PubMed Health⌕ Search

Biomedical subjects

J K Tsotsos

Publications and source records attributed to J K Tsotsos.

6 recordsLinked to original sources
PubMed2006

Direct neurophysiological evidence for spatial suppression surrounding the focus of attention in vision.

The spatial focus of attention has traditionally been envisioned as a simple spatial gradient of enhanced activity that falls off monotonically with increasing distance. Here, we show with high-density magnetoencephalographic recordings in human observers that the focus of attention is not a simple monotonic gradient but instead contains an excitatory peak surrounded by a narrow inhibitory region. To demonstrate this center-surround profile, we asked subjects to focus attention onto a color pop-out target and then presented probe stimuli at various distances from the target. We observed that the electromagnetic response to the probe was enhanced when the probe was presented at the location of the target, but the probe response was suppressed in a narrow zone surrounding the target and then recovered at more distant locations. Withdrawing attention from the pop-out target by engaging observers in a demanding foveal task eliminated this pattern, confirming a truly attention-driven effect. These results indicate that neural enhancement and suppression coexist in a spatially structured manner that is optimal to attenuate the most deleterious noise during visual object identification.

Adult↗
PubMed1986

Knowledge-based landmarking of cephalograms.

Orthodontists have defined a certain number of characteristic points, or landmarks, on X-ray images of the human skull which are used to study growth or as a diagnostic aid. This work presents the first step toward an automatic extraction of these points. They are defined with respect to particular lines which are retrieved first. The original image is preprocessed with a prefiltering operator (median filter) followed by an edge detector (Mero-Vassy operator). A knowledge-based line-following algorithm is subsequently applied, involving a production system with organized sets of rules and a simple interpreter. The a priori knowledge implemented in the algorithm must take into account the fact that the lines represent biological shapes and can vary considerably from one patient to the next. The performance of the algorithm is judged with the help of objective quality criteria. Determination of the exact shapes of the lines allows the computation of the positions of the landmarks.

Artificial Intelligence↗
PubMed1986

Ambient illumination and the determination of material changes.

The task of distinguishing material changes from shadow boundaries in chromatic images is discussed. Although there have been previous attempts at providing solutions to this problem, the assumptions that were adopted were too restrictive. Using a simple reflection model, we show that the ambient illumination cannot be assumed to have the same spectral characteristics as the incident illumination, since it may lead to the classification of shadow boundaries as material changes. In such cases, we show that it is necessary to take into account the spectral properties of the ambient illumination in order to develop a technique that is more robust and stable than previous techniques. This technique uses a biologically motivated model of color vision and, in particular, a set of chromatic-opponent and double-opponent center-surround operators. We apply this technique to simulated test patterns as well as to a chromatic image. It is shown that, given some knowledge about the strength of the ambient illumination, this method provides a better classification of shadow boundaries and material changes.

Color↗
PubMed1985

Computer assessment of left ventricular wall motion: the ALVEN expert system.

The current limited success of computer-assisted analysis of left ventricular (LV) dynamics is due to three main reasons: there is a strong tendency to remain within the realm of mathematical modeling for LV dynamics, and it is not at all clear that this is an adequate approach; in places where mathematical models alone may be insufficient, current computer science research into more sophisticated schemes is not yet complete, and thus, more basic research is required, particularly into artificial intelligence, representations of knowledge, and interpretation control structures, before applications such as LV performance can be solved, a view also stated in M. Boehm and K. Hoehne (in "Digital Image Processing in Medicine" (K. Hoehne, Ed.), Springer-Verlag, New York/Berlin, 1981); there is a distinct lack of knowledge about LV dynamics, in conjunction with disagreements about what is important to model and what terminology is to be used. Although each of these issues is addressed, the first two issues are concentrated on. Furthermore, a language for the expression of definitions for terminology has been designed, and a system for LV dynamics interpretation has been implemented.

Cineradiography↗