So I'm possibly interested in machine-learning models that receive laboratory data as inputs, then hopefully are able to churn out some kind of risk stratification to assign that particular patient.
But the first thought that crossed my mind was all the typical pitfalls associated with Machine Learning especially in relation to health data (e.g. the sheer downside of a false positive or false negative).
Are models like these used widely in this space? What vendors etc are generally the go-to for something like this, if accreditation is one of our main concerns. We're Asia-based if that makes a difference!
0 nhận xét:
Đăng nhận xét