# Brain subtle anomaly detection based on auto-encoders latent space analysis : application to de novo parkinson patients

## TL;DR

This paper introduces new unsupervised criteria based on auto-encoder latent space analysis for detecting subtle brain anomalies, demonstrating improved performance in Parkinson's disease classification.

## Contribution

It proposes two novel detection criteria derived from multivariate analysis that enhance anomaly detection in brain imaging, especially for subtle lesions.

## Key findings

- New criteria outperform traditional reconstruction error methods.
- Performance comparable or better than supervised methods.
- Effective in detecting subtle brain anomalies in Parkinson's disease.

## Abstract

Neural network-based anomaly detection remains challenging in clinical applications with little or no supervised information and subtle anomalies such as hardly visible brain lesions. Among unsupervised methods, patch-based auto-encoders with their efficient representation power provided by their latent space, have shown good results for visible lesion detection. However, the commonly used reconstruction error criterion may limit their performance when facing less obvious lesions. In this work, we design two alternative detection criteria. They are derived from multivariate analysis and can more directly capture information from latent space representations. Their performance compares favorably with two additional supervised learning methods, on a difficult de novo Parkinson Disease (PD) classification task.