According to the definition proposed by Baddeley (1) working memory is considered as a construct of a
temporary storage and an executive system. It can be further decomposed in an attentional control system, the
“central executive” and three subsidiary slave systems – the phonological loop, the visuospatial sketchpad and
the episodic buffer. The idea of storage systems which support the central executive is widely accepted in
literature. However, the organization and localization of the central executive or alternative systems (Stuss et al.
(2)) remains unclear.
This study investigated the mapping of working memory and attention to brain anatomy to prove controversies
about the link between brain function and structure. I used high-resolution brain structure magnetic resonance
imaging (MRI) data acquired from 32 patients with chronic stroke lesions. I applied a novel technique for
automated lesion detection in the framework of voxel-based morphometry (VBM) to correlate estimates of
grey/white matter volume with behavioral data from well-established test battery – the California Verbal
Learning Test (CVLT), Testbatterie zur Aufmerksamkeitsprüfung (TAP) and Wechsler Memory Scale-Revised
(WMS-R). Using a feature reduction technique, called principal component analysis (PCA), I looked for analytical
strategies to investigate principal behavioral tendencies from the plenitude of partially correlated subscores
within and between the applied tests. Here, I confirm empirically the established separation between storage
systems and a possible central executive. However, we failed to demonstrate a core WM network by using
currently available analytical techniques, which we ascribe to methodological issues. The non-significant results
in the principal component analysis are mostly due to the low statistical power of our sample (n=32). However,
the idea of using a PCA to reunion WM processes stays a potential access to investigate brain networks with
multiple functions.