Une interface cerveau-ordinateur est une voie de communication directe entre le tissu neuronal et un système informatique externe, court-circuitant les canaux de sortie neuromusculaires conventionnels pour traduire l'activité cérébrale enregistrée en commandes de l'appareil, en parole synthétisée ou en retour sensoriel restauré en temps réel.
Le domaine se divise selon un axe invasif : modalités non invasives - EEG du cuir chevelu, infrarouge fonctionnel, etc. spectroscopie - Les grilles d'électrocorticographie placées à la surface du cortex occupent une position intermédiaire. Enfin, les approches intracorticales, caractérisées par le réseau Utah et le système d'électrodes à fil flexible de Neuralink, enregistrent simultanément des pointes d'une seule unité provenant de centaines, voire de milliers de neurones, au prix d'une implantation chirurgicale, d'une réaction à un corps étranger et d'une dégradation à long terme de la stabilité des électrodes.
traitement du signal pipelines — pic tri, la décomposition du potentiel de champ local et, de plus en plus, les décodeurs d'apprentissage profond entraînés sur la dynamique des populations neuronales — traduisent l'électrophysiologie brute en signaux de contrôle de haute dimension, avec moteur Décodage cortical pour le contrôle du curseur et robotique L'activation des membres et le décodage de la zone de parole pour la synthèse vocale imaginée ou tentée représentent les deux pistes d'application clinique les plus avancées. boucle fermée Les architectures qui combinent l'enregistrement neuronal avec une neurostimulation corticale ou périphérique précisément synchronisée font progresser simultanément la réadaptation après un AVC, la dépression résistante au traitement et la prise en charge de l'épilepsie.
Les publications et les brevets répertoriés ci-dessous couvrent la science des matériaux d'électrodes, la conception d'amplificateurs frontaux ASIC, les algorithmes de décodage, la télémétrie neuronale sans fil, les études de biocompatibilité et la technologie de l'information. essai clinique sur l'ensemble du spectre de l'invasivité :
Voici notre dernière sélection de publications et de brevets internationaux en anglais sur les interfaces cerveau-ordinateur (ICO), issus de nombreuses revues scientifiques en ligne, classés et axés sur les ICO, les interfaces cerveau-ordinateur, les interfaces neuronales, l'électrocorticographie, les réseaux d'électrodes intracorticales, les réseaux de l'Utah, les implants Neuralink, les ICO à stentrode, les ICO basées sur l'EEG, le décodage du cortex moteur, le tri des potentiels d'action neuronaux, les ICO à potentiel de champ local, les décodeurs de réseaux neuronaux à impulsions, les amplificateurs de signaux neuronaux, la neurostimulation en boucle fermée, les neuroprothèses motrices ICO, les ICO vocales, le décodage de la parole imaginée, le contrôle du curseur ICO, les dispositifs de communication ICO, les algorithmes de décodage neuronal, le rejet des artefacts ICO, les sondes neuronales flexibles, les électrodes neuronales biocompatibles et la stabilité à long terme des ICO.
Electrical overvoltage protection
Patent published on the 2026-06-24 in EP under Ref EP4765622 by MAHLE INT GMBH [DE] (Batistic Zan [si])
Abstract: The invention relates to an electrical overvoltage protection with a first, second and third heat sink (2a, 2b, 2c). Furthermore, the overvoltage protection comprises an electrical diode circuit which has a total of twelve Zener diodes (Z1, Z2, Z1 , Z2 ) and is distributed over the three heat sinks (2a-2c). Each of the twelve Zener diodes is therefore arranged on one of the three heat sinks (2a-2c). Furthermore, the electrical diode circuit comprises a first and a second electrical output termin[...]
Our summary: The invention describes an electrical overvoltage protection system featuring three heat sinks and a diode circuit with twelve Zener diodes. The Zener diodes are arranged in pairs and distributed across the heat sinks, forming two subcircuits in series. Each subcircuit includes pairs of Zener diodes with phase input connections for enhanced electrical performance.
overvoltage protection, Zener diodes, electrical circuit, heat sinks
Patent
Neuronal avalanches as a predictive biomarker for guiding tailored BCI training programs
Published on 2026-05-29 by @MIT
Abstract: AbstractMotor imagery-based Brain-Computer Interfaces (BCIs) restore control in persons with motor impairments, but up to 30% of users struggle, a phenomenon known as “BCI inefficiency”. This study tackles a key limitation of current protocol: the use of fixed-length sessions training paradigms that ignore individual learning variability. We propose a novel approach based on neuronal avalanches, spatiotemporal cascades of brain activities, as biomarkers to characterize and predict user-speci[...]
Our summary: This study introduces neuronal avalanches as predictive biomarkers for personalized BCI training programs. It analyzes electroencephalography data to correlate avalanche characteristics with BCI performance. The findings support tailored approaches to enhance user success and reduce BCI inefficiency.
neuronal avalanches, BCI training, biomarkers, motor imagery
Publication
Large language models reveal the neural tracking of linguistic context in attended and unattended multi-talker speech
Published on 2026-05-07 by @MIT
Abstract: AbstractLarge language models (LLMs) capture long-range contextual structure in natural language and have recently been shown to align with the human brain’s contextualized linguistic encoding. This makes them a promising computational probe for studying how context-dependent linguistic information is represented during natural speech perception. Speech perception often occurs in multi-talker environments, where attention must dynamically select among competing streams, yet how contextual info[...]
Our summary: Large language models align with human brain encoding of linguistic context. The study explores how attention affects neural tracking of speech in multi-talker environments. Findings indicate that both attended and unattended speech streams contribute to neural predictions based on contextual information.
language models, neural tracking, speech perception, auditory attention
Publication
Suppression of inflammation associated with implants
Patent published on the 2026-05-07 in WO under Ref WO2026092500 by SUNMED THERAPEUTIC LTD [CN] (Sun Joseph [cn], Sun Dongxu [cn])
Abstract: Provided herein are methods and compositions for suppressing a foreign body reaction, such as an implant-associated inflammation in a subject, by using an anti-Galectin-3 antibody. Such methods and compositions can be used in various areas applications where an implant is introduced, such as in brain-computer interface (BCI).[...]
Our summary: Methods and compositions are described for suppressing implant-associated inflammation using an anti-Galectin-3 antibody. These approaches target the foreign body reaction in subjects receiving implants. Applications include areas like brain-computer interfaces (BCI).
anti-Galectin-3, inflammation suppression, foreign body reaction, implants
Patent
Towards the use of functional near-infrared spectroscopy as an assessment tool in disorders of consciousness
Published on 2026-04-30 by @MIT
Abstract: AbstractFunctional near-infrared spectroscopy (fNIRS) has emerged as a promising neuroimaging tool for assessing patients with disorders of consciousness (DoC). While functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) have advanced the detection of covert brain function, their use is often constrained by accessibility, medical and physical contraindications, and practical limitations. fNIRS offers a portable, safe, and cost-effective alternative capable of measuring he[...]
Our summary: Functional near-infrared spectroscopy (fNIRS) is a promising tool for assessing disorders of consciousness (DoC). It provides a portable and cost-effective alternative to fMRI and EEG for measuring brain function. Future research should focus on validation, multimodal integration, and ethical access to enhance DoC care.
fNIRS, disorders of consciousness, neuroimaging, brain-computer interfaces
Publication
A reference-less level-crossing adc with a bump-based adaptive-bias comparator
Patent published on the 2026-04-22 in EP under Ref EP4730654 by IMEC VZW [BE] (Yang Xiaolin [be], Xing Xiaonan [be], Sawigun Chutham [be], Mora Lopez Carolina [be])
Abstract: This disclosure relates to a comparator circuit and an ADC circuit for a neural interface. The ADC circuit includes the comparator circuit. The comparator circuit comprises a comparator to receive a first and a second signal, and internally amplify the first and the second signal. A bump bias circuit of the comparator circuit receives the amplified first and second signal, and causes the comparator to operate at a higher power level when a difference between the amplified first and second signal[...]
Our summary: The disclosure describes a reference-less level-crossing ADC featuring a bump-based adaptive-bias comparator. The comparator amplifies two input signals and adjusts its power level based on their difference. The ADC includes two comparator circuits and switching circuits controlled by a control circuit for capacitor state management.
ADC, comparator, adaptive-bias, neural interface
Patent
All spectral frequencies of neural activity reveal semantic representation in the human anterior ventral temporal cortex
Published on 2026-04-17 by @MIT
Abstract: AbstractIntracranial electrophysiology offers a unique insight into the nature of information representation in the brain—it can be used to disentangle information encoded in gamma and high gamma frequencies from information encoded in lower frequencies. We used regularised logistic regression to decode animacy from time-frequency power and phase extracted from electrocorticography (ECoG) grid electrode data recorded on the surface of human ventral anterior temporal lobe (vATL). Power in gamma[...]
Our summary: Neural activity in the anterior ventral temporal cortex encodes semantic information across various spectral frequencies. Intracranial electrophysiology reveals that gamma and high gamma frequencies contribute to reliable decoding of animacy. A broader frequency range enhances decoding accuracy, supporting the concept of a local vATL hub interacting with distributed cortical spokes.
neural activity, semantic representation, electrocorticography, frequency decoding
Publication
A novel hybrid BCI system combining single-channel SSVEP and PLR to improve classification accuracy and ITR
Published on 2026-03-24 by @OXFORD
Abstract: AbstractSteady-state visual evoked potential (SSVEP)-based brain–computer interfaces (BCIs) have been widely studied because they provide high classification accuracy and information transfer rate (ITR) without requiring user training. To further enhance BCI performance, this study proposes a novel hybrid BCI that integrates single-channel SSVEP with the pupillary light reflex (PLR). Twelve healthy subjects participated in experiments involving three paradigms: SSVEP, PLR, and hybrid. Each sub[...]
Our summary: This study presents a hybrid BCI system that combines single-channel SSVEP and PLR to enhance classification accuracy and information transfer rate. The hybrid paradigm achieved a classification accuracy of 95.70%, outperforming both SSVEP and PLR methods. Results indicate that this approach can significantly improve BCI performance, potentially facilitating broader adoption.
BCI, SSVEP, PLR, classification
Publication
