Application of genetic programming for the differential diagnosis of acid-base and anion gap disorders

Abstract

Differential diagnostics of very complex disorders in the acid-base status and in the accompanying electrolyte balance belongs to the most difficult diagnostic procedures in clinical practice. The goal of our study was to create computer genetic programs, able to aid the diagnostics of acid-base disorders, taking into account a set of laboratory gasometric and electrolyte measurements. Seven (7) single acid-base disorders, eleven (11) double acid-base disorders and six (6) triple complicated disorders with accompanying anion gap alterations were approached by our study. A set of simulated laboratory measurements was prepared, providing 250 results for the evaluation of the fitness function of the designed computer genetic programs, plus some additional results for testing.