With over ten thousand people on the waiting list for kidney transplantation, reduction of graft loss is of utmost importance. In the past decades there has been a vast increase of knowledge surrounding the mechanisms involved in acute/active renal allograft rejection, which has brought good therapeutic to the clinic. Unfortunately, the complex cellular interactions resulting in the key components of chronic rejection, interstitial fibrosis and tubular atrophy (IFTA), remain poorly understood. Alternatively activated macrophages play a central role in this complex network of cellular interactions, regulating T-cell behavior in the tissue microenvironment, inducing tissue repair and remodeling effects, and conferring a delicate equilibrium between immunosuppressive beneficial and fibrosis-inducing detrimental effects.
Key is combining dynamic mathematical models of macrophage-associated immunological and metabolic regulation with advanced biopsy evaluation and in-vitro experiments to dissect processes in the renal interstitium that converge into IF/TA.
In our study we combine Deep Learning with Periodic Acid Schiff (PAS) and multiplex immunohistochemically (mIHC) stained histopathological slides to extract objective information from protocol biopsies with the aim to answer research objectives revolving around kidney allograft rejection.
Media & Publications
Kim M (2018, March 29). ‘A Machine Learning Algorithm Segments Renal Tissue into Healthy and Pathological Structures.' Bioengineering Today