Independent component analysis for biomedical signals

Abstract

Independent component analysis (ICA) is increasing in popularity in the field of biomedical signal processing. It is generally used when it is required to separate measured multi-channel biomedical signals into their constituent underlying components. The use of ICA has been facilitated in part by the free availability of toolboxes that implement popular flavours of the techniques. Fundamentally ICA in biomedicine involves the extraction and separation of statistically independent sources underlying multiple measurements of biomedical signals. Technical advances in algorithmic developments implementing ICA are reviewed along with new directions in the field. These advances are specifically summarized with applications to biomedical signals in mind. The basic assumptions that are made when applying ICA are discussed, along with their implications when applied particularly to biomedical signals. ICA as a specific embodiment of blind source separation (BSS) is also discussed, and as a consequence the criterion used for establishing independence between sources is reviewed and this leads to the introduction of ICA/BSS techniques based on time, frequency and joint time-frequency decomposition of the data. Finally, advanced implementations of ICA are illustrated as applied to neurophysiologic signals in the form of electro-magnetic brain signals data.

Blind source separation to enhance spectral and non-linear features of magnetoencephalogram recordings. Application to Alzheimer's disease

“…BSS techniques estimate the set of n unknown components, s(t) = [s 1 (t), …, s n (t)] T , where T denotes transposition, which were linearly mixed by the full rank m × n matrix A (m ≥ n) to form m temporally and spatially correlated recordings, x(t) = [x 1 (t), …, x m (t)] T [15,16].…”

“…Several assumptions are needed to estimate s(t) and A from x(t) [16,17]. The most important one is that the components are mutually independent or, alternatively, that they should be decorrelated at any time delay.…”

“…BSS methods estimate the underlying components of the EEG and MEG signals without a priori information about those components (i.e., the components themselves and the process that produced the observed recordings are unknown) [15][16][17]. Since these techniques isolate specific physiological activities into different components, they have been used to reject artefacts [16][17][18][19].…”

“…Since these techniques isolate specific physiological activities into different components, they have been used to reject artefacts [16][17][18][19]. This application is based on the fact that BSS isolates the artefacts into a few components.…”

Neuronal oscillations in the EEG under varying cognitive load: A comparative study between slow waves and faster oscillations

“…ICA (in its basic form) is a blind source separation (BSS) technique that extracts statistically independent and spatially distinct components (ICs) from a set of measured (mixed) signals, without using any additional information about the underlying (unknown) sources (James and Hesse, 2005;Ziehe and Müller, 1998;Demanuele et al, 2009). The extracted band-limited rest and task ICs are then filtered in four (slow, delta, theta and alpha) frequency bands, and their amplitude and phase features compared by means of a neural network methodology.…”