StudyMoose App
24/7 writing help on your phone
Save to my list
Remove from my list
Artificial Neural networks help us to diagnose ASD and AI-assisted intervention technologies such as contemporary computer-aided systems (CAS), computer vision assisted technologies (CVAT), and virtual reality (VR) helps autistic people to communicate and learn as it is difficult for them to make communication with other human beings. To diagnose ASD in individuals, ASD tests are developed under the study of ASD Diagnosis with Artificial Neural Network. The ANN model shows an accurate diagnosis of ASD based on the evaluation of test results.
The survey of AI-assisted intervention technologies reveals that these technologies are beneficial as autistic people prefer machine communication over humans.
I found this topic interesting because movies like Extinction and I robot create the impression that AI is going to change the world in a negative way which I wanted to disagree with and hence proved most positive aspect of AI through this paper. As well famous personalities like Einstein, Newton have autism, which again triggers my interest in this topic.
4.7 (348)
“ Amazing as always, gave her a week to finish a big assignment and came through way ahead of time. ”
Milo is a humanoid robot, which is developed to treat autistic students. This robot can help students to learn the necessary communication skills. The study for diagnosis for Autism through ANN is based on the test results data fed as input to the network and its evaluation through the network. What if the data provided is misleading? Then, in that case, also will ANN diagnose autism correctly? Also, AI-assisted intervention technologies help children learn. They make them comfortable enough to communicate. So , during this procedure, aren't we pushing children in an artificial world? Moreover, to what extent we can provide children in this world and how right it is?
Centers for Disease Control provide data, which states that 1 out of 59 children, has autism.
This suggests that autism is no longer a rare disease. It can affect any person from all levels of society. Autism is a developmental disorder. This affects social skills, speech, and nonverbal communication and causes repetitive behavior (WWW1). These symptoms of autism can appear at the age of 2 to 3. (Naseer, 2019 ).
Early diagnosis of autism in an individual helps to deal with it in the later part of his or her life. A person diagnosed with autism earlier shows excellent improvement over acquiring skills than a person diagnosed later. Artificial Neural Network is a mathematical model based on artificial intelligence that can diagnose autism in an individual. This technology proves an excellent help for autistic people and their caregivers as well as family members. After a diagnosis of autism, autistic people can be treated in various ways. Traditional clinical settings were favorite in the past. Along with traditional clinical settings, now a day's technology-based interventions are also proving useful for autistic people to learn various skills. Researches carried out during the last decade states that people suffering from autism shows interest in technology-based interventions for learning and communicating (Jaliaawala, 2019 ). These interventions can be used in clinical settings also in homes and classrooms. The study is executed by Jaliaawala (2019) about researches in the past decade, in his article, he states that all researches indicate the positive effect of technology interventions on, verbal and nonverbal skills of an autistic person. Robots like Milo helps autistic children to learn and understand, communication, and social skills (WWW2).
Recently I came across an article which states that Albert Einstein, Amadeus Mozart, Sir Isaac Newton, Charles Darwin, these famous personalities had autism. This triggered my interest in autism. While searching further, I came to know that Autistic people have difficulties in social communication. The use difficulties can be overcome with technologies like Milo and QTrobot by implementing artificial intelligence. This makes me more fascinated by this field.
Autism Spectrum Disorder comes under Pervasive Developmental Disorder, which again subdivided into High Functioning Autism, Low Functioning Autism, and Social interaction impairment. High functioning autistic people are those with IQ level 70 or above but face difficulties in emotion regulation and expression, communication, and social interaction. Low functioning autistic people are those having cognitive impairments. The symptoms of LFA include impaired social communication or interactions, problems in sleep, strange behavior, self -injurious tendencies (Jaliaawala, 2019 ). Autism affects verbal as well as nonverbal skills of an individual. According to data provided by the Centers for Disease Control in 2018, 1 out of 59 children is autistic. 1 in 37 boys and 1 in 151 girls that mean boys are four times more prone to this disorder. 31% of children have intelligence quotient less than 70, 25% of children are in the range of 71 to 85, and 44% of children possess above 85 intelligence quotient level(WWW1).
Diagnosis of Autism using Artificial Neural Network Model Ibrahim M. Nasser conducted an experiment, in which autism diagnosis is carried out with the help of ANN model. This model used sample data out of which 80 % data used to train the model and 20% of data used to validate the model. This data is collected from the ASD test app. This is a mobile screening application to detect the presence of autism.
Test: -This application is designed such that all persons of all ages can use it. ASD behaviors are measured based on four modules of this app. Autism Spectrum Quotient is the screening tool which consists of self -observed questions. The answers to these question s can detect autistic traits in an individual. This 50 -question set is reduced to 10, to shorten the version. These questions examine autism's cognitive strength. The answer to the question ranges from definitely agree to disagree. On the 10 -question scale, a person who scores more than six is detected with autism.
Methodology: - The data from the autistic screening test is collected. This dataset is used for this experiment. ANN model used in this experiment is trained and validated against this data and conclusions drawn from it. This data consists of user-specific information like name, gender, age, ethnicity, history about jaundice, autism detected in a family member, relation with a participant of the experiment, country of residence, information about the usage of the application in the past, along with ten revised questions from autism spectrum quotient. Data collected from the screening app is preprocessed before forwarding to the ANN model. This is done for predictive analysis. The actual data fed to model consists of answers from 10 questions, age, gender, jaundice at birth or not, class/ASD attribute. Apart from the age attribute, all other data is in the form of Boolean numbers, which means only 0 and one values. Out of these data, all attributes from age to jaundice are considered as input attributes, and the class attribute is considered as an output attribute, which shows ASD presence or absence.
Results: - ANN model developed by ( Naseer, 2019 ) shows an accuracy of 100%. The model was correct in the detection of autism.
Importance of Technology-Assisted Intervention in the treatment of autism autistic people lacks social and communication skills. This makes them difficult to reach for teaching and letting them understand society. Many of them avoid social contact, so standard school settings and clinical settings make it even more challenging to teach them necessary skills like, understand and respond to facial expressions, greet peers or do any communication with the elderly or with colleagues.
A Comprehensive Survey of AI-Assisted Intervention for Autistic People
The methodologies which are reviewed in this survey are contemporary computer-aided systems, computer vision assisted technologies, and virtual reality or AI-assisted intervention. In this study, literature is presented related to interventions for facial expression impairments in autistic people. Interventions for facial expression impairments are again subdivided into three categories.
They are non -verbal communication skills, verbal social skills, and virtual reality. Non -verbal communication skills can be summarized as emotion recognition, affect recognition, maintaining eye contact, decision making, etc. Similarly, verbal communication skills can be summarized as social interactions, greeting to peers, etc. Virtual reality includes both non - verbal and verbal skills. All the reviewed researches categorized in these three sections and conclusion is drawn. One of the main questions, among researches, was, whether Computer-Aided Systems are more effective than traditional clinical settings for autistic people? Some ofthe researches provide positive outcomes about statistical inference. Though the efficiency of CAS over traditional clinical settings is not proven, it has advantages over traditional clinical settings for individuals who have ASD. Advantages of CAS over Traditional Clinical Settings for Autism Unlimited repeated treatments can be formulated with correctness. Hence improving constancy. Because of automation, access to more extensive and remote areas which in turn saves cost. As human intervention is reduced, chances of unintentional induction of maladaptive behaviors are also reduced.
Methodology and Result: -Result of the survey are tabulated. This table contains column heads with Author, Participant, Mean Age, Average IQ, Participants characteristics, Skill intervened, Technology used for intervention. Data from various researches are studied, sorted, and entered against respective columns of tables. For example, Bernard Opitz et al. (2001) performed research in which 16 persons participated, eight persons, having ASD, and eight persons with TYP syndrome with a mean age of 4 years for TYP syndrome and seven years for ASD. The average IQ of these people was 90. The characteristics of these individuals noted were, eight persons have ASD out of which 2 were females, and 6 were males. Eight persons have typical control out of which 3 were females, and 5 were males. Social skills and solution generation ability in conflicting situations were intervened through this research. The technological background consists of software - developed by the researcher, hardware - a personal computer with windows 95 for ten sessions. Results obtained from this research were also entered against the result column. For this research, they compared the enthusiasm scale for computer and instructor. After comparison, they found that autistic people have fewer issues with behavior and increased motivation towards computer-aided sessions.
Robokind, a robotics company, created a robot called Milo. This is two feet tall social robot developed to support children with autism. The mechanical inclinations of the robot make children relate quickly with a humanoid robot. Milo can engage children 87 % of their time.
Milo, along with instructor or therapist, helps children learn without getting frustrated and teaches through the number of repetition s. Milo can be used as a group setting or an individual setting. Milo needs an instructor to facilitate between MMilo and the child. Milo in the present moment only accessible for schools and therapist's use. Unfortunately, it is not available for home or personal use (WWW2). The only disadvantage of MMilo is, it needs some prerequisite skills possessed by children to continue learning with Milo. Caregivers or therapists have to confirm these skills in an autistic individual to work with MMilo (WWW2).
As autism is a developmental disorder, there is no specific treatment or cure to autism. Early diagnosis of autism is beneficial to deal with autism. ANN can detect autism 100 % correctly based on data provided. A comprehensive survey of all researches regarding AI intervention with autistic people shows that the autistic person shows a keen interest in technology-based intervention and these interventions can be used to teach autistic population necessary skills regarding verbal and nonverbal communication types. However, though proving effective, there is no uniformity amongst computer scientists, therapists, and AI developing teams. There is no standard protocol which is followed during researches concerning AI intervention technologies with individuals who have autism. Hence data collected so far cannot be effectively used (Jaliaawala, 2019 ). Robots like Milo can prove a great help in the learning processes of the autistic population. However, there must be control over every technology used with individuals who have autism. Because of this disorder, these people are already in disconnection with society. If not use in control manner, children will get obsessed with robots and can forget the human touch. If the situation gets uncontrolled, then it will become complicated for traditional settings also to deal with children. Hence control use of any technology could be the key to establish better results to treat autism with AI-assisted technology intervention.
Autism as a developmental disorder. (2019, Dec 17). Retrieved from https://studymoose.com/autism-as-a-developmental-disorder-essay
👋 Hi! I’m your smart assistant Amy!
Don’t know where to start? Type your requirements and I’ll connect you to an academic expert within 3 minutes.
get help with your assignment
