(Im)poster syndrome: preparing for Neuroscience 2024
A selection of posters from Neuroscience 2024 that might help bring general-purpose BMI closer

After two amazing days at the IEEE Brain Discovery & Neurotechnology Workshop (notes from it—TBD), I’m recharging before Neuroscience 2024 and reviewing the poster abstracts.
Unsurprisingly, general-purpose brain-machine interfaces (BMIs) or non-medical BMIs do not have a separate section. However, a few papers that might move us closer to the general-purpose BMI are scattered here and there. I’d structure these papers into four groups:
- Fundamental research offering insights into how the brain works that are relevant for developing BMIs;
- Software/hardware improvements for existing BMI systems, e.g., in terms of resolution, longevity, etc.;
- Generalisation/platformisation of BMI systems, for example, through multi-functionality or expansion across patient populations;
- Emerging systems, less established BMIs or imaging/microscopy systems.
If you are around and enthusiastic about the themes above, or building a neurotech company - reach out via p@e184.com.
1) Fundamental research
Motor BMIs: classical concepts of somatotopy and iBCI implantation (link); specialised regions within the motor cortex that guide distribution and coordination of neural signals across the cortex (link); similar neuronal dynamics in a different behavioural context and potential to generate varied motor outputs (link); internal dynamics of neural populations that do not directly encode movement kinematic variables but are still involved in the task (link).
Speech BMIs: a study into the extent to which the content of the inner speech was discernible using a decoder trained on attempted vocalised speech (link); the role of the 6v area in the speech planning hierarchy (link); exploring how hierarchical linguistic information (phoneme - words - phrases - sentences) is encoded in different cortical areas and how to incorporate higher level information to make iBCI speech decoders more accurate and efficient (link).
2) BMI improvements
Software: designing decoders that co-adapt to a user influencing user learning and opening new ways to achieve high-performing, individualised neural interfaces (link); a decoder with quick adaptation for practical everyday use and leverage available historical data of neural cursor control (link); comparison of spatial filtering techniques for Stentrode (link); a foundation model for motor tasks, achieving high performance with minutes to a few hours of a new task (link); pairing QWERTY keyboard with a language model for reducing decoder confusion (link); protocols for assessing the impact of manual spike sorting curation (link).
Hardware: exploring factors affecting recording quality of Stentrode (link); benchmarking Stentrode against scalp EEG (link); improved limb recovery when epidural electrical stimulation (EES) is synchronised with a patient's intentions via a BMI (link); using sEMG recordings for improving cortical decoding (link).
3) Generalisation and platformisation
Expanding across patient populations: patients with the loss of control over spontaneous respiration, and the lack of real-time auditory and somatosensory feedback (link); soft robotic glove control (link); a wearable robotic shoulder control (link).
Improving quality: decoding attempted speech amplitude (link); enhancing the strength of movement-related information in the neural control signal to improve control of external devices, e.g. the speed of cursor (link);
Multifunctional BMIs: combining rapid brain-to-text predictions (via attempted speaking) with cursor control for text correction (via intended arm/hand movements) (link); speech-to-text combined with hand gestures decoding for switching between rows of a virtual keyboard (link); placing arrays in vPCG (speech motor cortex) in order to operate both a cursor iBCI and a speech iBCI, instead of splitting arrays between ventral and dorsal motor areas (link); voice synthesis (link).
4) Emerging systems
Imaging/microscopy: lensless microscope for mesoscopic calcium imaging (link); photoemission electron microscopy (link); cell barcoding and highly multiplexed imaging (link); a remote scanning light-sheet microscope capable of imaging GEVI-expressing neurons (link).
BMIs: flexible and soft bioelectronics with tissue-like properties (link); a silicon probe with much smaller and denser recording sites than previous designs (link); a novel consumer EEG device (link); fiberscopes for optical imaging (link).
📬📬📬 Subscribe for notes like this and monthly updates on neurotech and neurotech investment/commercialisation activity.