2 ms·

Epileptic seizures can be predicted through researchers' software

• program 12y ago
  > Gras and Golestanti have patented the software, and now need a device that can constantly monitor a patient’s EEG. As far as I know there are no portable devices that can produce a clean EEG. Motiv/Neurosky devices don't count. Addendum: in the USA you can't drive a car during the 365 days following the last seizure
  • deutronium 12y ago
    You can now get Ti ADCs that can be used for capturing EEG data from electrodes for around $30 iirc, so surely it'd be possible to make one cheaply enough? 8 channel, 24 bit: http://www.ti.com/product/ads1299 http://www.ti.com/product/ads1299
  • plg 12y ago
    > Gras and Golestanti have patented the software, yuck imagine if the fast fourier transform had been "patented" imagine if the "mean" (sum of all values divided by the number of values) had been patented do they even give the full algorithm in the paper? if not how was the paper published? OK went to the actual paper: http://goo.gl/hkiQKp http://goo.gl/hkiQKp How can they patent the algorithm when the work was funded by the government (i.e. the public)? Even if it's legal, it's not ethical.
  • cinquemb 12y ago
    I'd agree, and the portable ones don't give access to raw data so you have to hack them. I wonder if a robust headset will come to market soon?
• pierrel3 12y ago
  A/d conversion is the least of the issues with mobile eeg monitoring. The signals are on the order of 10-300 micro volts so they need lots of amplification. If the amp is at the other end of an electrode motion of the wire through the earth's magnetic field generates signals of a comparable size. So ideally you would need to place the amp on or in the electrode, which is a bit more of a challenge. You also might have to worry about muscle signals like eye movement artifact