CLASSIF1 Classification

• < SLE

< fig.3      fig.4: Reclassification of the Learning Set

- Displayed are the triple matrix patterns of the first 10 SLEDAI0-8 and SLEDAI>8 patients of the immunophenotype classification of fig.2A.
- The disease classification mask fig.3A for the reference group of patients (SLEDAI 0-8) typically contains a sequence of (0) characters because the majority of the parameter values is located between the two percentile thresholds 10-90%, 15-85%, 20-80%, 25-75%, 30-70% of the value distributions. As an example, 80% of the values are (0) for the percentile pair 10-90%, 10% are below (-) the lower percentile and 10% above (+) the upper percentile.
- Unknown patients are classified according to the highest positional coincidence of their patient classification mask with any of the disease classification masks.
- The degree of coincidence between the patient mask and the best fitting disease mask is expressed by the mask coincidence factor. The coincidence factor is 1.00 for patient #DO15 with all triple matrix characters (0) (arrow <= ). The factor is also 1.00 for patient #DO02 (arrow <= ) despite the fact that not all triple matrix characters are (0). The reason for this is that triple matrix characters in those parts of the reference value distribution curves which are not occupied by the triple matrix characters of the SLEDAI>8 disease mask count as positional coincidences for the SLEDAI0-8 disease mask. With the SLEADI>8 disease mask being: +--++ the characters 00000 and -++-- belong to the SLEDAI0-8 disease mask.
- Eighteen patients are correctly classified while the two patients #DO12 and #DO04 are misclassified

• < SLE