Visual parameter optimisation for biomedical image processing.

BACKGROUND: Biomedical image processing methods require users to optimise input parameters to ensure high-quality output. This presents two challenges. First, it is difficult to optimise multiple input parameters for multiple input images. Second, it is difficult to achieve an understanding of underlying algorithms, in particular, relationships between input and output. RESULTS: We present a visualisation method that transforms users' ability to understand algorithm behaviour by integrating input and output, and by supporting exploration of their relationships. We discuss its application to a colour deconvolution technique for stained histology images and show how it enabled a domain expert to identify suitable parameter values for the deconvolution of two types of images, and metrics to quantify deconvolution performance. It also enabled a breakthrough in understanding by invalidating an underlying assumption about the algorithm. CONCLUSIONS: The visualisation method presented here provides analysis capability for multiple inputs and outputs in biomedical image processing that is not supported by previous analysis software. The analysis supported by our method is not feasible with conventional trial-and-error approaches.

Context dependent latent inhibition in adult humans.

Learning the association between one stimulus (a condition stimulus, CS) and another (unconditioned stimulus, US) can be impaired by prior exposure to the CS alone--latent inhibition (LI). Current theories attempting to elucidate the cognitive deficit in schizophrenia have used the abolition of LI in schizophrenia as an indicator of attentional dysfunction. However, it has always been unclear if human and animal LI are measuring the same psychological processes. It is obviously important to clarify this relationship so that theoretical and experimental developments in the rat do not mislead the investigation of brain-behaviour relationships in schizophrenia. LI in the rat is strongly dependent upon context. Our aim was to examine the context specificity of LI in humans and specifically to: (1) investigate whether participants' belief that they are in a different context is sufficient to abolish LI, even though there is no physical change in the environment; (2) produce a context manipulation that is immune to alternative interpretation in terms of stimulus generalization decrement; and (3) investigate whether a "tonic" change of context reduces or abolishes human LI, thus complementing previous reports using a "phasic" change of context. In two experiments we manipulated context in either the real world or a virtual world, and showed that LI is abolished by a change of context in adult humans.

Manual and virtual rotation of a three-dimensional object.

An orientation-matching task, based on a mental rotation paradigm, was used to investigate how participants manually rotated a Shepard-Metzler object in the real world and in an immersive virtual environment (VE). Participants performed manual rotation more quickly and efficiently than virtual rotation, but the general pattern of results was similar for both. The rate of rotation increased with the starting angle between the stimuli meaning that, in common with many motor tasks, an amplitude-based relationship such as P. M. Fitts' (1954) law is applicable. When rotation was inefficient (i.e., not by the shortest path), it was often because participants incorrectly perceived the orientation of one of the objects, and this happened more in the VE than in the real world. Thus, VEs allow objects to be manipulated naturally to a limited extent, indicating the need for timing-scale factors to be used for applications such as method-time-motion studies of manufacturing operations.