AMC harnesses AI for early kidney transplant rejection diagnosis
It focused on leveraging surface-enhanced Raman spectroscopy to detect low-concentration analytes.
A research team at South Korea’s Asan Medical Centre (AMC) has utilised biomarker detection and artificial intelligence (AI) based discriminant technology for early diagnosis of kidney transplant rejection using a small amount of serum.
The team focused on leveraging surface-enhanced Raman spectroscopy (SERS) to detect low-concentration analytes by increasing sensitivity through localised surface plasmon resonance (LSPR) modes of metallic materials.
They hypothesised that analysing Raman patterns produced by different biomarkers in the serum could improve the precision of rejection diagnoses when combined with AI technology.
The team analysed the prognosis of rejection in patients and classified them into groups with no transplant rejection, antibody-mediated rejection, and T-cell-mediated rejection.
They validated their AI-based Raman signal analysis by evaluating post-transplant kidney damage and function.
According to AMC, linear discriminant analysis results showed 93.53% accuracy and partial least squares discriminant analysis reached 98.82% accuracy.