I am working on a project to study the association between change in the value of a quantitative variable (a predictor) with the occurrence of an outcome (a binary variable). This quantitative variable called DH is derived from radiographic imaging.
In the real world, this quantitative variable improves with the ongoing treatment however the question of interest is if this improvement is statistically associated with clinical improvement such as less occurrence of chest pain. Data consists of serial values of DH per patient along with the occurrence (or non-occurence) of chest pain at the time of imaging.
I am wondering which statistical model do I have to apply. Looking forward to your advice.