> EOG interface to a computer to collect the data, and then use the computer to
> control the alarm. The only real problem is having to apply and stabilize
> the elctrodes each night before going to sleep. There might be some simpler
>> Joey King
How about activating the alarm when a major body movement takes place
in the morning? Quoting Jukka Alihanka's thesis, Gastaut and Broughton
found in 1965 that these movements were associated with EEG changes
ranging from a single K-complex to an intense awaking reaction, but
usually only passage into lighter phase of sleep. Body movements can
be detected without electrodes, e.g. by means of the "static charge
sensitive bed" (SCSB) designed by Alihanka et al. It is basically a
mattress between two layers of thin metal folio which form a
capacitor. When the body moves, the capacitance changes. The mattress
is quite sensitive even to small movements and it is being used in
many hospitals in Scandinavia.
Alpo Varri, DSP lab., Tampere University of Technology, Finland.