The study investigates the feasibility and the effects of training at home with personalised neurofeedback in children with an autism spectrum diagnosis and comorbidities. The personalisation relies on quantitative electroencephalography recordings performed in a clinic, which are also used to evaluate the efficiency of the home-based training, along with standardized questionnaires. The active phase is expected to last roughly 6 months, during which children are expected to complete multiple hours of training and complete an interim evaluation. Whether the training's effects are long lasting is evaluated in a follow-up phase, 3 months after the active phase.

Autism spectrum disorders (ASD) affect one in 59 children in the developed world and are characterised by significant impairments in social abilities and repetitive/restrictive behaviours. ASD are heterogeneous in nature and manifestations, and usually come in conjunction with other neurodevelopmental disorders, of which attention deficit hyperactivity disorder (ADHD), epilepsy and intellectual disability (low Intelligence Quotient; IQ) are the most common. Interventions to alleviate ASD are in many instances hard to quantify in terms of effectiveness, leading to an inability of caregivers and practitioners to assign patients to an appropriate treatment approach.

Accumulating evidence highlights the utility of electroencephalography and its mathematical analysis, quantitative electroencephalography (qEEG), in discriminating between ASD and typically developing children in early infancy and throughout development. Additionally, previous studies show that EEG-based neurofeedback can be used to correct brain activity abnormalities. This approach builds on models of brain plasticity and uses operant conditioning of the patient's electroencephalogram (EEG) to induce brain adaptation and self-regulation, which further tunes cognition, emotions and behaviour. Prior scientific evidence suggests impaired mirror neuron activity in ASD, which connects to mu-rhythm abnormalities, as recorded on an EEG. Neurofeedback-mediated training of mu-rhythm power at C4, a EEG electrode position, in children with ASD resulted in improved social behaviour in previously published research. A similar approach using the company's Brain Hero® neurofeedback game in ASD children with or without ADHD and/or epilepsy and irrespective of IQ, which participated as beta testers in a previous usability study, resulted in improved concentration and relaxation abilities, and improved mu-rhythm at C4.

The objective of this feasibility study is therefore to track the efficiency of home-based personalised Brain Hero® neurofeedback in children with ASD and comorbidities. To identify whether the brain activity at other electrode positions and frequencies has been influenced by this neurofeedback approach and to track overall normalization of qEEG parameters in comparison with typically developing children, the investigators propose to engage the patients in a longitudinal clinical study evaluating changes in whole brain activity. To this end, they will employ a commercially available qEEG device to record whole scalp EEGs from Brain Hero® users, once before beginning neurofeedback and then again once every month over the course of neurofeedback. These EEGs will be analyzed using Neuroguide® (Applied Neurosciences) and power and coherence compared with their database of typically developing individuals. The points and frequencies of the neurofeedback training will be individually adjusted to specifically correct the EEG abnormalities detected in the patient. In comparison with the current C4 alpha-mu training and C4 alpha-, beta- and theta-tracking using spectral power measurements (available now for all Brain Hero® beta users), this approach has the advantage that it enables treatment personalization (e.g. training other electrode points and power frequencies in addition to C4 alpha-mu, as well as coherence training) and a more thorough tracking of treatment efficiency (by evaluating changes in spectral power and coherences over the entire scalp). In addition, the investigators will track changes in behavior using the Social Responsiveness Scale (SRS) 2 and the Behaviour Rating Inventory of Executive Function (BRIEF), as well as with personalized interviews, which will be completed by primary caregivers every three months.

Children or adolescents, male or female, aged 6-18 years, with an ASD diagnosis according to the DSM-5 or ICD-10, irrespective of IQ, with or without ADHD or epilepsy will be included in the study. Children with a diagnosis of schizophrenia, severe generalised anxiety disorder, major depression, or other neurological disorders, or with a medical disorder requiring systemic chronic medication with confounding psychoactive effects, as well as children with muscle spasms affecting the facial muscles and interfering with electrode measurements will be excluded from the study. Because of the high heterogeneity of ASD and its extremely frequent association with other neurodevelopmental disorders (ADHD, epilepsy, low IQ), the investigators expect that data from a minimum of 30 patients will be necessary. This is also the sample size recommended for the case in which the differences in pre- and post- values are not normally distributed. Because the aim of the study is beyond statistical significance towards biological relevance, the investigators are aiming for an average effect size similar to the ones reported in previous studies. Considering that this is a longitudinal study and a significant number of patients might drop out (dropout rate estimated at 70% in studies employing two groups and involving office/clinic visits for qEEGs), the investigators aim at recruiting a total of 100 patients.

The study will be conducted over 9 months, of which 6 months are active training and 3 months are follow-up phase, since the long-term effects of the neurofeedback training are also of interest. A patient undergoing all training sessions and evaluations is considered to have successfully completed the study.

The study will have reached its endpoint once complete data (all four time points, all qEEGs and all questionnaires) will be gathered from 30 patients diagnosed with an ASD, whether or not associated with comorbidities.

Inclusion Criteria:

• Children or adolescents, male or female, aged 6-18 years, with an ASD diagnosis according to the DSM-5 or ICD-10, irrespective of IQ (IQ will be determined for each patient), with or without ADHD or epilepsy
• Signature on informed consent form
• Wireless internet connection for using the game and training
• Tablet or Smartphone for Brain Hero® training
• Confirmation that the child will be provided a quiet and undisturbed environment at home for playing Brain Hero® and that he/she is assisted by the caregiver with questions throughout play sessions to facilitate the game hero's flight in the direction of interest

Exclusion Criteria:

• Established diagnosis of schizophrenia, severe generalised anxiety disorder, major depression, or other neurological disorders
• Medical disorder requiring systemic chronic medication with confounding psychoactive effects
• Muscle spasms affecting the facial muscles and interfering with electrode measurements
• Inability to cope with qEEGs (sit still for a duration of 20 minutes) or inability to use the neurofeedback device (tablet use, headset set-up, understanding instructions)
• Plan to start other ASD treatments, including applied behaviour analysis (ABA), music therapy, dietary interventions etc. within the next 6 months
• Plan to start pharmacological treatment (medication) within the next 6 months
• Any estimated major events leading to changes in child routine planned in the next 6 months (e.g. school/kindergarten change, caregiver change, home move etc.)
• Pregnant and nursing mothers