M. Garciabonafe et A. Moragas, DIFFERENTIAL-DIAGNOSIS OF MALIGNANT AND REACTIVE CELLS FROM SEROUS EFFUSIONS - IMAGE AND TEXTURE ANALYSIS STUDY, Analytical cellular pathology, 12(2), 1996, pp. 85-98
The interrelationship of textural primitives which define morphologica
l texture can be estimated by quite different descriptors; the discrim
inant value of which varies considerably. In the present study three d
ifferent approaches to the texture analysis of nuclear chromatin were
assayed to correctly allocate 332 cells from a pool obtained from sero
us effusions (six malignant mesotheliomas, six reactive mesothelial pr
oliferations and five pleural metastases of lung adenocarcinoma). In a
ll cases, initial cytological diagnosis posed considerable problems an
d final diagnosis was established by histologic examination of surgica
l specimens. The three approaches were based on binarization of the im
age obtained by edge detection, gradient analysis and pattern spectrum
by morphological opening-closing, respectively. Characteristics affor
ded by each method were: (a) spatial distribution of heterochromatin,
besides geometric features, (b) features related to transitions and co
ntrast between dark and light chromatin primitives, and (c) granulomet
ric characteristics of the theoretically biphasic heterochromatin-euch
romatin image defined by mathematical morphology. The three methods we
re applied to the raw grey-tone image and did not require interactive
handling. Although each of the three approaches yielded relatively sat
isfactory results, with percentages of well-classified cells in the te
st set ranging from 61.45 to 67.47, the best results (78.31% of well-c
lassified cells) were obtained taking into consideration the three typ
es of variables (area, 2nd opening, 5th closing, and S.D. of the ampli
tude image). A point to be stressed is the considerably high proportio
n of correctly-allocated reactive mesothelial cells (82.0%) in a field
where subjective assessment commonly yields rather poor results. Neve
rtheless, classification yielded 14.8% and 3.3% false positives as ade
nocarcinomatous and malignant mesothelioma cells, respectively. In the
theoretical situation devised in the study, results on a cell-by-cell
basis are encouraging and suggest that a textural approach might be u
seful in a dedicated expert system or on a more real case-by-case basi
s. Copyright (C) 1996 Elsevier Science Ireland Ltd.