CLASSIFICATION OF RED-BLOOD-CELLS AS NORMAL, SICKLE, OR OTHER ABNORMAL, USING A SINGLE IMAGE-ANALYSIS FEATURE

Sickle cell anemia is a disease for which there is currently no effect ive treatment. One method of evaluating clinical status is the countin g of cell types based on morphology. There is a need for a rapid, repr oducible method, superior to human inspection, for classification of t hese cells. Quantitative digital-image analysis is being applied to th is need. Blood from 24 patients with sickle cell anemia (SS) and SC di sease and ten hematologically normal volunteers (AA) was stressed by b ubbling with nitrogen. One hundred fifty cells were analyzed from each sickle specimen, and 100 were analyzed from each nonsickle specimen. Expert observers classified each cell as normal (N), sickle (S), or ot her abnormal (A). Cells were analyzed with a custom, high-resolution i mage-analysis instrument. A total of 42 features including metric, opt ical density-derived, and textural features were extracted. The metric feature Form Factor (4 pi Area/Perimeter(2)) was selected by recursiv e partitioning analysis as the sole feature needed for segregating cel ls into the classes of N, A, and S. The agreement of automated classif ication (using cutpoints determined by recursive partitioning analysis ) with a human expert for specimens from individuals with sickle cell anemia was 89% for N-, 73% for A-, and 92% for S-classified cells. For specimens from AA individuals, the agreement was 92% for N and 76% fo r A. For specimens from individuals with sickle cell anemia, rates of agreement between two human experts were compared and found to be 86% for N, 84% for A, and 80% for S. For specimens from AA individuals, th e agreement was 90% for N and 87% for A. (C) 1994 Wiley-Liss, Inc.