Computerized assessment in neuropsychiatry using CANTAB: discussion paper.

This paper illustrates how recent developments in computing technology can be exploited for the neuropsychological assessment ofneurodegenerative diseases. Computerizedtests have many advantages over traditional paper and pencil tests, including precision, speed an-d reliability. For example, it is possible to measure -response latencies to' millisecond accuracy, necessarfor reaction time studies. Furthermore, the 6iomputer can 'be used to give feedback to the subjet in a iystematic 'and objective fashion. These generl' desgir 'features have now been utilized extengiMvelf inis set of tests knownas the Cambridge Neuropsycho6gic1a Test Automated Battery (CANTAB). -Other iijortant design1features of these tests are: (1) They are designed to test different aspects of mental functioning so that a profile ofperormance can be constructed'for a particular patient group. (2) Where appropriate they are graded in difficulty in order to assess abroad range ofcognitive ability. (3) They employ non-verbal stimuli and require nonverbal responses. This, of course, is essential for patients with language impairment, but is also an important advantage for -cross national studies. (4) They are designed to be-visually attractive and interesting, using positive feedback which has a 'game-like' quality and maintains motivation. CANTAB consists ofthree separate batteries oftests measuring visual memory, attention and planning. They are designed to run on an Acorn BBC Master Microcomputer and more recently on IBM PC and compatible machines. In both cases the programs employ high resolution graphics and responses are made using a touch sensitive screen. The theoretical rationale of the tests is based on two main themes: (i) adapting those tests of animal neuropsychology that have proved useful in establishing the neural substrates" of certain types of cognitive function and; (ii)'.undertaking a componential analysis ofthe processes comprising particular forms of cognitive function; for example spatial planning may involve the ability to store and sequence particular spatial representations, as well as manipulate them in working memory. A patient could fail a planning test because of impairments in any one of these constituent functions. The tests are comprehensive in terms of the range of functions they assess, and have been shown' to be sensitive to deficits, as well as progressive decline, in patients early in the course of dentia of the