Sensory abnormalities are reported to be central to autistic experience. The symptoms include hyposensitivity, hypersensitivity, multichannel receptivity, processing difficulties and sensory overload. Given the intensive and cumbersome nature of current methods of diagnosis, there is a need of behavior-based approach to identify the children at risk for autism spectrum disorder (ASD). Children with autism often like technology and robots. So, using robots can help them connect better with others and improve their social and emotional skills.
GW researchers have developed a novel system that uses facial expression and upper body movement patterns to detect ASD. A robot resembles typical everyday experiences like uncontrolled sounds and lights or tactile contact with different textures to watch how children react. A camera records the robot-child interactions, and a convolutional neural network (Artificial Intelligence Algorithm) evaluates multimodal data collected from those recordings, and identifies the signs of autism. The robot also helps children learn to express their feelings about things that bother them, instead of reacting strongly.
Figure: Diagram of robot aided platform
Advantages:
Applications: