A machine learning approach for quantifying age-related histological changes in the mouse kidney

Area of Science:

• Computational pathology
• Gerontology
• Histology

Background:

• Quantifying age-related histological changes is crucial for evaluating interventions targeting health span.
• Existing methods may lack the reproducibility and throughput needed for large-scale aging studies.

Purpose of the Study:

• To develop and validate a machine learning architecture for quantifying age-related changes in mouse kidney histology.
• To provide reproducible and quantifiable age scores for histological samples.

Main Methods:

• A machine learning architecture was trained to detect and quantify aging in mouse kidney histological samples.
• Model validation was performed using held-out data and compared against pathologist scores from the Geropathology Research Network aging grading scheme.
• Trained classifiers for Hematoxylin and Eosin (H&E)-stained slides were provided, along with tutorials for creating additional classifiers.

Main Results:

• The machine learning model accurately quantified aging-related histological changes in mouse kidneys.
• Model performance correlated well with expert pathologist assessments.
• The approach provided reproducible and quantifiable age scores, independent of specific lesions.

Conclusions:

• The developed machine learning architecture offers a high-throughput solution for quantifying mouse aging studies, particularly in kidney tissues.
• This tool facilitates the evaluation of interventions aimed at extending health span.
• The provided resources enable broader application in histological analysis and classifier development.