Efficacy of rTMS as an Adjunct to AI Enabled Remedial Intervention in Children With Dyslexia

• Consecutive School going children fulfilling DSM V criteria for SLD with dyslexia will be screened in the study centre for eligibility.
• After applying inclusion and exclusion criteria, legal guardians of the participants will be counselled and written informed consent will be taken.
• Baseline data of the participants will be collected.
• GLAD (Grade level Assessment Device Tool ), CTOPP-2 (Comprehensive Test Of Phonological Processing Tool, 2nd edition), WRAT-5 (Wide Range achievement test, 5th edition ) will be administered on the participants at baseline.
• fMRI at baseline will be done for 10 sequential participants whose legal guardians are willing to give consent for the same.
• Intervention Study group will be subjected to 6 weeks of AI enabled remedial education with Hf-rTMS (high frequency - repetitive Transcranial Magnetic stimulation) as per protocol developed based on previous studies of rTMS in SLD and from experience from other studies using rTMS for other indications.
• Participant will be started on AI enabled remediation therapy as described below and after ensuring 1st session of remediation the participant will be called for the rTMS session as detailed below.

AI enabled remedial intervention -

• It will be administered via a Web application based phonological training program (web application address : http://readable.com). This program was Developed at AIIMS, New Delhi (Dept of Pediatrics, Child Neurology division) in collaboration with Delhi IIT, as part of preliminary work done under Dyslexia program under Child Neurology, AIIMS by Dr Sachendra Badal in 2019 (Chief guide ; Professor Sheffali Gulati). This was further adapted to enhance learning by AI technique as part of research work by Dr Sayoni Roy Chowdhary in 2021 (Chief guide; Professor Sheffali Gulati).
• It is a cloud-based application design, which uses a bottom-up strategic learning approach. It was developed for Indian context based on Components of effective remedial program as recommended internationally. It is based on the principle of reinforcement learning with primary focus on phonic awareness, phonics and sight word recognition.
• Structure and Duration - It consists of introductory teaching session which would be administered by the primary investigator for teaching 8 modules over 5 classes (32 hours each) over 1st 10 days using online platform. This would be followed by AI based structured remediation under parental supervision for 1 hr/day, 7 days/week for total 6 weeks . A total of 8 modules would be taught through the AI enabled program.
• Follow up would be done at set intervals after intensive training for 10 days: weekly for two weeks and subsequently fortnightly for rest of duration to ensure proper use of intervention by the primary investigator using telephonic calls to check for compliance. Total number of online sessions, total count of attempted words and total count of correctly pronounced words would be monitored by the web application. Compliance will be defined by (total 52 hours of intervention) Good: > 80% i.e. > 42 hours of intervention; Average: 60-80% i.e. 31-42 hours of intervention; Poor: \<60% i.e. \< 31 hours of intervention.

rTMS procedure

• Handedness will be determined via Edinburgh handedness inventory - Laterality Index
• Electromyography will be recorded with surface electrodes from the right Abductor pollicis brevis (APB) using a belly tendon montage on the dominant hand
• A figure of eight TMS coil connected to stimulator coil will be placed over the contralateral motor cortex tangentially to the scalp with the handle pointing backward. The motor hot-spot will determined as the location on the scalp where TMS produced the largest MEP from the APB at rest
• Obtaining Motor Cortex Threshold- Lowest stimulator output over the motor hot-spot that could elicit EMG response in APB in 5 out of 10 trials
• Prior to start of each cycle, on day 1 cortical excitability parameters will be recorded using standard protocols for RMT (Resting Motor Threshold), MEP (Motor Evoked Potential), LICI (Long interval intracortical inhibition) and SICI (Short interval intracortical inhibition).
• Site of rTMS - established using 10-20 EEG system at Left IPL (inferior parietal lobule , corresponding to angular and supramarginal gyrus), and left STG (superior temporal gyrus) at P3 and P5 respectively,
• TMS Protocol used for each session - 5 Hz High frequency rTMS, at 100% RMT, 60 pulses per train, total 10 trains , gap between 2 consecutive trains of 30 sec
• Duration - Total 3 cycles per participant will be administered.1st cycle of 10 sessions over 10 consecutive days. 2nd cycle of 5 sessions over 5 consecutive days. 3rd cycle for 5 sessions over 5 consecutive days. Gap of 6 weeks between 2 consecutive cycles.
• Participants will be monitored for adverse events during and after the study period. Appropriate measures such as immediate supine placement upon symptoms or signs of hypotension, intranasal midazolam in case or seizures will be taken.

Follow up

• Collection of data at 6 weeks - Cortical excitability parameters, GLAD, CTOPP-2, WRAT-5 scores and CGI-C (Clinical Global Impression of Change) for both teacher and parents.
• Collection of data at 12 weeks - Cortical excitability parameters, GLAD, CTOPP-2, WRAT-5 scores and CGI-C for both teacher and parents.
• At 12 weeks (±7 days) fMRI will be repeated for those 10 sequential participants whose legal guardians had given consent for the same.
• Follow up after 3 months post intervention - 24 weeks - Reassessment of the participant with GLAD, CTOPP-2 and WRAT-5 tests, Cortical excitability parameters and CGI- for parents and teachers.
• Participants will be assessed telephonically every week for compliance with remediation therapy and any adverse effects of rTMS and will also be instructed to report any adverse event as and when they occur.

Control - For controls investigators will be using the retrospective data of children with dyslexia who received AI enabled remedial intervention alone (for 6 weeks) as a part of a previous study by Dr Sayoni under the guidance of Prof Sheffali Gulati as Chief guide.

•Assessment Tools

1. GLAD (Grade level Assessment Device Tool): Assesses the level of academic performance in children up to VI standard , developed at National Institute of Mental Health, Secunderabad and has been validated in Indian population. Format I contains Test booklets of class I to class IV worksheets in English, Hindi \& Mathematics and Tasks are verbal/ gestural and written responses to questions. Number of item varies in each section and also in each class level, which is scored and converted to a percentage. Format II is to be used by the clinician for noting observations while the child is performing on format I and has three sections and a summary sheet. There are three broad categories: Independent (score >70%), Instructional (score 40-70%) and Frustrational level (score \<40%).
2. CTOPP-2 (Comprehensive Test Of Phonological Processing Tool, 2nd edition): The Comprehensive Test of Phonological Processing (CTOPP) was published in 1999 and revised in 2013 to meet the need for an assessment of reading-related phonological processing skills. The CTOPP-2 has four principal uses: to identify individuals who are significantly below their peers in important phonological abilities, to determine strengths and weaknesses among developed phonological processes, to document individuals' progress in phonological processing as a consequence of special intervention programs, and to serve as a measurement device in research studies investigating phonological processing. It is individually administered to children between 4 years to 24 years 11 months with a total testing time of 40 minutes. CTOPP2 measures phonological abilities in four domains; as composite scores of phonological awareness, alternate phonological awareness, phonological memory and rapid symbolic/ non-symbolic naming through core and supplemental subsets (elision, blending words, phoneme isolation, blending nonwords, segmenting nonwords, memory for digits, nonword repetition, rapid digit naming, rapid letter naming, rapid color naming, rapid object naming). Composite scores of CTOPP2: Phonological Awareness Composite Score (PACS) comprises the standard scores of three subtests-Elision, Blending Words, and Sound Matching-for 4 through 6 year olds, and Elision, Blending Words, and Phoneme Isolation for 7 through 24 year olds. The PACS represents the examinee's awareness of and access to the phonological structure of oral language. Phonological Memory Composite Score (PMCS) comprises the standard scores of two subtests: Memory for Digits and Nonword Repetition for all individuals. The PMCS represents the examinee's ability to code information phonologically for temporary storage in working or short-term memory. The Rapid Symbolic Naming Composite Score (RSNCS) comprises the standard scores of two subtests: Rapid Digit Naming and Rapid Letter Naming for all individuals. The RSNCS measures the examinee's ability to include efficient retrieval of phonological information from long-term or permanent memory and execute a sequence of operations quickly and repeatedly. Rapid Non-Symbolic Naming Composite Score (RNNCS) comprises the standard scores of two subtests-Rapid Color Naming and Rapid Object Naming and offers an alternative for young children, ages 4 through 6 year olds, not familiar with letters and numbers. The RNNCS measures the examinee's ability to include efficient retrieval of phonological information from long-term or permanent memory and executing a sequence of operations quickly and repeatedly using objects and colors. The Alternate Phonological Awareness Composite Score (APACS), an alternate composite for measuring phonological awareness is available for 7 through 24 year olds. Combining the scaled scores from Blending Nonwords and Segmenting Nonwords forms it. The APACS measures the examinee's phonological awareness exclusively with nonwords. The CTOPP-2 yields six types of normative scores: age equivalents, grade equivalents, percentile ranks, subtest scaled scores, composite indexes, and developmental scores.
3. WRAT-5 (Wide Range achievement test, 5th edition): It is a tool applied for children more than 5 years of age. The subsets include Word reading, Sentence comprehension, Spelling, Math computation and Reading composite. It is available in Indian version (WRAT-5 India blue form) and validated for Indian population (aged 6 years to 19 years, 11 months) since 2019.