Research project uses artificial intelligence to try to tackle organ shortage for kidney transplant

Translation by Diana Taborda

Approximately 50 percent of all donated kidneys are rejected because the current methods for classifying kidney biopsies, which are used by doctors to determine whether a donor's organ is suitable for transplant, are subjective and error-prone.

To overcome this problem, a team of researchers from the Faculty of Medicine of the University of Coimbra (FMUC) and doctors from the Coimbra Hospital and University Centre (CHU) - in collaboration with the University of Buffalo, USA - are developing an intelligent algorithm to help specialists in the complex task of evaluating deceased donor kidney biopsies at the time of procurement, the so-called 'zero time'.

The biopsy provides vital information for assessing the quality of the organ and checking that it is suitable for transplantation into the recipient. However, the current methods for classifying kidney biopsies are visual, semiquantitative and sometimes imprecise, i.e. "it is a procedure resulting from human observation and depends heavily on the experience of the specialist interpreting the results, in addition to being difficult to predict the evolution of the organ in the long term. It is a semiquantitative analysis because it depends on the way the specialist classifies an image. The doctor observes, evaluates certain structures and decides on the quality of the organ. It is a very manual, tedious and subjective process, which could lead to the waste of organs that could be used," explains Luís Rodrigues, principal investigator of the project, which was recently awarded a prize by the Portuguese Society of Nephrology.

"The best therapeutic option to treat patients with severe kidney failure, who depend on haemodialysis, is a transplant, but we have the huge problem of organ shortage", says Luís Rodrigues.

In Portugal, the prevalence of treated end-stage renal disease is one of the highest in Europe, and the waiting list for transplantation is growing every year, so there is an urgent need to develop tools to increase the number of organs available for transplantation and optimise their allocation, thus improving the survival and quality of life of organ recipients," stresses the doctor from the Coimbra Hospital and University Centre (CHUC).

The project team is convinced that Artificial Intelligence (AI) can be an important ally in preventing organ waste. Luís Rodrigues adds that the main objective of the research, which is part of his doctoral thesis supervised by Professor Rui Alves, is to develop an intelligent image analysis algorithm that "not only increases the effectiveness and accuracy of morphological characterisation of donated kidneys, but also improves organ allocation, with a subsequent longevity match between donor and recipient".

"The data obtained through computational analysis can significantly improve our ability to predict transplant outcomes and optimise organ use and allocation. The longer the organs we transplant last, the lower the likelihood of a second transplant and the lower the probability of needing another donor," says the FMUC researcher.

If the development of a new artificial intelligence-based diagnostic tool that allows a robust and systematic approach to kidney biopsies is successful, Luís Rodrigues expects that "between 10 and 25% of the organs that are currently rejected could be used". However, he points out that there is still a lot of work to be done before the team has a reliable algorithm that can be applied to kidney pathology. For now, the researchers are collecting information from kidney patients involved in the project. Data from clinical samples will be used to learn and train the algorithm.

The project takes place at the Renal Transplantation Unit of the CHUC, with the collaboration of the departments of Nephrology, Urology, Renal Transplantation and Pathology. In addition to Luís Rodrigues and Rui Alves, Vitor Sousa, Rui Almeida, Ana Pimenta, Catarina Romãozinho, Lídia Santos, Edgar Silva and Arnaldo Figueiredo are also part of the team.