Background Electroencephalography (EEG) provides a low-cost, non-invasive view of millisecond-scale brain dynamics; however, its clinical value depends on reliable probabilities and deployment-aware evaluation, rather than accuracy alone. Objective This study aimed to deliver a simple EEG approach for Alzheimer’s Diagnosis (AD) that returns calibrated subject-level probabilities and reports pre-specified clinical operating points, all under subject-wise, leakage-free validation. Methods Resting-state, eyes-closed EEG from OpenNeuro ds004504 was pre-processed (resampled to 128 Hz, 0.5–45 Hz band-pass, 50 Hz notch, average reference), segmented into 8-s epochs with 4-s overlap, and quality-controlled (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\ge\)</EquationSource> </InlineEquation> 20 valid epochs per subject). Wavelet Scattering Transform features were then extracted under two configurations (<InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(J=6/7, Q=8\)</EquationSource> </InlineEquation>), pooled to lobar regions with mild weights, and aggregated into subject-level statistics. Feature learning used an <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(\ell _1\)</EquationSource> </InlineEquation>-penalized logistic selector followed by an <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(\ell _2\)</EquationSource> </InlineEquation>-regularized logistic classifier, with performance estimated via 5-fold GroupKFold to produce subject-wise out-of-fold (OOF) logits. Base probabilities were calibrated using isotonic regression, combined through linear ensembling, and then recalibrated. We summarized discrimination (AUC, PR-AUC), calibration (Brier score, ECE), threshold behavior, and clinically oriented operating points (Sens@Spec, Spec@Sens); uncertainty was quantified using bootstrap confidence intervals derived from OOF predictions. Results After quality control, 59 subjects (31 AD/28 HC; 6,957 epochs) remained. The calibrated ensemble achieved an AUC of 0.930 and a PR-AUC of 0.931; the Brier score improved from 0.107 to 0.102, and the ECE decreased from 0.051 to <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(\sim 0.000\)</EquationSource> </InlineEquation>. Bootstrap resampling (<InlineEquation ID="IEq6"> <EquationSource Format="TEX">\(N=1000\)</EquationSource> </InlineEquation>) yielded mean AUC <InlineEquation ID="IEq7"> <EquationSource Format="TEX">\(=0.932\)</EquationSource> </InlineEquation> (95% CI <InlineEquation ID="IEq8"> <EquationSource Format="TEX">\([0.864,\,0.980]\)</EquationSource> </InlineEquation>). Thresholds <InlineEquation ID="IEq9"> <EquationSource Format="TEX">\(t=0.380\)</EquationSource> </InlineEquation> (F1-optimal) and <InlineEquation ID="IEq10"> <EquationSource Format="TEX">\(t=0.5\)</EquationSource> </InlineEquation> produced identical <InlineEquation ID="IEq11"> <EquationSource Format="TEX">\(F_1=0.839\)</EquationSource> </InlineEquation> and accuracy <InlineEquation ID="IEq12"> <EquationSource Format="TEX">\(=0.831\)</EquationSource> </InlineEquation>. Clinically, <InlineEquation ID="IEq13"> <EquationSource Format="TEX">\(t=0.375\)</EquationSource> </InlineEquation> prioritized sensitivity (0.935), whereas <InlineEquation ID="IEq14"> <EquationSource Format="TEX">\(t=0.667\)</EquationSource> </InlineEquation> prioritized specificity (1.00; no false positives). Conclusions The lightweight, interpretable EEG workflow produced reliable, calibrated probabilities under subject-wise, leakage-free evaluation and supported explicit clinical operating points. While external, multi-center validation remains necessary, these findings support probability-aware EEG decision support for AD.

Alzheimer’s diagnosis from EEG with reliable probabilities: subject-wise, leakage-free evaluation and isotonic calibration

期刊论文
出版模式：

开放获取
发表日期：

2025年12月1日

查看全文

Hamed Shamsi

摘要

Background

Electroencephalography (EEG) provides a low-cost, non-invasive view of millisecond-scale brain dynamics; however, its clinical value depends on reliable probabilities and deployment-aware evaluation, rather than accuracy alone.

Objective

This study aimed to deliver a simple EEG approach for Alzheimer’s Diagnosis (AD) that returns calibrated subject-level probabilities and reports pre-specified clinical operating points, all under subject-wise, leakage-free validation.

Methods

Resting-state, eyes-closed EEG from OpenNeuro ds004504 was pre-processed (resampled to 128 Hz, 0.5–45 Hz band-pass, 50 Hz notch, average reference), segmented into 8-s epochs with 4-s overlap, and quality-controlled ( \(\ge\) 20 valid epochs per subject). Wavelet Scattering Transform features were then extracted under two configurations ( \(J=6/7, Q=8\) ), pooled to lobar regions with mild weights, and aggregated into subject-level statistics. Feature learning used an \(\ell _1\) -penalized logistic selector followed by an \(\ell _2\) -regularized logistic classifier, with performance estimated via 5-fold GroupKFold to produce subject-wise out-of-fold (OOF) logits. Base probabilities were calibrated using isotonic regression, combined through linear ensembling, and then recalibrated. We summarized discrimination (AUC, PR-AUC), calibration (Brier score, ECE), threshold behavior, and clinically oriented operating points (Sens@Spec, Spec@Sens); uncertainty was quantified using bootstrap confidence intervals derived from OOF predictions.

Results

After quality control, 59 subjects (31 AD/28 HC; 6,957 epochs) remained. The calibrated ensemble achieved an AUC of 0.930 and a PR-AUC of 0.931; the Brier score improved from 0.107 to 0.102, and the ECE decreased from 0.051 to \(\sim 0.000\) . Bootstrap resampling ( \(N=1000\) ) yielded mean AUC \(=0.932\) (95% CI \([0.864,\,0.980]\) ). Thresholds \(t=0.380\) (F1-optimal) and \(t=0.5\) produced identical \(F_1=0.839\) and accuracy \(=0.831\) . Clinically, \(t=0.375\) prioritized sensitivity (0.935), whereas \(t=0.667\) prioritized specificity (1.00; no false positives).

Conclusions

The lightweight, interpretable EEG workflow produced reliable, calibrated probabilities under subject-wise, leakage-free evaluation and supported explicit clinical operating points. While external, multi-center validation remains necessary, these findings support probability-aware EEG decision support for AD.