Functional neuroimaging of implicit and explicit reading in patients with pure alexia

Pure alexia is an acquired disorder characterized by severely impaired reading abilities with relative sparing of all other linguistic abilities, including writing and the recognition of orally spelled words. As a compensatory strategy, patients with pure alexia typically name individual letters within words aloud, or subvocally, in order to identify words. "Letter-by-letter reading" allows patients to gain access to the lexical representation of words. However, recent investigations of patients with pure alexia suggest that whole word identification may be observed in these patients. This type of lexical access has been demonstrated when patients are presented words at a rate too fast to allow for letter-by-letter reading and patients are instructed to make their best guess about the lexical status of words. Patients have shown above-chance accuracy in word identification using such implicit reading methods. The present study examined the behavioral performance, as well as the neuroanatomical correlates underlying explicit and implicit reading in two patients with pure alexia and four neurologically-intact participants. Previous investigators have postulated left hemisphere involvement in patients with pure alexia during both explicit and implicit processing. Alternatively, others have suggested that while the left hemisphere is involved during explicit reading, the right hemisphere is responsible for implicit reading in patients with pure alexia. These hypotheses were directly tested by employing functional magnetic resonance imaging (fMRI). Functional and behavioral data were collected during two lexical decision tests; one involved suprathreshold exposure in order to measure explicit reading capacities, and the other utilized subthreshold exposure to examine implicit reading ability. In addition, the two patients with pure alexia performed a third experiment that investigated implicit semantic classification of words and pictures. Results suggest support for both the left and right hemisphere hypotheses with regard to explicit reading, but strong evidence against the right hemisphere hypothesis during implicit reading. In addition, semantic classification results were mixed in that one patient with pure alexia demonstrated exclusive left hemisphere involvement, while the other demonstrated primarily right hemisphere activation. These finding provide important information about hemisphere involvement during explicit and implicit processing after left hemisphere damage.