Idea Description
Supplementary Information
Innovation 'Elevator Pitch':
Early identification lies at the heart of the Mental Illness Prevention strategy. PredictX’s proven machine learning model for identification of not school ready children forms the basis for wider application.
Overview of Innovation:
The “Get It Right, First Time” program is an ambitious effort to transform the mental health treatment to focus more on early diagnosis and intervention. The benefits of early intervention are clear and well documented. PredictX (PX) believe that we can the application of Machine Learning (ML) based systems improves the identification of at-risk individuals, especially children.

At Essex CC, we developed and deployed a working ML system to identify children and households where the children had a higher probability of not being school ready. As the report states, “ the most productive and cost-effective element of any prevention of mental ill-health strategy will be during a child’s early years.” The PX school readiness model has proven to be effective in the early identification of not school ready children - beyond previous methods. The system not only identifies at-risk children for early intervention but also indicates the most impactful intervention for a locality.

The PX model not only applies to school readiness but is a flexible framework that adapts to new use cases and available data to solve other similar challenging issues. The model uses data from housing, education, social care, police, health, benefits, and population can use this data for applications beyond school readiness. For instance, early work suggests that the model will be effective predicting domestic abuse - another causal precursor of childhood mental health issues. Because such a wide range of data is accessed, the model is resilient to sparse or missing data from any particular data source.

We believe that the school readiness model is one branch among several that leads to a much more holistic and accurate identification of individuals, families, and locations at risk of mental health issues. Moreover, the model can be tuned to further identify those who would be positively impacted by intervention thus aiding the most effective resource allocation of the intervention team.

The model has the further advantage of having real-world analogues and explainability. This means that the model will aid the experts by identifying the key causal factors most prevalent in a particular locality, thereby systematically aiding the work with Schools and Education to identify the most critical vulnerabilities and risk markers for mental illness and aiding building the most impactful preventative resilience and wellbeing practices.

Please see:
Stage of Development:
Close to market - Prototype near completion and final form may require additional validation/evaluation and all CE marking and regulatory requirements are in place
Similar Content3


Very interesting innovation - have any clinical trials been done - or feedback from trials already conducted?

Created by

Share and Follow