It has been observed that certain qualitative and quantitative features of visual objects presented to both arthropod and vertebrate visual systems give rise to neural signals that are specific for those features. Features in this context are simple, contrasting, geometric properties of the object, which can be moving or stationary. That is, the feature of an object can involve shape, contrast, and a velocity vector. For example, a type of neuron in the medulla of the OL of the lubber grasshopper will significantly increase its firing rate if a long black stripe is moved from front to back over the ipsilateral CE. Motion from back to front silences the unit, and a black or white spot given the same motion has a negligible effect on the random, background firing of the neuron. Also, motion of the stripe at right angles to its preferred direction produces little response. Such units that respond only to long stripes moving in a preferred direction have been called "vector edge units" (Northrop, 1974).
Prewired feature extraction operations offer neural economy in simple nervous systems. It will be seen that the features extracted have survival value for the animal, particularly in the detection of prey, avoidance of predators, stabilization of flight, and perhaps even in the location of a mate.
Examples of a visual neural array are the retina in the case of vertebrate eyes, and the OLs of a CE, including the lamina ganglionaris, the medulla, and the lobula when arthropod eyes are considered. Both retinas and OLs are highly organized neural networks dedicated to visual data preprocessing and feature extraction..
It is supposed that most of the visual feature extraction used by arthropods occurs in their OLs; the OLs probably serve most of the functions of the vertebrate retina, tectum, and visual cortex. A number of well-defined nerve tracts carry information to and from the OLs. Bullock and Horridge (1965) describe such tracts in the bee and locust. For example:
1. There is a tract of fibers leaving the back of the lamina ganglionaris that runs tangential to it toward the front of the lobe, thence to a primary optic association center in the dorsal posterior part of the protocerebrum.
2. Two tracts arise in the medulla; one leaves from the dorsal edge, and the other from the ventral edge. Both tracts run to the calyces of the corpora pedunculata.
3. The anterior optic tract runs from the lobula to the optic tubercle on the anterior brain, thence to the ipsilateral corpora pedunculata.
4. The superior medial optic tracts are short and long tracts from the lobula to the a and p lobes of the corpora pedunculata.
5. The inferior medial optical tract has fibers running from the lobula to the ventral center of the brain.
6. Decussating tracts between left and right OLs include two tracts between dorsal and ventral medullas, and one between left and right lobulas.
7. Giant descending neurons from the lobula project into giant ventral nerve cord fibers. (These are the DCMD and DIMD VNC neurons described by Rowell, et al., 1977.)
It is apparent that the arthropod OLs are "well-connected" to the other components of the animal's CNS. Thus, it is not surprising to find signals in OL neurons that respond to visual stimuli, as well as to other sensory modalities (touch, sound, airflow). These are the multimodal units described in Locusta by Horridge et al. (1965), in moths by Blest and Collett (1965), and in Romalea by Northrop and Guignon (1970). Other visual units (N3 vector units) described by Northrop (1974) responded to moving edge visual stimuli and stimulation of aerodynamic hair patches (wind velocity sensors) on the head. More will be said about these units below.
Early workers examining CE, OL unit responses explored a variety of arthropod visual systems. Burtt and Catton (1960), Ishikawa (1962), Schiff (1963) and Dingle and Fox (1966) examined single unit OL unit responses to simple changes in illumination in locusts, silkworm moths, mantis shrimp, and crickets, respectively. Not surprisingly, units were found responding to ON, OFF, and ON/OFF of general illumination, as well as sustaining units that fired steadily in steady-state light or dark. Because an eye operates in a complex visual environment where many object features can be found (spatial frequency, contrast, color, relative velocity, etc.), some workers have explored how visual systems respond to certain object features, as well as to changes in overall and local illumination. Both approaches, i.e., the study of responses to simple changes in illumination, and of responses to certain object features are necessary to provide a complete understanding of how the CE visual system works.
The seminal work on visual feature extraction was done by Lettvin et al. (1959) on the frog retina. What was new in their work were three important findings: (1) Object size (hence spatial frequency) was a factor in determining the strength of response of some ganglion cells. (2) Object motion, and direction of motion were also response factors for some ganglion cell units. (3) Both size and motion were factors, as well. Other workers (see Section 6.3), have extended their approach to a variety of other vertebrate retinas and CNSs.
Horridge et al. (1965) were the first workers to explore systematically feature extraction in the CE system of Locusta, in which they ambitiously classified some 20 types of OL units. The feature extraction approach has also been applied to flies (diptera) by Mimura (1970; 1971; 1972; 1974), McCann and Dill (1969), Bishop et al. (1968), and DeVoe (1980), and to lepidoptera (moths and butterflies) by Swihart (1968) and Collett (1971). Wiersma and colleagues (254, 257) have also examined the CE visual system of crabs and other crustaceans from a feature extraction point of view. These works will be discussed below.
By combining the properties of responses to changes in illumination with the detailed responses to moving objects, several of the workers mentioned above have identified large numbers of units classes (20 for Horridge et al., 1965 in Locusta; 12 OL classes by McCann and Dill, 1969, in flies). The author, working with the grasshopper Romalea microptera, adopted a more parsimonious approach and designated eight OL units with distinct properties based on form and motion, as well as ON and OFF responses These are described in detail below.
Was this article helpful?
This guide will help millions of people understand this condition so that they can take control of their lives and make informed decisions. The ebook covers information on a vast number of different types of neuropathy. In addition, it will be a useful resource for their families, caregivers, and health care providers.