DEPARTMENT FOR SCIENCE, INNOVATION AND TECHNOLOGY

Screen4SpLDs - Development of an Automated Pre-Screening Tool for Specific Learning Disabilities in Children.

Project disclaimer

Disclaimer: The data for this page has been produced from IATI data published by DEPARTMENT FOR SCIENCE, INNOVATION AND TECHNOLOGY. Please contact them (Show Email Address) if you have any questions about their data.

Description

Specific Learning Disabilities (SpLDs) is a common term in today's society, which manifests in different ways and can cause various difficulties in daily life. For one person it might be the lack of attention, for another, it might be struggling to read fluently or conduct basic mathematical calculations; these are different groups of Learning Disabilities. Early detection and treatment of SpLDs are crucial, as it enables the start of interventions that support the best outcomes for children living with SLDs. Not addressing SLDs at a young age has a major influence on development into adulthood and results in a high economic cost, exceeding the lifetime costs of asthma, intellectual disability, and diabetes have a huge shortage of special educators to conduct SLDs screening and subsequently providing treatment post diagnosis. There are nearly 90% of the world's children reside in Low and Middle-Income Countries (LMICs). The challenge of early detection and early intervention of SpLDs is exacerbated by limited expertise, including limited screening, diagnostic and treatment resources in LMICs. For instance, in the Global South, the skilled human resource and tools to assess SpLDs are very limited. Thus, these children are undiagnosed and negatively reinforced by the community by stigmatizing and labelling them. These factors all lead to low self-esteem and behavior problems that further interfere with their ability to learn. In vulnerable communities, which are often already poverty stricken, this operates as a vicious cycle, simply because optimal education is the main method of breaking this vicious cycle. We aim to target these developmental issues by developing and piloting low-cost mobile app-based solution for the screening of SpLDs that will lead to early intervention. Specific learning disorder may affect handwriting in a way that can be visually distinguished. The purpose of the proposed research is to evaluate the ability of deep learning to distinguish between those who have SpLDs and those who do not, from their handwriting. The proposed solution requires no more than taking a photo of the handwritten image on a mobile phone and passing it to the prediction model and getting the prediction results. Based on the proposed solution, the SpLDs screening can be conducted at home, in a school study area without any additional special setting. The important factors of this app are simplicity, ease of use, less training requirement, the accuracy of the results, and reliability. This app can serve from individual to national level for screening SpLDs in children. This will reduce the burden of the shortage of special educators, and this will be a huge relief for LMICs. This will, in general, reduce the inequalities faced by vulnerable and marginalized children, by providing an opportunity to receive optimal health and educational services. This will lead to the improvement of quality education received by ALL which in turn will contribute to wider societal improvements. In addition to the direct impact on the child, the spillover effects on the family and community development are significant. Further, creating an opportunity to screen a larger population will increase societal awareness of SpLDs and reduce the stigma

Objectives

The primary goal of this project is to develop an automated screening mobile app for SpLDs in children. Even though different issues are identified in children with SpLDs, the handwriting issue is noted as common in SpLDs. Therefore, by analysing the photos taken from notebook pages, the app could be able to predict the possibility of SpLDs. This will be achieved by the following objectives: 1.Develop smartphone app to scan students' handwritten notebooks and store the scanned images in a secured database. 2.Pre-processing steps such as normalisation of colour space, illumination and unwanted artifacts removal will be designed to enable comparison of scanned images. 3.Develop efficient SpLDs prediction model by employing appropriate deep learning algorithms. Apply this model to predict the handwritten images pre-processed from (2) as normal or SpLDs. 4.Apply explainable AI algorithms such as SHAP to show the visual explanation of the predicted result. 5.Implement efficient offline mobile based SpLDs screening app by integrating (1), (2), (3) and (4) and deploying into the mobile phone. This will provide an instant prediction result to classify the scanned images as normal or SpLDs along with visual explanation. 6.Conduct trails in local primary schools to validate the app. 7.Special educators and teachers will be asked to evaluate the ease of the mobile app.

---

Location

The country, countries or regions that benefit from this Programme.

South Asia, regional

Disclaimer: Country borders do not necessarily reflect the UK Government's official position.