Extraction of Fusiform Gyrus Area of Brain to Analyze Autism

Extraction of Fusiform Gyrus Area of Brain to Analyze Autism

Mousumi Bala

Department of Computer Science and Engineering Jahangirnagar University

Savar, Bangladesh

Sujan Kumar Das

Department of Computer Science and Engineering Stamford University Bangladesh

Savar, Bangladesh

Abstract Autism is a complex disorders of brain development. Now it is a great threat to the present world. It is a serious neurodevelopmental disorder that impairs a child's ability to communicate and interact with others. It also includes restricted repetitive behaviors, interests and activities. These issues cause significant impairment in social, occupational and other areas of functioning. Recent studies show abnormalities in the various regions of brain in autistic persons. To investigate the causes of autism, fusiform gyrus of brain is responsible for face processing tasks has been studied for the control and autistic individuals. In carrying out the work, three categories of faces that is, familiar faces, stranger faces and combination of both familiar and stranger faces are considered of fusiform gyrus from functional magnetic resonance imaging (fMRI) images. Detect the activation area in fusiform gyrus of these images using edge detection methods as Robert, Prewitt and Sobel operators. Extract the activation area in fusiform gyrus using thresholding. Then these images are applied for segmentation. After segmentation the values of activation areas for both control and autistic individuals are calculated using the binarization method. All values of activation areas are compared for both control and autistic individuals. It has been observed that fusiform gyrus regions are hypoactive in patients with autism than in control.

Keywords Autism; Autistic Spectrum Disorders(ASD); Fusiform Gyrus; Functional Magnetic Resonance Imaging(fMRI).

1. INTRODUCTION

   Autism appears to have its roots in very early brain development. However, the most obvious signs of autism and symptoms of autism tend to emerge between 2 and 3 years of age. It is part of a spectrum disorders characterized by a triad of symptoms, including deficits in all aspects of social reciprocity; pragmatic communication deficits and language delays; and an assortment of behavioral problems, such as restricted interests, sensory sensitivities and repetitive behaviors [1]. It is a developmental neural disorder, which is known as autistic spectrum disorders (ASD) [2]. The worldwide prevalence of ASD is about 6 per 1,000, with about four times as many males as females [3]. It affects information processing in the brain by altering nerve cells and their synapses. However, the real cause of autism occurrences is not well understood yet [4]. Autism has a strong genetic basis, although the genetics of autism are complex and it is unclear whether ASD is explained more by rare mutations, or by rare combinations of common genetic variants [3], [5]. In rare cases, autism is strongly associated with agents that cause birth defects [6], [7]. Autism is

   characterized by ASD first appears during infancy or childhood, and generally follows a steady course without remission and establishes by age two or three years [8], [9]. Autistic people understand both surroundings and human behavior uniquely since they react in an abnormal way to input stimuli makes problematic human engagement, restricted interests and inability in the environmental generalization [10]. A key feature of normal social functioning in humans is the processing of faces, which allows people to identify individuals and enables them with the capacity to understand the mental state of others [3]. It is well recognized from functional magnetic resonance imaging (fMRI) studies that the fusiform gyrus is consistently active when normal humans view faces [4]. Patients with autism can perform face perception tasks but there is strong evidence that the fusiform gyrus, as well as other cortical regions supporting face processing in controls, is hypoactive in patients with autism [4,6-8]. It has been proposed that the failure to make direct eye contact may explain the observed hypo-activation of the fusiform gyrus in face perception tasks in autism [9]. This review will specifically focus on face perception deficits in autism, describing current literature on abnormalities in the fusiform face area and the amygdala. It will be argued that an abnormality early in development in the amygdala can give rise to later social perceptual deficits in face identity and facial expression perception. There is evidence that the fusiform gyrus receives input from the visual cortex and provides the major input into an extended system consisting of cortical regions and sub-cortical regions such as amygdala indicating that the altered function of the fusiform gyrus in patients with autism. Three categories of faces that is, familiar faces, stranger faces and combination of familiar and stranger faces are considered for examine of fusiform gyrus from fMRI images. The proposed approach has been implemented in MATLAB. The fMRI images are taken as the input image and detect the edge of these images using different edge detection operators and calculate the value of activation area in fusiform gyrus of brain for control and autistic individuals. The results obtained after segmentation is taken as input for binarization operation and compared to the control based on the calculated value of activation area of fusiform gyrus of brain.

2. FUSIFORM GYRUS

   The fusiform gyrus is a part of the human visual system that, it is speculated, is specialized for facial recognition. It is located in Brodmann Area 37. It is also known as the (discontinuous) occipitotemporal gyrus [11]. Other sources

   have the fusiform gyrus above the occipitotemporal gyrus and underneath the parahippocampal gyrus [12].

   There is still some dispute over the functionalities of this area, but there is relative consensus on the following:

   1. processing of color information

   2. face and body recognition

   3. word recognition

   4. number recognition

   5. within-category identification

   Some researchers think that the fusiform gyrus may be related to the disorder known as prosopagnosia, or face blindness. Research has also shown that the fusiform face area, the area within the fusiform gyrus, is heavily involved in face perception but only to any generic within-category identification which is shown to be one of the functions of the fusiform gyrus [13]. Fusiform gyrus has also been involved in the perception of emotions in facial stimuli [14]. Fig.1 shows the fMRI of fusiform gyrus of brain for the control and autistic individuals which clearly indicates the differences among them (Consider three types of faces- all faces, familiar faces and stranger faces)[15].

   Control Autistic

   1. All faces.

      Control Autistic

   2. Familiar faces.

      Control Autistic

   3. Stranger faces

      Fig.1 fMRI of fusiform gyrus both of control and autistic for

      1. all faces, b) familiar faces and c) stranger faces.

3. MATERIAL AND METHODS

   In carrying out the work fMRI scan images of the fusiform gyrus of brain for control and autistic individuals are taken as input. Three categories of faces that is, familiar faces, stranger faces and combination of both familiar and stranger faces are considered of fusiform gyrus from fMRI images as input. Detect the edge of activation area in fusiform gyrus of input images using Roberts, Prewitt and Sobel operators.

   Roberts Oerator

   The Roberts operator is given by the equations: Gx= W9-W5

   Gy= W8-W6

   -1	0
   0	1

   0	-1
   1	0

   -1	0
   0	1

   0	-1
   1	0

   (a) (b)

   Fig.2 (a) Roberts Mask for Horizontal Direction (b) Roberts Mask for Vertical Direction

   Sobel Operator

   Consider the arrangement of pixels about the pixel:

   W1	W2	W3
   W4	W5	W6
   W7	W8	W9

   The Sobel operator is given by the equations: Gx = (W7+2W8 +W9) (W1+2W2+W3)

   GY = (W3+2W6 +W9) (W1+2W4+W7)

   Where, W1 to W9 are pixels values in a sub image as shown in Fig.3

   -1	-2	-1
   0	0	0
   1	2	1

   -1	-2	-1
   0	0	0
   1	2	1

   -1	-2	-1
   0	0	0
   1	2	1

   -1	-2	-1
   0	0	0
   1	2	1

   1. (b)

   Fig.3. (a) Sobel Mask for Horizontal Direction (b) Sobel Mask for Vertical Direction

   Prewitt Operator

   The Prewitts operator is given by the equations: Gx = (W7+W8 +W9) (W1+W2+W3)

   GY = (W3+W6 +W9) (W1+W4+W7)

   -1	-1	-1
   0	0	0
   1	1	1

   -1	0	1
   -1	0	1
   -1	0	1

   -1	-1	-1
   0	0	0
   1	1	1

   -1	0	1
   -1	0	1
   -1	0	1

   (a) (b)

   Fig.4 (a) Prewitt Mask for Horizontal Direction (b) Prewitt Mask for Vertical Direction

   Extract the activation area in fusiform gyrus using thresholding. Then these images are applied for segmentation. After segmentation, the values of activation areas for both control and autistic individuals are calculated using the binarization method. That is the image having only two values either black or white (0 or 1). Here 256×256 jpeg image is a maximum image size. The binary image can be

   represented as a summation of total number of white and black pixels.[16] Area of an image is the total number of the pixels present in the area which can be calculated in the length units by multiplying the number of pixels with the dimension of one pixel:

   Image, I = 255=0 255=0[f(0) + f(1)]

   f (0) = white pixel (digit 0) f (l) = black pixel (digit 1)

   Pixels = Width (W) X Height (H)

   = 256 X 256

   No_ of_ white pixel P = 255=0 255 =0[f(0)] Where,

   P = number of white pixels (width*height)

   Finally, the calculated activation values are used for comparing the surroundings of fusiform gyrus to show the graphical representation for both control and autistic. The proposed work is implemented using Matlab.

4. RESULTS AND DISCUSSIONS

   The work is performed for the calculation of activation areas in fusiform gyrus of control and autism individuals. In carrying out the work fMRI scan images of the control and autism are taken as input and the corresponding images are produced by edge detection methods (Roberts, Prewitt and Sobel), thresholding, segmentation and binarization operation. The images of the input and produced output for the control and the autistic individuals are shown separately. Fig. 1 shows the input images of the fusiform gyrus of brain both of control and autistic for a) all faces, b) familiar faces and c) stranger faces.

   A. Simulated images for all, familiar and stranger faces

   Roberts Operator

   To detect the neural pathway surrounding the fusiform gyrus area of brain, Roberts operator is used for the control and the autistic in. Fig.5 shows the processed images of the control and autistic for all, familiar and stranger faces.

   Control Autistic

   All Faces

   Familiar Faces

   Stranger Faces

   Fig. 5. Processed images of the fusiform gyrus using Roberts operator of control and autistic for all faces, familiar faces and stranger faces.

   Prewitt Operator

   To detect surrounding edge of the fusiform gyrus area of brain, Prewitt operator is used for the control and the autistic in. Fig.6 shows the processed images of the control and autistic for all, familiar and stranger faces.

   Control Autistic

   All Faces

   Familiar Faces

   Stranger Faces

   Fig. 6. Processed images of the fusiform gyrus using Prewitt operator of control and autistic for all faces, familiar faces and stranger faces.

   Sobel Operator

   To detect the neural pathway surrounding the fusiform gyrus area, sobel edge detection operator is used for the control and the autistic. Fig.7 shows the processed images.

   Comparing the produced images, it is clearly visualized that the edges for the control are sharply observed which indicates that the neural pathways for the control are more effective surrounding the fusiform gyrus than autistic.

   Control Autistic

   All Faces

   Familiar Faces

   Stranger Faces

   autistic individuals. It also shows extracted activation area of fusiform gyrus of brain for both control and autistic individuals.

   Control Autistic

   All Faces

   Fig. 7. Processed images of the fusiform gyrus using Prewitt operator of control and autistic for all faces, familiar faces and stranger faces.

   Thresholding

   Fig.8 shows the result of thresholding which gives the accurate edge detected images using sobel operator of the fusiform gyrus of brain for control and autistic for all faces, familiar faces and stranger faces.

   Control Autistic

   All Faces

   Familiar Faces

   Stranger Faces

   Fig. 8. Edge detected images of the fusiform gyrus of brain for control and autistic using thresholding.

   Image Segmentation

   It has been observed that Sobel edge detection operator is computationally more expensive compared to Prewitt and Roberts operator. It means using Sobel operator the edge detection of fusiform gyrus of brain is very clear from Roberts and Prewitt operators. Consider processed images using Sobel operator for extraction fusiform gyrus area of brain both control and autistic using segmentation methods. Fig. 9 shows the segmented images for both control and

   Familiar Faces

   Stranger Faces

   Fig. 9. Segmented images of the fusiform gyrus of brain of control and autistic for all faces, familiar faces and stranger faces.

   Calculation activation area of fusiform gyrus

   Consider segmented processe images for calculating the values using binarization operation of activation area fusiform gyrus of brain of control and autistic for all faces, familiar faces and stranger faces.

   Table 1. Indicated activation area and calculated values of both control and autism for all faces.

   For All faces		Area in pixels
   Control	Left side	219
   	Right side	154
   Autistic	Left side	102
   	Right side	47

   Table 2. Indicated activation area and calculated values of both control and autism for familiar faces.

   For Familiar faces		Area in pixels
   Control	Left side	107
   	Right side	76
   Autistic	Left side	92
   	Right side	54

   Table 3. Indicated activation area and calculated values of both control and autism for stranger faces.

   For Stranger faces		Area in pixels
   Control	Left side	126
   	Right side	63
   Autistic	Left side	65
   	Right side	26

   Table 1. shows activation area is indicated as pink color for left and right sides of fusiform gyrus of brain both of Control and Autistic for all faces. The values are calculated for indicated area of fusiform gyrus. For left side, the values are 219 (in pixels) for Control and 102 (in pixels) for Autistic. For right side, the values are 154 (in pixels) for control and 47 (in pixels) for autistic. From these calculated values, it is observed that the values for left and right sides of Autistic are smaller than Control.

   Table 2. shows also activation area is indicated as pink color for left and right sides of fusiform gyrus of brain both of control and autistic for familiar faces. The values are calculated for indicated area of fusiform gyrus. For left side, the values are 107 (in pixels) for control and 92 (in pixels) for autistic. For right side, the values are 76 (in pixels) for control and 54 (in pixels) for autistic. From the result values, it is observed that the values for left and right sides of autistic are also smaller than control but the difference of values are very close.

   Table 3. shows also activation area is indicated as pink color for left and right sides of fusiform gyrus of brain both of control and autistic for all faces. The values are calculated for indicated area of fusiform gyrus. For left side, the values are 126 (in pixels) for control and 65 (in pixels) for autistic. For right side, the values are 63 (in pixels) for control and 26 (in pixels) for autistic. From these calculated values, it is observed that the values for left and right sides of autistic are smaller than control.

   From the calculating values, it has been observed that the values of activation area for autistic are smaller than control. The compared results are shown in Fig. 5, Fig. 6 and Fig. 7 with graphical representations.

   Fig. 5. Graphical representation of both Control and Autistic for all faces.

   Fig. 6. Graphical representation of both Control and Autistic for familiar faces.

   Fig. 7. Graphical representation of both Control and Autistic for stranger faces.

5. CONCLUSION

By using Sobel operator the edge detection of fusiform gyrus of brain is very clear from Roberts and Prewitt operators. Then applying threshoding, segmentation and binarization operation on fMRI scan images of control and autistic individuals, it is clearly visualized that there are differences in calculated values of activation area of the fusiform gyrus for three types faces (all faces, familiar faces and stranger faces). From these differences, it has been observed that the activation areas in fusiform gyrus are hypoactive in patient with autism than in control. It is a simulation model to understand the risk of autistic child. The causes of these differences needs to be investigated in details which demands further-study.

REFERENCES

1. American Psychiatric Association, 1994. Diagnostic and statistical manual of mental disorders, fourth ed. (DSM-IV). American Psychiatric Association, Washington, DC.

2. Volkmar, F., Lord, C., Bailey, A., Schultz, R.T., Klin, A., 2004. Autism and pervasive developmental disorders. J. Child Psychol. Psychiatry 45 (1), 135170.

3. Baron-Cohen S, Ring H, Moriarty J, Schmitz B, Costa D, Ell P.Recognition of mental state terms. Clinical findings in children with autism and a functional neuroimaging study of normal adults. Br J Psychiatry 1994; 165: 6409.

4. Kanwisher N, Stanley D, Harris A. The fusiform face area is selective for faces not animals. NeuroReport 1999; 10: 1837.

5. Schultz RT. Developmental defecits in social perception in autism: the role of the amygdala and the fusiform face area. Int J Dev Neurosci 2005; 23:12541.

6. Pierce K, Mu¨ller RA, Ambrose J, Allen G, Courchesne E. Face processing occurs outside the fusiform face area in autism: evidence from functional MRI. Brain 2001; 124: 2059 73.

7. Pierce K, Haist F, Sedagat F, Courchesne E. The brain response to personally familiar faces in autism: findings of fusiform activity and beyond. Brain 2004; 127: 270316.

8. Bolte S, Hubl D, Feineis-Matthews S, Prvulovic D, Dierks T, Poustka F. Facial affect recognition training in autism: can we animate the fusiform gyrus? Behav Neurosci 2006; 120: 2116.

9. Dalton KM, Nacewics BM, Johnstone T, Scheafer HS, Gernsbacher MA, Goldsmith HH, et al. Gaze fixation and the neural circuitry of face processing in autism. Nat Neurosci 2005; 8: 51926.

10. Johanson, D. C. (1996). From Lucy to language. New York: Simon and Schuster, p. 80.

11. Nature Neuroscience, vol7, 2004.

12. nervsystemet.se – Hjärnatlas.

13. McCarthy, G et al. Face-specific processing in the fuman fusform gyrus.J. Cognitive Neuroscicence. 9, 605-610(1997).

14. Radua, Joaquim; Phillips, Mary L.; Russell, Tamara; Lawrence, Natalia; Marshall, Nicolette; Kalidindi, Sridevi; El-Hage, Wissam; McDonald, Colm et al. (2010). "Neural response to specific components of fearful faces in healthy and schizophrenic adults". NeuroImage 49 (1): 939946. doi

    :10.1016 / j . neuroimage .2009 . 08 .030.

15. Karen Pierce and Elizabeth Redcay, Fusiform Function in Children with an Autism Spectrum Disorder Is a Matter of Who, BIOL PSYCHIATRY 2008; 64:552560.

16. Mr.Rohit S. Kabade et al." Segmentation of Brain Tumou and Its Area Calculation in Brain MR Images using K- Mean Clustering and Fuzzy C-Mean Algorithm" International Journal of Computer Science & Engineering Technology (IJCSET), Vol. 4 No. 05 May