Functional Integrity of Dorsal and Ventral Visual Streams in Children with Autism

Background: Visual information in the brain is processed by two distinct pathways (the dorsal and the ventral visual streams) which operate in a parallel and distributed manner. The dorsal stream extends from the occipital towards the parietal lobe and is associated with locating objects in space, and the ventral stream is commonly activated during tasks that involve recognizing objects, and information in this system travels in the direction of the temporal lobes (Ungerleider and Mishkin, 1982). Since people with autism may identify objects differently (enhanced ability to locate objects in space) from typically developing persons, it will be particularly relevant to examine the functional integrity of the brain areas mediating these functions. It is also equally relevant considering the recent findings of widespread functional underconnectivity among several brain areas in autism (Just et al., 2004; Kana et al., 2006).

Objectives: The primary objective of this study is to examine the functional integrity of cortical networks, especially the dorsal and ventral visual streams, involved in visual processing in children with autism.   

Methods: Four high-functioning children (age range: 10-15 years) with autism, and two typical control participants took part in this fMRI study (data collection is in progress). The fMRI data collected from the Siemens 3.0T Allegra scanner at the UAB Civitan International Research Center is analyzed using SPM2 (Wellcome Department of Cognitive Neurology, London, UK). The stimuli consist of a series of common household objects presented in blocked design format in two experimental conditions. In the object recognition condition, the participants named an object from four choices, and in the location identification condition, they identified the location of a given object with respect to a cross on the screen. 

Results: The fMRI data from this study showed that the dorsal and ventral visual streams were differentially recruited in the control participants for location and object conditions respectively. While the control participants activated inferior parietal area in the location identification condition and inferior temporal gyrus in the object recognition condition, participants with autism showed activation in parietal areas in both experimental conditions. In addition, unlike control participants, the participants with autism showed no activation in left inferior frontal gyrus in the object recognition condition.  Behavioral results showed that the autism group made more errors in their responses; however both groups showed no differences in latency for either condition. The results are preliminary at this point and reveal only a trend.

Conclusions: The participants with autism seem to recruit more posterior brain areas, such as intraparietal sulcus in accomplishing both tasks suggesting their increased reliance on visuospatial processing. This type of processing is also reflected in the lack of reliance on language areas (reduced activation in left inferior frontal and left posterior superior temporal areas) when recognizing and naming an object.  As the data collection is progressing, functional connectivity analysis with adequate sample size will be important in determining the integrity of the dorsal and ventral visual streams in autism.