This paper proposes a new approach for medical diagnosis based on laboratory empirical data using fuzzy set theory. The presented approach is based on intuitionistic fuzzy set (IFSs) theory as a special kind of fuzzy set using two new distance measures between intuitionistic fuzzy sets as a tool in pattern recognition. We will apply them in a part of progress of medical diagnosis in order to recognize Bacillus colonies and then we compare its results with similar measures in regular fuzzy. Numerical results show that the proposed measures are more applicable in medical diagnosis than the similar measures in regular fuzzy ones.