Outlier rejection method in an automatic clinical analyzer

• US 7,039,561 B2
• Filed: 06/10/2004
• Issued: 05/02/2006
• Est. Priority Date: 06/10/2004
• Status: Expired due to Fees

• Current Assignee: Siemens Healthcare Diagnostics Incorporated (Siemens AG)
• Original Assignee: Dade Behring, Inc. (Siemens AG)
• Inventors: Wei, Tie Quan
• Primary Examiner(s): Barlow, John
• Assistant Examiner(s): KUNDU, SUJOY K
• Application Number:
  US10/865,081
• Publication Number:
  US 20050288908A1
• Time in Patent Office: 691 Days
• Field of Search: 702/191, 702179-181, 356319-326
• US Class Current: 702/191
• CPC Class Codes: G16H 10/40 for data related to laboratory analysis, e.g. patient specimen analysis

Abstract

Automatically rejecting outlier measurements in an analytical method employing multiple photometric values by determining the photometric value that is involved in the pair of photometric values having the greatest variance between individual mean rates and the mean rate of all of the multiple photometric values.

Claims

1. A method for rejecting outlier measurements in an analytical method having multiple photometric signal values by:

identifying and rejecting the photometric value that forms one of the pair of photometric values having the greatest variance between all possible mean rates between the totality of multiple photometric values and the mean rate of all of the multiple photometric values, wherein said identifying and rejecting comprises;

a) identifying the totality of different pairs of measurement values;

b) calculating rates of data changes for all of said pairs of measurement values;

c) calculating the mean rate of all of the data changes;

d) calculating the variance between the mean rate and the plurality of data changes;

e) identifying the pairs of measurement values having the maximum variance between the mean rate and the plurality of data changes;

f) re-calculating the mean rate of the plurality of data changes, excluding a first one of the two measurement values identified in step e);

g) re-calculating the mean rate of the plurality of data changes, excluding a second one of the two measurement values identified in step e);

h) identifying the absolute values of the differences between the mean rate and the re-calculated mean rates of step f) and step g); and

, either,

i) rejecting the first one of the two measurement values if the absolute value of the difference between the mean rate and the re-calculated mean rate of step f) is larger than the absolute value of the difference between the mean rate and the re-calculated mean rate of step g), or

j) rejecting the second one of the two measurement values if the absolute value of the difference between the mean rate and the re-calculated mean rate of step g) is larger than the absolute value of the difference between the mean rate and the re-calculated mean rate of step f).

2. The method of claim 1 wherein step e) comprises numerically ranking all mean rates from lowest to highest and then repeating steps f) through j) for either one of the lowest or the highest ranked mean rates.

3. The method of claim 1 further comprising repeating steps a) to j) at least once.