Information, discrimination and divergence in cytology. V. General symmetry of total discrimination and total divergence.

Total discrimination and divergence are derived from Baye's theorem and based on backward (a posteriori) probabilities. Total forward (a priori) discrimination and divergence can be computed from a test matrix from which total backward discrimination and divergence were calculated for optimization of the classifications of gynecologic cytology (Papanicolaou smears) and quality control in a laboratory. The total forward discrimination and divergence appears to behave in parallel with the total backward discrimination and divergence, and the discrepancies in the backward and forward discrimination/divergence were the smallest near the optimum classification scheme. Conversely, the discrepancies between backward and forward discrimination and divergence may be helpful in finding the best classification scheme for gynecologic cytology. The general symmetries of the total backward and forward discrimination/divergence may be related to the human cognitive process of preferring symmetry and to the historic process in which cytologic classification followed histologic classification and continuously checked for matches in two directions--i.e., from histology to cytology and vice versa--thus resulting in the preservation of symmetry.     

Information, discrimination and divergence in cytology. IV. Quality control in diagnostic cytology

The performance of diagnostic cytology on Papanicolaou smears can be periodically monitored by calculating the total discrimination or the total divergence of the cytologic diagnoses against the histologic diagnoses on samples obtained by colposcope-directed biopsies. Using these measures, the annual performances of the Gynecologic Cytology Laboratory of the University of Minnesota between 1980 and 1988 were retrospectively analyzed. For those years, the total discrimination and total divergence behaved similarly and were sensitive to the performance of the total system, including specimen sampling errors and laboratory precision. The lowest limits of the permissible range of the total discrimination and total divergence were 0.15 and -1.21 decits, respectively, for a single-slide Papanicolaou test if an 80% "hit" rate was accepted as the lowest threshold for each category. The optimal numbers of category-states were not a sensitive indicator of the quality of a laboratory; i.e., the optimal number of diagnostic categories remained at three throughout the period studied.     

Information, discrimination and divergence in cytology. I. Theoretical aspects and definitions

The results of medical tests update the probability of diagnosis (diagnosability) of a patient from the population prevalence of a disease. This paper demonstrates a method of combining several mutually dependent tests as the sum of log odds ratios, and of separating the log odd ratios of the test results and those of prevalences. Extending Kullback's discriminant function between two states to m states (m greater than or equal to 2), it is shown that the information of a test can be measured by discrimination and divergence, which are well-defined measures in information theory and test theory. The performance of diagnostic cytology can then be compared to the results of histopathologic diagnoses or to a peer-reviewed consensus without the use of arbitrarily given scores or without unfounded assumptions about underlying continuous variables. They can also provide a way to optimize the categorization (classification) of Papanicolaou smears and a means of quality control for morphologic tests. These methods can be used to evaluate the performances of cytology laboratories that use two classifications consisting of different numbers of categories (classes) and states.     

Information, discrimination and divergence in cytology. II. Total discrimination as a measure of performance

Discrimination is the expectation of log odds ratios. Performance in diagnostic cytology, as compared to the results of histopathology or to a peer-reviewed consensus, can be measured by the total discrimination, which is a well-defined measure of information in information theory and test theory. The total discrimination, as a measure of performance, was calculated for the Gynecologic Cytology Laboratory of the University of Minnesota for the years 1985 through 1987, based on 3,545 sets of single-slide Papanicolaou smears and colposcope-directed biopsies. Similar calculations were made for the performances of physicians and technicians on peer-reviewed "clear-cut" Papanicolaou smears, as reported in a study from the Centers for Disease Control (CDC). For fair comparison of the two different sets of observations, the cytologic categories and corresponding histologic states were merged into five categories and states. The cytologic performance of our laboratory, tested against the histologic diagnoses, and the performances of the physicians and technologists tested against peer-reviewed check samples by the CDC were 0.53, 0.45 and 1.17 decits, respectively. These values generally agree with the conclusions derived by more conventional methods used by the CDC. When sampling errors of cytology specimens are taken into account, the performance of our laboratory, measured by the total discrimination of the cytology-histology confusion matrix, was approximately equal to the performance of the group of technologists studied by the CDC, which was in turn significantly better than that of the group of physicians studied by the CDC. This study demonstrates the value of using the total discrimination for quantifying the performance of a cytology laboratory, a cytopathologist or a cytotechnologist, without the arbitrary means usually used to evaluate such performances.     

Information, discrimination and divergence in cytology. VI. Biases and errors of measurement in small samples.

Total discrimination and total divergence have been shown to be useful measures of performance in diagnostic cytology. The sample size, Ns (observed number of cytology-histology pairs, essential components of a confusion matrix), may be small for the comparison of two or more laboratories or periodic quality control using observed values. From actual data from previous reports in this series, the best estimation of the confusion matrix of a population was obtained by fitting a Gaussian-type curve after pseudoscalar transform of ranks (row and column numbers). Small sample confusion matrices were generated by Monte Carlo simulation to 2.5 x 10(-7) resolution. To keep measurement biases within +/- 0.5 decits, we found that 100-200 samples of cytology-histology pairs were required in the best classifications of three, four and five category-states. At these sample sizes, measurement errors (standard errors) were also contained within +/- 0.5 decits. This study also confirmed that previously reported overestimated propagated errors in small samples were in fact overestimation and that their use for testing a null hypothesis was valid. The number of samples with indefinable statistics due to a zero denominator can be as high as 30% when the sample sizes are 500 for three, four and five category-state classifications. Biases due to small samples were positive for most category-states except for the optimum three category-states, in which bias changed to negative ("bias inversion"), and observed errors of discrimination and divergence paradoxically decreased as Ns decreased ("error-sample paradox") for a small sample size (Ns less than 700).(ABSTRACT TRUNCATED AT 250 WORDS)     

Significance of atypia in conventional Papanicolaou smears and liquid-based cytology: a follow-up study

The diagnosis of atypical squamous epithelial cells, borderline nuclear changes, is associated with some controversy, as it encompasses benign, reactive, as well as possible neoplastic conditions. The aim of this study was to evaluate the follow-up diagnoses of cytological atypia in conventional Papanicolaou smears (CP) and liquid-based samples by the ThinPrep Pap Test (TP). A total of 1607 CP smears from 1 January 2000 to 31 December 2000 and 798 TP samples from 1 January 2002 to 31 December 2002 diagnosed as atypia were included. The results show that the detection rate of atypia in cervical cytological samples was reduced by 41.3% (P < 0.001) in TP compared with CP. Cytological and histological follow-up data showed the presence of neoplastic lesions in 34.7% of patients screened by TP versus 22.3% of patients screened by CP, corresponding to a 55.6% increase in TP (P < 0.001). Follow-up diagnosis of mild dysplasia was seen more than twice as often in TP than in CP (12.8% versus 5.0%, P < 0.001). The prevalence of moderate and severe dysplasia was significantly increased with 26.7% in TP compared with CP (21.9% versus 17.2%, P < 0.01). In conclusion, the ThinPrep Pap Test yielded a significant decrease in atypia rates compared with the conventional Papanicolaou test. In subsequent follow-up the percentage of neoplastic lesions was significantly increased in the ThinPrep Pap Test samples.     

Morphologic scales and relative distances between lesions identified on Papanicolaou smears.

We fit a Gaussian-type curve to a nonmonotonic transform of initial arbitrarily given scales (pseudoscalar transform) of a two-way discrete classification table by maximizing the likelihood (entropy) and computing morphologic scales of clusters of objects identified by visual judgments. The scales give the relative distance between pairs of categories or states. Morphologic scales of histologic lesions (states) identified on Papanicolaou smears were computed from a confusion matrix consisting of 3,545 matched pairs of observations on Papanicolaou smears and colposcopically directed biopsies available from the Gynecologic Cytology Laboratory, University of Minnesota, between 1985 and 1987. The original (uncollapsed) confusion matrix consisted of eight cytologic categories and histologic states: normal plus atypical benign, reactive atypia (including condylomatous changes); mild, moderate and severe dysplasia; squamous carcinoma in situ; invasive squamous cell carcinoma; and other malignancies. The morphologic scales are expressed numerically, which reflects a degree of confusion between two diagnostic categories or states. Except for malignancies other than squamous cell carcinoma, morphologic scales of histologic states seen from cytology retained the order of clinical severity. We detected a high degree of confusion between severe dysplasia and carcinoma in situ. Malignancies other than squamous cell carcinoma fell between moderate and severe dysplasia. Such a transposition of the scales in this group containing adenocarcinoma was likely due to frequent association of adenocarcinoma with cervical intraepithelial neoplasia. Morphologic scales of cytologic categories seen from histology showed high degrees of confusion and transpositions between mild and moderate dysplasia and between severe dysplasia and carcinoma in situ.(ABSTRACT TRUNCATED AT 250 WORDS)     

Papanicolaou staining of air-dried smears: value in rapid diagnosis

Air-dried material normally submitted for Diff-Quik (modified Romanowsky stain) was rehydrated in normal saline, then fixed for a short period in formol alcohol, before staining by a modified Papanicolaou technique. Staining was performed by a rapid manual technique (<2 min) if urgent or routinely on an automatic stainer. Comparisons were made between wet-fixed Papanicolaou-stained specimens and air-dried Papanicolaou-stained material. Air-dried material stained after rehydration showed superior nuclear definition compared with wet-fixed material; the removal of erythrocytes enhanced the staining of the remaining epithelial cells.     

Electron microscopy of cells showing viral cytopathic effects in Papanicolaou smears

A technique is described whereby cells showing cytologic changes suggestive of virus infection by light microscopy can be processed further for examination in the electron microscope so that virus particles present in the cell can be visualized directly. We present the results of electron microscopy of over 100 Papanicolaou-stained smears processed this way. The morphologic changes in the cells and the ultrastructural appearances of the virus particles are demonstrated. This technique is particularly valuable for retrospective studies of mounted cytologic or histologic material. It has also proven to be a valuable research tool in the study of human polyomavirus and human wart virus infection.     

Quantitative cytology of the positive region in flow sorted vaginal smears.

Fifty-one gynecologic specimens were collected from three women's hospitals and mailed in a prefixed status to our laboratory. The specimens were classified into a negative, a suspicious, a postradiation, and a positive group. After single cell dispersion the samples were stained for DNA and protein, analyzed, and sorted in the dual laser equipped Heidelberg flow analyzer sorter (HEIFAS). Particles with elevated DNA values (beyond 3.5 ploidy) and with intermediate protein values were sorted as the positive fraction directly on microscopic slides. After restaining according to Papanicolaou, they were re-evaluated cytologically and identified as tumor cells, dysplastic cells and false alarms. The latter consist of doublets and aggregates of more than two cells, binucleated cells, sperm aggregates and epithelial cells contaminated with bacteria. The different groups showed significant differences regarding the total rate of aggregates to single cells. In general, false alarms were very frequent in the positive region and impeded the statistical classification of the sample. The reduction of false alarms is a prerequisite for prescreening with flow instrumentation.     

Detection of malignant melanoma of the uterine cervix from Papanicolaou smears. A case report

A case of malignant amelanotic melanoma of the uterine cervix in a patient presenting with right hemiparesis and enlarged lymph nodes was diagnosed in Papanicolaou-stained cervical smears showing many melanoma cells. Melanoma cells with bizzare nuclear and cytoplasmic abnormalities, rarely seen in other tumors, helped to establish a positive diagnosis. The diagnosis was confirmed by histopathologic study of the endocervical surgical specimen, including a positive immunoperoxidase staining for S-100 protein.     

A multi-spectral approach for scene analysis of cervical cytology smears.

A multi-spectral approach for the scene analysis of cervical cytology smears, using multiple images of a scene digitized through suitably chosen color filters matched to the Papanicolaou stain, has been proposed here. This technique involves clustering of two-dimensional data for extracting cytoplasm of the epithelial cells. Its performance on an experimental data set of 233 scenes involving more than 10 types of normal and malignant epithelial cells has been compared with density and gradient thresholding techniques. This resulted in an approximate 83% rate of success compared to approximately 40% for the rest of the other techniques.