Is there a way to run an opensource QEEG esk program without have to spend thousands of dollars on a clinical software. I am thinking of the typical z-score QEEG map.
Not that I am aware of. The standard packages such as Neuroguide, BrainDx, HBI, qEEG-Pro, etc., all seemingly cost so much because of the time and effort that went into creating and maintaining the carefully controlled normative databases. These typically contain 100's to 1000's of controlled 'normative' client QEEGs. That are then averaged to compute the z-score deviation for the subject in question.
Nice comparison page, pdf links on that page are a bit out of date, new ones are below.
Pete van Deusen's system avoids using QEEGs. By instead analyzing for certain patterns he has determined over decades of experience. It works. Once the assessment recording is made, Pete's software processes that into recommended protocols. These are NOT z-score training protocols. But instead various band dynamics based on relative amplitudes, coherence, etc.
Alex, it is possible that a university or research institute could create a public open access QEEG database. But such a project would be costly and take much time. There are these resources of anonymized datasets:
Existing QEEG databases contain 100's to 1000's of extremely carefully controlled 'normative' client QEEGs. This cannot be done by hobbyists or distributed research groups. But instead require a centralized funded system of continuous support.
If you are interested in alternatives, see the previous comment links.
I know this is an old thread, but i've recently been fascinated by the idea of neurofeedback training relative to the normative database and since i am currently too poor to allow myself a QEEG based system like NeuroGuide or the like, i would like to exercise some DIY'ism and create my own (sort of) norm just for the sake of an interesting exercise.
It consists of more than hundred of subjects in EO conditions, has info regarding age/sex and even results of some psychological tests. I will select a relative age group (my age +/- 3-5 years), exclude outliers in terms of bad performance on tests and will import them one by one into BioEra and extract metrics like coherence, phase, amplitude, etc for locations that coincide with networks that i suspect are poor performing in me. This is just an interesting project/exercise and i understand that this has nothing to do with proper database generation that was used in NeuroGuide, for example.
The question is. This particular dataset uses 'average reference'. From what i understand, in this kind of montage earlobes are not used and all the signals are referenced to an average of all other signals? As i get it, this kind of montage can be re-referenced using tools such as EEGLab. Since this dataset does not contain recordings of earlobes (A1/A2) i can not use 'linked ears' for re-referencing. What if i reference it to, say, Cz and then use Cz in my trainings as well? I know that linked ears is preferred in NFB, but since i'll compare my personal metrics using the same montage as was used during creation of the normative DB based on this particular dataset and it was reconfigured to use Cz as reference, won't i be comparing apples to apples? Or there are some other reasons why other reference montages are not useful?
Another common "off 10-20 map" location used for Ground or Reference is AFz, just above FPz. Or CPz. This allows you to still train at all the normal 10-20 sites. And Cz is a common training site, as well as C3, C4.
Comments
Ben, hi.
Not that I am aware of. The standard packages such as Neuroguide, BrainDx, HBI, qEEG-Pro, etc., all seemingly cost so much because of the time and effort that went into creating and maintaining the carefully controlled normative databases. These typically contain 100's to 1000's of controlled 'normative' client QEEGs. That are then averaged to compute the z-score deviation for the subject in question.
Nice comparison page, pdf links on that page are a bit out of date, new ones are below.
https://qeeg.pro/database/
https://qeeg.pro/wp-content/uploads/2017/09/ISNR2015_Keizer.pdf
https://qeeg.pro/wp-content/uploads/2017/08/Connectivity_comparisons.pdf
Pete van Deusen's system avoids using QEEGs. By instead analyzing for certain patterns he has determined over decades of experience. It works. Once the assessment recording is made, Pete's software processes that into recommended protocols. These are NOT z-score training protocols. But instead various band dynamics based on relative amplitudes, coherence, etc.
https://brain-trainer.com/
Another alternate system which has good success is the ClinicalQ, which would be considered open source.
https://www.google.com/search?q=clinicalq+neurofeedback
Regards, William
Here is a good article on ClinicalQ, by Dr. Paul Swingle, the creator.
Clinical versus Normative Databases: Case Studies of Clinical Q Assessments
https://www.researchgate.net/publication/318573194_Clinical_versus_Normative_Databases_Case_Studies_of_Clinical_Q_Assessments
That's too bad there is no open source QEEG dataset available. Perhaps it is time to start compiling one?
Alex, it is possible that a university or research institute could create a public open access QEEG database. But such a project would be costly and take much time. There are these resources of anonymized datasets:
https://openneuro.org/
https://github.com/meagmohit/EEG-Datasets
Existing QEEG databases contain 100's to 1000's of extremely carefully controlled 'normative' client QEEGs. This cannot be done by hobbyists or distributed research groups. But instead require a centralized funded system of continuous support.
If you are interested in alternatives, see the previous comment links.
Regards, William
I know this is an old thread, but i've recently been fascinated by the idea of neurofeedback training relative to the normative database and since i am currently too poor to allow myself a QEEG based system like NeuroGuide or the like, i would like to exercise some DIY'ism and create my own (sort of) norm just for the sake of an interesting exercise.
I have found this dataset - https://openneuro.org/datasets/ds003775/versions/1.0.0
It consists of more than hundred of subjects in EO conditions, has info regarding age/sex and even results of some psychological tests. I will select a relative age group (my age +/- 3-5 years), exclude outliers in terms of bad performance on tests and will import them one by one into BioEra and extract metrics like coherence, phase, amplitude, etc for locations that coincide with networks that i suspect are poor performing in me. This is just an interesting project/exercise and i understand that this has nothing to do with proper database generation that was used in NeuroGuide, for example.
The question is. This particular dataset uses 'average reference'. From what i understand, in this kind of montage earlobes are not used and all the signals are referenced to an average of all other signals? As i get it, this kind of montage can be re-referenced using tools such as EEGLab. Since this dataset does not contain recordings of earlobes (A1/A2) i can not use 'linked ears' for re-referencing. What if i reference it to, say, Cz and then use Cz in my trainings as well? I know that linked ears is preferred in NFB, but since i'll compare my personal metrics using the same montage as was used during creation of the normative DB based on this particular dataset and it was reconfigured to use Cz as reference, won't i be comparing apples to apples? Or there are some other reasons why other reference montages are not useful?
Hope i was clear enough
Another common "off 10-20 map" location used for Ground or Reference is AFz, just above FPz. Or CPz. This allows you to still train at all the normal 10-20 sites. And Cz is a common training site, as well as C3, C4.
https://www.google.com/search?q=10-10+eeg+map
Regards, William