NPJ Digital Medicine

Brain-wave model predicted Alzheimer’s progression in people

In a cohort of 459 people followed for 10 years, model-derived brain-wave measures predicted clinical progression and improved predictions when added to established biomarkers.

Figure 1. Study outline: diagnostic and prognostic predictions based on DADD-derived digital biomarkers.
Open the figure at full size
Figure 1. Study outline: diagnostic and prognostic predictions based on DADD-derived digital biomarkers.Study outline: diagnostic and prognostic predictions based on DADD-derived digital biomarkers.Amato et al.

Our read

Biomarkers
Studied:
Cohort study of 459 people

Editor of record: Boris Djordjevic · AgingResearch.News

Follow @agingbrief

The 459 participants in a cohort study had either subjective cognitive decline—self-reported worsening of thinking—or mild cognitive impairment, involving mild memory or thinking problems. The Digital Alzheimer’s Disease Diagnosis computer model extracted measures from electroencephalography, which records the brain’s electrical activity.

The team compared its predictions of clinical progression over 10 years with established Alzheimer’s indicators, including genetic information, brain scans and spinal-fluid markers. The model-derived measures outperformed standard brain-wave measures. They matched established Alzheimer’s indicators in predicting the risk and timing of clinical progression and outperformed them among participants with subjective cognitive decline. Combining the established indicators with the model-derived measures improved prediction accuracy.

Why it matters

Alzheimer’s prognosis often relies on tests that are invasive or require expensive equipment. Noninvasive brain recordings could make assessment of future progression more accessible.

Caveats

Prediction accuracy in these participants does not establish how well the model would perform in people without subjective cognitive decline or mild cognitive impairment.

From the paper

Participants

The second cohort (recruited in Florence, Italy, see “Methods”) comprised 102 participants, 58 SCD and 44 MCI.

Results, DADD-derived digital biomarkers generalize to independent cohorts

Quoted word for word from the full text on nature.com.

Comments

Be the first to comment.