receiving electrocardiogram (ECG) data associated with a patient;

determining that the ECG data is valid for use with a trained detection model, comprising providing the ECG data to a trained quality model, wherein the trained quality model is trained to detect low quality input data within the ECG data based on a set of quality indices, wherein the low quality input data is not valid for use with the trained detection model, wherein the trained detection model is associated with detection of a medical condition;

transforming the ECG data into a detection model input for the trained detection model;

at the trained detection model, determining a prediction metric based on the detection model input, wherein the prediction metric is indicative of a likelihood that the patient has the medical condition;

determining that the prediction metric is greater than a threshold value;

in response to determining that the prediction metric is greater than the threshold value, based on the prediction metric, determining an output indicative of the patient and the medical condition; and

providing the output to a user.