Case-Based Reasoning with Biofeedback to Help the Learning Process of Autism Children

Home/Articles/Case-Based Reasoning with Biofeedback to Help the Learning Process of Autism Children

Autism spectrum disorders (ASD) are characterized by qualitative abnormalities in social behavior and communication skills and restricted, repetitive, and stereotyped patterns of behavior, interests, and activities. Behavioral training programs based on Applied Behavior Analysis (ABA) are evidence-based treatments that are effective in the reduction of the core features of ASD. According to World Health Organization (WHO), it is estimated that worldwide about one in 160 children has ASD. This estimate represents an average figure, and reported prevalence varies substantially across studies. Some well-controlled studies have, however, reported substantially higher figures. The prevalence of ASD in many low- and middle-income countries is unknown. Case-based reasoning (CBR) is a methodology for developing knowledge-based systems where the central issue is past cases or experiences. Case-based reasoning means solving problems based on past experiences, remembering previous cases to guide the solution to current problems, and adapting past solutions to new problems. A previous experience can play several roles: suggest a solution to a new problem or a way of interpreting a situation, warn of a problem that will arise, or allow the potential effects of a proposed solution to be predicted. CBR views analogical reasoning as the centerpiece of our ability to function as human beings. It posits that our most natural and powerful learning strategies are the automatic ones that situate learning in real-world experience. Biofeedback is a technique you can use to learn to control some of your body’s functions, such as your heart rate. During biofeedback, you’re connected to electrical sensors that help you receive information about your body. This feedback helps you make subtle changes in your body, such as relaxing certain muscles, to achieve the results you want, such as reducing pain. In essence, biofeedback gives you the ability to practice new ways to control your body, often to improve a health condition or physical performance. Combining the CBR and biofeedback it might be possible to help ASD children with a new system called Adaptive Choice Case-Based Reasoning (ACCBR) system, which learns from and interacts with a user to assist them in achieving greater concentration and productivity based on the biofeedback.

Biofeedback helps you make subtle changes in your body, such as relaxing certain muscles, to achieve the results you want, such as reducing pain. In essence, biofeedback gives you the ability to practice new ways to control your body, often to improve a health condition or physical performance. There are a variety of biofeedback methods depending on the health problems and goals. Types of biofeedback include Brainwaves, breathing, heart rate, muscle contraction, sweat gland activity, and temperature. In this article the proper biofeedback is the brainwaves, this type uses scalp sensors to monitor your brain waves using an electroencephalograph (EEG). Previous research has confirmed that neurofeedback can help children with autism increase their functioning by improving attention, behavior, and sensory-motor skills.

The Flow Choice Architecture or FCA can be developed to integrate the mental state and learning performance of an ASD child to suggest “nudges” that can help to stabilize the mental state to be ready to learn and accept any lessons. Nudges can be rituals, sounds, speech, or other “mental hacks” that subtly encourage behavior change. If FCA classifies children’s biosignals as a distraction, it may “nudge” the children by showing task completion lists, highlighting the days, or verbalizing encouragement.

The figure shows the interaction of the FCA with ASD children as they learn. The FCA uses mental and task states, as well as rewards, to assess whether a nudge is necessary. This information, including whether there have been earlier nudges, is used to calculate the step that is most likely to return ASD children to a normal mental state. Nudges with a minor impact on the operator’s mental state will be chosen less frequently than those with a rapid and positive impact.

Case-based reasoning for EEG has been successfully implemented to recognize depression with a very high accuracy result. EEG Biofeedback includes converting data on brainwave activity to a computer, which further translates it into game-like representations that can be aural, visual, or both. During a typical session, EEG electrodes are placed on the scalp and/or ear lobe(s). These sensors only measure a person’s brainwaves; no electrical current enters the brain. Individuals utilize their brainwaves to learn to control the feedback they instantly receive about the amplitude and synchronization of their brain activity. As a child learns to control and improve brainwave patterns, the game scores increase, and progress occurs. The only way to succeed at the games is for children to improve their brainwave function. Integration of decision trees and CBR was used to classify patient EEGs into one of seven different psychological or physical disorders, including migraines and ADHD. Their work showed that even a very coarse discretization of EEG signals in conjunction with psychological and behavioral features can produce an accurate classification of different disorders. ACCBR computes features from the raw EEG signals and combines them with other information to construct a case to be classified. However, ACCBR differs from this system in that the determination to make nudges is a continuous process, happening throughout the working day.

The“Nudge Controller” is implemented in ACCBR. The goal of ACCBR is to see if the operator’s current mental and task state signals that he or she could need a little help getting into or staying in deep, focused work. To accomplish this, it continuously balances the user’s stream of EEG signal classifications with task and historical data to conclude whether to produce a nudge or remain silent.

The FCA starts with user demographic and trait information, task knowledge, and task rewards, and then moves on to the work at hand. The operator and task state are observed, and the EEG signals are classified into one of the four conventional quadrants of the valence arousal scale: neutral, sad, frightened, and happy. When the classification has high confidence (i.e., when the same classification has been made several times in a row), it is combined with the task state and other contextual information to create a probe into case memory. The probe will retrieve one to five cases from case memory, and a response will be generated depending on the returned cases. The response is determined by the response distribution and the history of past nudges. The nudge is activated, and FCA tracks the operator’s status to see how effective it is. The change of mental state to a more positive mental state, as well as the task completion status, are used to gauge efficacy. This process continues until the operator is mentally exhausted or the work tasks are completed.

(ASD) isn’t a learning disability. But it does affect learning and sometimes in ways like learning disabilities. And kids who have autism are often eligible for special education services. By using case-based reasoning, neurofeedback from the ASD children’s biofeedback, the ACCBR can be developed to help all ASD children to learn better and easier. the ACCBR will read the EEG result from the ASD children and use the nudges to help the ASD children at a normal mental state and ready to accept lessons. The flow control architecture will also help them by giving them tasks or lessons based on their current mental state.