UNSTRUCTURED
The age of deep learning has brought high-performing diagnostic models for a variety of healthcare conditions. Deep neural networks can, in principle, approximate any function. However, this power can be considered both a gift and a curse, as the propensity towards overfitting is magnified when the input data are heterogeneous and high dimensional coupled with an output class which is highly nonlinear. This issue can especially plague diagnostic systems which predict behavioral and psychiatric conditions that are diagnosed with subjective criteria. An emerging solution to this issue is crowdsourcing, where crowd workers are paid to annotate complex behavioral features in return for monetary compensation or a gamified experience. These labels can then be used to derive a diagnosis, either directly or by using the labels as inputs to a diagnostic machine learning model. Here, I describe existing work in this field. I then discuss ongoing challenges and opportunities with crowd-powered diagnostic systems. With the correct considerations, the addition of crowdsourcing into machine learning workflows for prediction of complex and nuanced health conditions can rapidly accelerate screening, diagnostics, and ultimately access to care.