my story

FCMs

Mon, 3 Oct 2011 06:47:19 -0500

Patterns and pattern making are an essential part of the way the brain works. Migratory birds are born with a pattern, that is, a twice annual alarm telling them when and where to flock. Grid cells fire whenever a rat is at the vertex of an equilateral triangle, no matter where the rat is initially placed.

So too humans are able to form patterns. These patterns link sets of information in at least an internally consistent way (think theoretical systems). More importantly, they link information is a causally meaningful way. The “math” behind these linkages are called fuzzy cognitive maps.

Segmentation theory, film and psychopathology -2

Sat, 23 Jan 2010 15:00:55 -0600

From this perspective, segmentation occurs when the comprehensibility (i.e., an unpredicted change) of an event breaks down. However, the order of events allows for a higher level pattern recognition. This both “fine grained” and “coarse grained” hierarchical order allows for prediction.

There is evidence to suggest that the organization of events into fine-grained and coarse-grained sequences depends on different processes and presumably different brain regions. Subjects with frontal lobe lesions or schizophrenia show selective impairments in coarse-grained segmentation.

Of course, whether or not such segmentation deficits result in, or from the underlying pathology of schizophrenia remains to be seen. But here’s where I get to speculate...because it is my blog, after all.

When we speak about an increase in processing load, or faltering comprehension I believe that we are referring to a probabilistic shift. That something that has now entered our perceptual awareness is outside the normal distribution of objects, or characteristics of that event. If, however, the salient elements that we have used to define the set (or event) are fairly irrelevant, we will be unable to meaningfully detect event boundaries and the whole hierarchy breaks down.

Is this what happens in psychosis? Would fMRI or some other functional imaging technique applied while watching film help us to determine brain regions associated with segmentation “errors” among people with schizophrenia? Hmmm.

Segmentation theory, film and psychopathology -1

Sat, 23 Jan 2010 14:17:53 -0600

This is the first entry on the topic of how studying film can help us to better understand the cognitive neuroscience of psychopathology. Kurby and Zacks (Trends in Cognitive Sciences, 12 (2); 2007, 72-79) offer an excellent review of Event Segmentation Theory. EST posits that we spontaneously parse a seemingly continuous stream of experience into events. Studies indicate that we form such segments when there is an event boundary, the latter consisting of a significant change in characters, places, things, or causal contingencies. Essentially, any change that results in an increase in “processing load”, that is, a demand in increasing newly encountered information, may demarcate a segment. Various tasks have been used in studies investigating the biological correlates of segmentation, including readings, music, visual cues and, best of all, movies. Activity in a number of different brain regions appears to be associated with subjects’ perceptions of event boundaries. For example, subjects viewing a simple film were then asked to segment the film into events. Using fMRI, those event changes were associated with changes in bilateral posterior occipital, temporal, and right lateral frontal cortex (Zacks, JM et al, (2001) Human brain activity time-locked to perceptual event boundaries. Nat. Neurosci. 4, 651-655). In another study involving classical music, segment boundaries were associated with fMRI early changes in a ventral network (right sided temporal and ventrolateral prefrontal cortex) and later changes in a dorsal network (dorsolateral prefrontal cortex and posterior parietal, all on the right side) (Sridharan, D. et al. (2007) Neural dynamics of event segmentation in music: Converging evidence for dissociable ventral and dorsal networks. Neuron 55, 521-532)
EST suggests two purposes for segmentation. The first relates to memory consolidation. Working memory appears to be updated at event boundaries. So, segmentation “chunks” information into more manageable packets for neat storage. A second function for segmentation is as an aide to the cognitive task of prediction.