Too preliminary to change practice; no validation data against real-world clinical populations or comparator diagnostic tools have been provided, so clinicians should not yet rely on this automated classifier.