It’s time to garner the potential of digital health transformation
A study carried out in Spain and Portugal using data on the incidence of different types of cancer once again illustrates the potential of health research when access to large repositories of labeled data is available and the appropriate algorithms are applied. Another recent study in the United States using data from blood samples of military personnel has established a link between multiple sclerosis and the incidence of mononucleosis.
Other studies, carried out by technology companies such as Apple (on cardiovascular health or menstrual symptoms) or Google (large longitudinal study), use data obtained from users of their devices or applications, and producing interesting data that enables significant advances in health care. In fact, the success of Apple’s HealthKit and ResearchKit demonstrates that patients, when they understand that their data is adequately protected, tend to prefer it to be used to improve research aimed at making progress in the treatment of their ailments.
The application of machine learning to the medical records of millions of patients offers such a high potential for medical research that delaying its possibilities is simply irresponsible. What do machine learning algorithms need? The answer is obvious: correctly labeled data, such as that generated by physicians on a daily…
