Personalized modelling of patient lab readout and application to understanding age-related disease
Approved Research ID: 64658
Approval date: January 11th 2021
Physicians are evaluating patients health using a variety of laboratory tests and additional diagnostic procedures. The interpretation of lab results is using standard "normal" ranges: patients with one or more lab readouts that fall out of the normal range are subjected to further diagnostics and possible treatment. It is however clear that not all healthy individuals are similar in their lab readout and that even lab values that are within the normal range are extremely useful for understanding ongoing disease processes in patients. Our study is aiming at the development of new computational tools that interpret the entire patient's clinical history to devise quantitative scores for the personalized "normality" of current lab test. These scores take into account patients age, gender, genetic background and more, and are sufficiently powerful to consider the holistic clinical state of an individual rather than defining each lab value as universally normal or abnormal. For these models to become universally applicable, we are seeking validation in multiple communities and populations, including the UK biobank cohort.