A timely medical diagnosis is crucial to increasing the chances of successful treatment of diseases, as it allows for early intervention and improved patient outcomes. In recent years, intelligence techniques have played an increasingly important role in this field, facilitating the early detection of conditions such as different types of cancer and heart disease and in the application of various medical treatments. Intelligence techniques can analyze large volumes of medical information accurately and quickly, detecting patterns that might go unnoticed by humans. In this work, the design of Interval Type-3 Fuzzy Systems is proposed to perform the medical classification. The design is performed seeking the configuration of 3 Generalized bell Type-3 Membership functions and fuzzy rules. The proposed design is tested using 3 health information datasets: Cryotherapy, Haberman’s Survival, and Immunotherapy, where the accuracy averages achieved are 88.15, 76.50, and 83.61, respectively. The results achieved are compared versus Trapezoidal Type-3 Membership Functions.