Stanford researchers have developed a contrastive learning approach that can significantly reduce the amount of labeled electrocardiogram (ECG) data required for downstream healthcare tasks, such as arrhythmia identification.
Scientists at Stanford have developed a machine learning program with broad potential for diagnostic applications which analyzes mass spectrometry data profiling metabolites in a patient sample ("metabolomics" data) and predicts infection status.
The CheXbert labeler accurately detects the presence or absence of 14 common medical conditions in radiology reports, converting unstructured radiology text into a structured format.
Stanford researchers have developed CheXpert which can reduce noise and identify several pathologies on x-rays with very high accuracy via machine learning. CheXpert can read photos of x-rays from a mobile phone and is robust to noise.