The international Banff classification is commonly used by physicians to analyze biopsies of the transplanted kidneys. This is used to diagnose the rejection of kidney transplants; however, this classification has become more complex which leads to variability in its application and misclassification. These misclassifications can lead to physicians failing to modify or incorrectly modify the treatment regimen to protect the patient.
The use of Artificial Intelligence (AI) in the industry has become far more prevalent as it has many different uses including the ability to make decisions based on rules embedded within its programming. An algorithm was created by researchers for this very purpose to use with the Banff system to help prevent misdiagnoses of kidney transplants. This was then studied in 2 prospective clinical trials.
During the trials, the algorithm successfully reclassified 83 of 279 antibody-mediated rejection cases and 57 of 105 T cell-mediated rejection cases in adults. The algorithm obtained similar rates in a pediatric study at 30.77% for both antibody and T cell-mediated rejection. The success of the algorithm resulted in important clinical implications for kidney transplant recipients. The algorithm is designed to help pathologists make more accurate, efficient diagnoses and fewer errors. The use of AI and Machine Learning (ML) has become an important topic with the FDA as they have reviewed a growing number of devices using AI and ML to improve the functionality of their legally marketed devices across many different fields of medicine.
If you need support in understanding how AI use can apply in medical devices, the experts at EMMA International can help! Call us at 248-987-4497 or email us at firstname.lastname@example.org to learn more.
Taylor, N. P. (2023, May 5). Algorithm may help avoid 40% of kidney transplant rejection misdiagnoses: Study. Algorithm may help avoid 40% of kidney transplant rejection misdiagnoses: study. https://www.medtechdive.com/news/algorithm-predict-kidney-transplant-rejection/649521/