Blood group (ABO) incompatible transplants carry an increased risk of rejection. This risk could be dramatically reduced by the removal and suppression of the antibody types that would attack the donor organ. A prerequisite for this removal is an experimental procedure that can estimate the binding affinities of multiple antibodies from patient blood samples. This paper presents the usage of surface plasmon resonance (SPR) experiments with a pre existing mathematical model and a recently created expanded version of it that can estimate multiple antibody binding affinities from parallel experiments. SPR experiments were conducted on purified patient antibody samples of the IgG and IgA isotype, as well as mixed antibody from the other isotypes. The result of these experiments was analyzed with both mathematical models, and the expanded model was demonstrated to give a vastly improved fit. Estimates of antibody binding affinity were compared between samples and the non IgG/IgA protein sample was seen to have the highest binding affinity.