#AI reads Urine#Deep learning analysis of urine-derived stem cell mitochondrial morphology as a non-invasive Alzheimer's disease biomarker
Published 16 February, 2026
Addressing the lack of convenient and non-invasive biomarkers for mitochondrial assessment in Alzheimer’s disease, this study developed an artificial intelligence framework that utilizes mitochondrial fluorescence imaging of live urine-derived stem cells to explore differences between individuals with cognitive impairment (including those with Alzheimer’s disease and mild cognitive impairment) and cognitively normal subjects. First, mitochondrial fluorescence images of living HeLa cells were segmented, and two binary classification models based on the ResNet-18 convolutional neural network were trained to identify mitochondrial hyperfission and hyperfusion relative to normal morphology. These models demonstrated robust performance in detecting intermediate mitochondrial states during validation. When applied to urine-derived stem cells, the system effectively distinguished mitochondrial patterns associated with cognitive impairment. The study recruited 26 participants divided into three groups, and through steps such as collecting urine samples, culturing stem cells, and acquiring mitochondrial fluorescence images, the results showed that this method achieved an area under the curve of 0.952 for predicting Alzheimer’s disease and 0.871 for predicting mild cognitive impairment, highlighting its potential for the early detection of Alzheimer’s disease. However, further validation in larger, independent cohorts is still needed.
Neurotherapeutics. 2025 Dec 16:e00813. doi: 10.1016/j.neurot.2025.e00813.
Youhe Gao
Statement: During the preparation of this work the author(s) used Doubao / AI reading for summarizing the content. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article.
For earlier AI Reads Urine articles：https://www.keaipublishing.com/en/journals/advances-in-biomarker-sciences-and-technology/ai-reads-urine/