Extensive testing has been carried out, including biocompatibility, packaging, and sterilization assessments. A comprehensive test plan for first-in-human use has been established, and the investigational medical device dossier structure for regulatory submission has been created.
Software and Algorithmic Advances
On the software side, the team transitioned from the earlier monolithic NeuroKey platform to a modular, high-performance software architecture, enabling faster iteration and broader compatibility; BOLT. This modular approach has supported the successful deployment of brain-to-speech and movement decoding algorithms, culminating in demonstrations where participants controlled communication software directly using their brain signals, using EEG.
Movement decoding
Research on movement decoding has revealed how visual cue design affects brain responses. Additional studies showed that fatigue throughout the day leads to reduced EEG amplitudes and classification accuracy.
Further, EEG-inspired decoding methods applied to ECoG data from patients with epilepsy and ALS enabled precise, asynchronous gesture decoding and uncovered circadian fluctuations in brain signals.
Speech Decoding
In the area of speech decoding, algorithms for single-word recognition were successfully implemented and validated in real time. ECoG data collected from epilepsy patients demonstrated that 30 repetitions per word are needed to achieve 90% classification accuracy, with the ventral sensorimotor cortex emerging as a key region for decoding.
Outreach and Impact
The project has received media coverage across TV and radio and produced eight scientific publications and preprints. A spin-out company from the Wyss Center has been established; ABILITY NeuroTech, to further develop and commercialize the ABILITY SD system. Collaboration with KU Leuven is planned to expand recruitment into Dutch-speaking regions.
Hardware and Manufacturing Progress
Significant achievements were made in the development and production of the ABILITY implant and associated components. These efforts were accompanied by process and yield improvements that have optimized manufacturing workflows. Fully functional prototypes have been manufactured and used to set up dedicated test equipment. Extensive testing has been carried out, including biocompatibility, packaging, and sterilization assessments. A comprehensive test plan for first-in-human use has been established, and the investigational medical device dossier structure for regulatory submission has been created.
Software and Algorithmic Advances
On the software side, the team transitioned from the earlier monolithic NeuroKey platform to a modular, high-performance software architecture, enabling faster iteration and broader compatibility; BOLT. This modular approach has supported the successful deployment of brain-to-speech and movement decoding algorithms, culminating in demonstrations where participants controlled communication software directly using their brain signals, using EEG.
Movement decoding
Research on movement decoding has revealed how visual cue design affects brain responses. Additional studies showed that fatigue throughout the day leads to reduced EEG amplitudes and classification accuracy.
Further, EEG-inspired decoding methods applied to ECoG data from patients with epilepsy and ALS enabled precise, asynchronous gesture decoding and uncovered circadian fluctuations in brain signals.
Speech Decoding
In the area of speech decoding, algorithms for single-word recognition were successfully implemented and validated in real time. ECoG data collected from epilepsy patients demonstrated that 30 repetitions per word are needed to achieve 90% classification accuracy, with the ventral sensorimotor cortex emerging as a key region for decoding.
Outreach and Impact
The project has received media coverage across TV and radio and produced eight scientific publications and preprints. A spin-out company from the Wyss Center has been established; ABILITY NeuroTech, to further develop and commercialize the ABILITY SD system. Collaboration with KU Leuven is planned to expand recruitment into Dutch-speaking regions.
During the INTRECOM project, we will make it possible to record over one hundred times more neural data than current implantable BCIs, producing cleaner and richer signals from the sensorimotor cortex. Combined with INTRECOM’s machine learning decoding software, these signals can be translated directly into computer control or speech with less than one second of delay, enabling real-time communication in everyday settings. The device therefore addresses an urgent need for individuals with locked-in syndrome, allowing independent interaction with caregivers, family, and society.
INTRECOM’s hardware development focuses on three main innovations. The first is the ABILITY SD implant itself, a compact, high-channel-count amplifier system powered wirelessly. The second is the development of implantable high-channel connectors that enable safe replacement of electronic components without removing the electrodes. The third is the optimisation of ECoG grid production through the use of double parylene layers, which strengthen the grid structure while maintaining flexibility. Together, these advances move BCIs closer to real-world clinical and commercial use.
INTRECOM’s hardware development focuses on three main innovations. The first is the ABILITY SD implant itself, a compact, high-channel-count amplifier system powered wirelessly. The second is the development of implantable high-channel connectors that enable safe replacement of electronic components without removing the electrodes. The third is the optimisation of ECoG grid production through the use of double parylene layers, which strengthen the grid structure while maintaining flexibility. Together, these advances move BCIs closer to real-world clinical and commercial use.