Ipsilon App

We developed musically gamified cognitive personality analytic app that evaluates the user's learning traits with their cognitive-oriented personality.  It examines the executive functions, inhibitory control (impulsivity), working memory and visual attention.  It has been scientifically validated with the pilot studies with executive function deficit screening.  The same trait for the clinical version is shared among those who have ADHD and similar conditions which may involve learning deficits.  Music involves several cognitive domains that is perfect to screen the strengths and weaknesses while it can evaluate the learning capabilities in non-language-based way.  

We use music notes as visuospatial instructions and collect time-stamped finger tapping response time and its accuracy.  From those data points, we generate parameters with cutoffs to compile cognitive-personality learning types that have been categorized into 12 types.  We use machine learning algorithm to finetune the data for not only for the current assessment but also provide with the predictable outcome. 

By screening the learning types by our app and systematically enhance the strength while supplement the weakness in one's cognitive and learning traits not only help build successful and confidence-building learning experience, but also more well-balanced one.  The app is web-based, so it does not involve specific device (though some tables are more recommended over the others for screen sensitivity), and it is usable where internet is available.  We recommend quarterly use of our app to create the learning profile of the user and update it with necessary follow-ups as the user grows academically. 

Our app is usable on a tablet where internet is available, such locations as library and schools to help teachers manage students' academic progresses.  Learning deficits are often intertwined with the impulsivity and other cognitive traits that may be present due to conditions that are undiagnosed and/or unaware by the caretaker of the child.  By profiling the learning type, strength and weakness, it helps teachers to gather each student's learning information more systematically, creating the overview of students and help them guide each student with individual needs. 

We are already active as cognitive screening app for elderly and neurodegenerative disease patients and have been equipped with providing our screening and analyses with healthcare-grade data protocol.  The foreseeable difference is that our intended users are much younger with academic-purpose.  Our advantage is that our solution is gamified and it is much easier for users to adapt it, especially when the target age is set at k-8 students.  The foreseeable potential issues can be how to utilize our assessment effectively to reflect what needs to be improved and what needs to be enhanced in their learning for the child.  We see the importance in this project to work with the actual teachers who serve the community to ensure that our learning assessment results are well-understood and used by the teachers.  We are open to such "learning" experience and opportunity to serve the community for its best interests. 

We have done several pilot studies among Alzheimer's and Parkinson's disease patients and their healthy counterparts in the age group over 65.  Our cognitive-personality learning profiling analysis has been tested among older adults and younger volunteers (over 18) which needs to be adjusted to the much younger user age.   We also see the needs to work with the educators who serve the target communities to ensure the smooth relay between our assessments and what can be done at the actual educational arena to compensate for the weakness and fortifying the identified strength of each students.  

We use music playing as a part of cognitive-learning assessment response generating tool.  Music uses multiple parts of the brain which is ideal to evaluate executive functions and core cognitive domains that are necessary in successful learning.  We established an analytic method to identify learning types in 12 traits, and aggregate the data to enable predictive outcomes near future as the expected trajectory of the current learning ability.  By knowing what is your strength and weakness from our evaluation, it helps to tailor the learning in more systematic and wholistic way.  

By evaluating cognitive traits and how one tends to learn, it helps to establish a profile of one's study and help catalyze the learning in the most optimizing way. 

We use reinforcement learning which is a branch of machine learning - human learning is literally the analogue version of reinforcement learning and we mimic the learning behavior to finetune the predictive model of our solution.  Our model helps create the machine version of the learning profile of the user and enable the prediction of the achievable threshold in learning capabilities in the near future. 

Our clinical version of the same application has been in pilot studies (total N=72), in which the data generating application and analytic methodologies have been validated thoroughly.  We established a digital marker for cognitively healthy to identify mild cognitive impairment among older adults and such healthy marker cutoffs can be also translated into profiling learners.  While we established such clinical product, we had several younger volunteers with their cognitive data which has been used as core part of the user data to create the cognitive-personality learning profiling. 

In addition, our data-related protocols in clinical studies also helped us tremendously to abide with data compliance obligations mainly in EU and North America.

•Sakaki, Y. “Clinical application of music playing: quantification of learning to train and monitor the cognitive functionality – a journey of a pianist.” Medizinisch Wissenschaftlische Verlagsgesellschaft 2019. XPOMET©: Innovation in Medicine and Care. Ulrich H. Pieper and Alois G. Steidel, Eds. ISBN: 978-3-95466-436-8

•Torrado JC, Husebo BS, Allore HG, Erdal A, Fæø SE, Reithe H, et al. (2022) Digital phenotyping by wearable-driven artificial intelligence in older adults and people with Parkinson’s disease: Protocol of the mixed method, cyclic ActiveAgeing study. PLoS ONE 17(10): e0275747. https://doi-org.ezproxyberklee.flo.org/10.1371/journa...

•Sakaki, Y. (2022, February 15). Potential of piano playing as digital biomarker for detecting prodromal phases of dementia and feasibility of computational model establishment for its better understanding and predictive markers development (A. Kormilitzin, Dr., Ed.). From talks.ox.ac.uk website: https://talks.ox.ac.uk/talks/i...

•Armengaud E, Sakaki, Y., & Wenzek, H. Mild Cognitive Impairment Screening for Safe, Inclusive and Human-Centered Road Transport. European Association on smart systems integration (EpoSS), conference Paper, 2023.

Our clinical version has been built to combat healthcare access inequity.  We enable remotely operable selfcare for cognitive functionality of those who suffer from conditions that have dementia and/or cognitive decline as comorbidity.  From such approach, our cognitive-personality learning assessment version is also easily accessible as a web-based app which does not choose specific device to operate, and operable where fast internet is accessible.  Our app also uses gamification with music playing so that the assessment does not heavily rely on language which can be often a cause for potential deficit if the student is not fluent in English language.  Our app identifies the cognitive traits that are often associated with learning deficits that can be compensated by modifying the curriculum by the educator.  We help optimize the learning by identifying the strength of the students while suggesting what needs to be supplemented.  This way, we hope to offer more well-balanced learning experiences systematically. 

1 full time, 2 part time team members, and software developer subcontractor

We adapt our clinical version to accommodate the classroom environment and prepare the team with different user population for data aggregation.  Our business model also needs to adapt to larger user population to make the pricing more reasonable while business is sustainable.  For such, we are likely to expand our team dedicated for this specific project to operate separately from our clinical solutions. 

We believe that iIf we have large user populations, our pricing can be significantly lowered  than our clinical version and still remain as sustainable business.  We need to find the middle point in terms of operable pricing and run more thorough market analysis and operational assessments to ensure such capability.  

In terms of infrastructure for our operation, we may need to include the tablets as a part of our offering to designate certain numbers of the devices to each classroom to ensure the access to the device while renegotiation with our current cloud server and data compliance management subcontractor can be an additional option to keep the price low. 

We believe that education is something that is the most powerful "equipment" one can have to have successful future.  In order to offer it equally to unprivileged population, pricing and delivering capabilities for offering systematic analyses for cognitive and learning types for the large user population may be a challenge.  We also believe that working with the SOLVE and Bill & Melinda Gates Foundation can help us network with potential competitive subcontractors from their network to overcome such obstacles and successfully deliver our offerings.