Many studies point to neurofeedback as a potent technique to treat particular mental disorders, such as epilepsy and ADHD.
To me it seems that NF therapy may be broken down into 1. determining the NF-protocol to be applied, which is based on anamnesis/questionnaires and comparison of a multichannel qEEG measurement to Z-scored norms, and 2. the actual application of this protocol (NF-training).
Validating a proper qEEG measurement to Z-scored norms and deciding what to do based on your psychological profile implies experienced professionals -- at least to me it does. This professional result however, is a fairly transparent NF-training recipe, involving eeg-feedback via only a limited number of channels.
Intending to cut on cost (yes, indeed), i decided to start building the openBCI Cyton board (8 channels), find software and do some programming.
At the moment of writing, I ordedered PCB's and components following http://www.thinkering.de/cms/?p=703. Since the RFDuino seems to be obsolete, i considered http://www.fredrikolofsson.com/f0blog/?q=node/664.