© 2022 IEEE.Childhood of Apraxia of Speech (CAS) is a type of disorder that affects the speech of the children. Children who have a CAS found it difficult to control and coordinate the brain's neutrons in order to create sounds for the phrases. Therefore, while talking they behave themselves in abnormal way. This research work proposes the Computer Vision based solution in order to diagnose the patients agains CAS based on video content. We will be using the PoseNet Convolutional Neural Network (CNN) in order to track the pose of the child and fix abnormal behaviour while giving the interview. As a result we will be able to automatically identify whether the child has CAS or not.