On the use of Pairwise Distance Learning for Brain Signal Classification with Limited Observations
This work addresses the challenge of diagnosing neuronal diseases like Schizophrenia with limited clinical trial data, representing an incremental advance in brain signal classification.
The paper tackled the problem of classifying Schizophrenia from brain signals with limited data by proposing a pairwise distance learning approach using a Siamese neural network, resulting in a 10 percentage point improvement in accuracy and sensitivity compared to spectral features.
The increasing access to brain signal data using electroencephalography creates new opportunities to study electrophysiological brain activity and perform ambulatory diagnoses of neuronal diseases. This work proposes a pairwise distance learning approach for Schizophrenia classification relying on the spectral properties of the signal. Given the limited number of observations (i.e. the case and/or control individuals) in clinical trials, we propose a Siamese neural network architecture to learn a discriminative feature space from pairwise combinations of observations per channel. In this way, the multivariate order of the signal is used as a form of data augmentation, further supporting the network generalization ability. Convolutional layers with parameters learned under a cosine contrastive loss are proposed to adequately explore spectral images derived from the brain signal. Results on a case-control population show that the features extracted using the proposed neural network lead to an improved Schizophrenia diagnosis (+10pp in accuracy and sensitivity) against spectral features, thus suggesting the existence of non-trivial, discriminative electrophysiological brain patterns.