About this Research Topic
Please note that abstract submissions are not mandatory for a submission to this collection and abstracts can be submitted after the deadline too.
Mental health disorders are underlain by a wide diversity of influencing factors, and they exert their impact across multiple domains in a patient's life. As such, mental health
research has greatly benefited from the proliferation of large and diverse databanks with many and new types of data, often covering a large sample of or even whole populations. However, traditional epidemiological and statistical techniques have proven to be insufficient to tackle the complexity of mental illness. While such data have spurred important advances in the area of mental health, they have also introduced new limitations that risk stalling progress.
In this context, machine learning provides, in theory, novel opportunities to leverage the potential of databanks and propel population mental health research forward. Indeed, after a slower introduction compared to other health specialties, the popularity of machine learning in this field has soared in recent years. Unfortunately, on many occasions, the use of machine learning has overshadowed epidemiological practice, thus limiting the production of results that have genuine clinical relevance. Without careful integration, there is a real risk of losing the trust and confidence that machine learning has slowly gained within the clinical field.
This Research Topic aims to illustrate examples of mental health research that exploit new data sources or their linkage using machine learning in a clinically/epidemiologically sound way. We welcome Original Research, Systematic Review, and Methods submissions, particularly (but not limited to) those focussing on self-harm and suicide prevention, severe mental illnesses, physical and mental health co-morbidity, children and young people, and excluded and under-represented groups. Submissions must pay special attention to the full or partial validation of machine learning models and their results, to support their clinical relevance.
Mental health disorders are underlain by a wide diversity of influencing factors, and they exert their impact across multiple domains in a patient's life. As such, mental health
research has greatly benefited from the proliferation of large and diverse databanks with many and new types of data, often covering a large sample of or even whole populations. However, traditional epidemiological and statistical techniques have proven to be insufficient to tackle the complexity of mental illness. While such data have spurred important advances in the area of mental health, they have also introduced new limitations that risk stalling progress.
In this context, machine learning provides, in theory, novel opportunities to leverage the potential of databanks and propel population mental health research forward. Indeed, after a slower introduction compared to other health specialties, the popularity of machine learning in this field has soared in recent years. Unfortunately, on many occasions, the use of machine learning has overshadowed epidemiological practice, thus limiting the production of results that have genuine clinical relevance. Without careful integration, there is a real risk of losing the trust and confidence that machine learning has slowly gained within the clinical field.
This Research Topic aims to illustrate examples of mental health research that exploit new data sources or their linkage using machine learning in a clinically/epidemiologically sound way. We welcome Original Research, Systematic Review, and Methods submissions, particularly (but not limited to) those focussing on self-harm and suicide prevention, severe mental illnesses, physical and mental health co-morbidity, children and young people, and excluded and under-represented groups. Submissions must pay special attention to the full or partial validation of machine learning models and their results, to support their clinical relevance.
Keywords: population mental health, data linkage, routine data, machine learning, epidemiology
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
