Dataview
LIST
FROM #Collection
WHERE file.name = this.Entry-For๐ -> Week 3: The BCI Pipeline
๐ค Vocab
โ Information
OpenBCI:
- A BCI that records EEG data from the scalp
- EEG (analog signal) -> OpenBCI (digital signal)
The Pipeline
The Brain -
Produces the electrical signals internally, that is received from the scalp by the EEG
Scalp Signals -> Signal Acquisition
Signal Processing Hardware -
Obtaining the signal is a crucial step, and OpenBCI is a popular candidate
- Open source
- High compatibility
Sampling frequency - Rate at which the EEG captures / polls the data
OpenBCI (digital signal) -> Preprocessing (digital signal to be read by the computer)
Preprocessing -
Brainflow / MNE
- Both are highly advanced python libraries designed to take, read, and process brain activity data
Preprocessing mainly consists of filtering the data such that it is in a usable state to feed into the next step of the process: - Bandpass filter - Removes all the high and low frequencies and keeps a band of them in the middle
- High Pass - Gets rid of high freqs
- Low Pass - Gets rid of low freqs
Preprocessing occurs in both hardware and software - Hardware filters are significantly faster than digital filters
Fourier Transform Vid
Preprocessing (digital EEG signal) -> Processing (internal python data structures)
Algorithm / Processing
Brainflow / MNE
- The algorithm/processing stage processes the data to find what we want
- This can involve creating a classifier
Processing (internal python data structures) -> Decision
What are classifiers?
A classifier is a model that is able to discriminate between different types of inputs. In our case we want out BCI to be able differentiate between different types of signals and artifacts.
Decision
The stage of the pipeline where the BCI makes a judgement regarding its data
- BCI looking for changes in brain activity based on two different frequencies and rendered a binary decision based on that
After this, the pipeline will be restarted
โ๏ธ -> Scratch Notes
๐งช-> Example
- List examples of where entry contents can fit in a larger context