Flex2Chip: slice recording and acute mouse seizure mapping, 2023
Lab or organization
Slice recording and acute seizure mapping
The 2023 Flex2Chip hardware paper demonstrates two different tissue-end arrays. Stanford affiliation is grounded in the paper; the institution link is not a claim that its human BCI laboratory conducted these experiments.
Cerebellar-slice recording
A 720-site array was mounted on a 6 mm stainless-steel ring and placed on an acute cerebellar slice. The array had 20 µm-diameter electrode pads and fenestrations for nutrient and oxygen diffusion. The authors isolated 36 individual units. Regular firing was consistent with Purkinje-cell behavior, not a claim that every unit had an independently confirmed anatomical cell identity.
This is an ex vivo slice result. It is not 720 isolated neurons and does not establish chronic living-animal survival.
Acute mouse surface recording
A separate 504-site ECoG array, with a 760 × 760 µm active area, was placed over exposed motor and somatosensory cortex through a 2 mm craniotomy in Scn8a heterozygous mice with absence epilepsy. Recordings were made within two hours of placement, with head-fixed animals able to run on a treadmill.
Phase-based analysis found traveling waves whose direction and velocity changed even within a spike-wave discharge. The report does not establish a universal seizure-onset site, an invariant propagation trajectory or human diagnostic accuracy.
Boundaries
The underlying chip’s simultaneous readout is limited to 1,024 channels; physical connector and recording-site counts are separate. Incubator aging reported in the hardware characterization is not chronic implanted recording evidence. No clinical study, therapeutic seizure control or assistive-device-control result is claimed.