[Eeglablist] PSI correction and new perspective on clinical assessment and scalp EEG validation

[נתי שטרן]

Dear EEGLAB community,

Following my earlier PSI message, I am sharing a new perspective and
proposed validation protocol that corrects the clinical and measurement
framing:

*Reframing the Psychiatric State Index Through Clinical Assessment and
Measurement Validation*
Full manuscript with references
<https://urldefense.com/v3/__https://github.com/netanelcyber/psi-qeeg-bis/blob/18e22b6dd4ddc8747b6c2034cc7f457ec1aff28f/docs/research/PSI_CLINICAL_PERSPECTIVE.md__;!!Mih3wA!EQKFbYNsgHm5eiuJjPjvf9KpWr8fIkJo8PC09Q1COsX_DpS-teuQBRAAmdIx8o9jvvpAe1lwWDy7LkF5HoU$ >

This is an unreviewed manuscript for discussion; psychiatric review has
been requested and remains pending. The present software supports offline
research analyses. Available datasets do not establish imminent meltdown or
psychosis prediction.

The proposed primary human analysis excludes scalp gamma and scalp
theta–gamma PAC. Any secondary interpretation would require synchronized
EMG/EOG/movement measurements and expert artifact review. Numerical QC,
notch filtering or a finite spectral estimate does not establish neural
origin; entropy and spectral edge measures also require scrutiny.
Intracranial LFP findings remain a separate measurement domain.

I would welcome methodological advice on practical synchronized artifact
measurements, prespecified raw-signal review, and reporting unavailable
data during active movement/distress in EEG workflows. The article also
separates full journal evidence from preprints and conference abstracts,
and clinical assessment from numerical benchmarks.

Code and documented corrections
<https://urldefense.com/v3/__https://github.com/netanelcyber/psi-qeeg-bis/pull/3__;!!Mih3wA!EQKFbYNsgHm5eiuJjPjvf9KpWr8fIkJo8PC09Q1COsX_DpS-teuQBRAAmdIx8o9jvvpAe1lwWDy7Ttbbnmc$ >. No attachment, in
accordance with list guidance.

Best regards,
Netanel Stern


On Wed, 07 Oct 2026 22:23:00 -0700, נתי שטרן <nsh531 at gmail.com> wrote:

Dear EEGLAB community, I am sharing a perspective article proposing a novel
wearable EEG device — the Psychiatric State Index (PSI) — designed to
differentiate autistic meltdowns from psychotic episodes in real time using
EEG-derived biomarkers and machine learning. As an autistic researcher, I
combine lived experience of sensory overload with computational
neuroscience to propose two composite indices computed from an 8-channel
montage: - Sensory Overload Index (SOI): alpha stability, gating
efficiency, gamma load, frontoparietal coherence, autonomic surge - Reality
Disconnection Index (RDI): temporal beta/gamma CSD, delta/alpha ratio,
bifrontal bicoherence, hemispheric asymmetry, frontotemporal connectivity
The article also addresses BIS monitor calibration failures in
neuropsychiatric populations (autism BIS=72 vs neurotypical 97 at
equivalent awareness), proposes personal baseline calibration with z-score
normalization, and includes worked clinical vignettes with epoch-by-epoch
numerical calculations. Full article (PDF) available upon request. Netanel
Stern Independent Security Researcher and Clinical AI Developer Rosh
HaAyin, Israel