Evaluation of recognition ability of altered images in relation to age
KOTSAVASILOGLOU C., TRAKA C., BALOYANNIS S.

One of the most important properties of the human brain is its ability to recognize objects based only on a subset of the necessary information. This ability, in other words the way CNS intakes, stores and elaborates information, is inherent of the brain function. For example an image arrives to the occipital lobe as a stream of elementary information. All this elementary information is recombined in the various associative areas of the occipital lobe in order to create the perception of one object. Even when this stream carries only a subset of the information related to that object mixed with other unrelated information, the brain can still recognize the object.

In this paper we investigate the decline of the ability to recognize altered images in relation to the age. For this purpose a software tool was developed in our clinic. This software tool contains a library of images, words and sounds that are related to each other. The user of this software tool can create a new test by setting up various parameters such as the information (images, words and sounds), that will be administered to the subjects, the duration of appearance of every information unit and the quantity of information units. Other important settings are the kind and the intensity of the alteration that can be applied in every information unit. The alteration of an information unit is used in order to make it difficult for a person to recognize.

Every test administered in this work consists of a sequence of images shown to the subject on the screen of a computer. For every test the images were altered in a specific manner. The subjects had to recognize and name the main theme of every altered image in a few seconds (4-10"). The computer program stores the response as positive (recognition) or negative (no recognition), for every altered image, for every test, for every subject.

Four of these tests were given to 52 healthy volunteers, aged 20 to 80 years old. The subjects were divided into 3 groups based on age.

Group I: 23 subjects, 20-39 years old. Group II: 16 subjects, 40-59 years old. Group III: 13 subjects aged over 60.

The subjects of Group III had a detailed neurological examination so that any subclinical manifestation involving the brain could be excluded. There were no pathological findings.

The statistical analysis of the results showed statistically significant differences between the groups in two of the tests. These two tests had the major degree of alteration of their images. In one of these two tests the significant difference was higher (0,01 and 0,05). There is a proportional relationship between the alteration of the image and the statistically significant difference. Group I had the best performances in these two tests, whereas group III had the worst performances. The results suggest that there is a clear decline in the ability to recognize altered images with the progression of age. This decline becomes evident after the age of 40 years old and is more marked after 60 years old.

It is well known that ageing is associated with gradual loss of various abilities in cognitive tasks. In these tests we investigate a specific area of the occipital lobe and the posterior areas of the temporal lobe. The differences in the performances in the tests, between persons of different age, may suggest a gradual loss of synaptic connections in these cortical areas.

The tests we used in this work were sensitive for detecting the differences in the pattern recognition ability during those specific periods of age in healthy people. It is well known that cognitive decline in persons with dementia starts many years before the first diagnosis. This kind of tests may prove very useful for the evaluation of persons with mild cognitive disorders.

Key words: Ageing, pattern recognition, visual perception, neuropsychological tests.