Skip to content

Signal Acquisition Technology

The performance ceiling of a BCI is determined on the acquisition side: different electrodes observe entirely different signal scales, bandwidths, and channel counts. This chapter introduces acquisition technologies in order of decreasing invasiveness, and also covers stimulation (write-in) technology and the preprocessing pipeline.

The role of this chapter. It turns the "neurophysiology picture" from Chapter 02 into measurable engineering. Each hardware platform here maps to a company in Chapter 11: Utah array → BrainGate / Pitt team, flexible threads → Neuralink, endovascular → Synchron, thin-film → Precision, EEG → Muse / Emotiv. Understanding the scale ladder of electrodes (single-cell spikes → mm-scale LFP → cm-scale ECoG → scalp EEG) also explains why EEG BCIs can never reach Utah-array decoding accuracy, why Stentrode trades bandwidth for reversibility, and why fMRI cannot drive a real-time BCI.

Recommended reading order. Readers oriented toward clinical or hardware engineering should start with Invasive Electrodes and Minimally Invasive Interfaces to master the platforms that have actually been implanted in patients. Readers oriented toward consumer or algorithmic work can start with Non-Invasive Recording and compare the trade-offs among EEG / MEG / fMRI / fNIRS. Either path must end with Stimulation Technology and Signal Preprocessing: the former underpins the sensory writing in Chapter 09, while the latter (filtering / spike sorting / artifact removal) is mandatory groundwork before any decoding pipeline can begin.

Chapter contents:

评论 #