Prediction of cervical neoplasia diagnosis groups. Discriminant analysis on digitized cell images

The purpose of this study was to develop discriminant analysis models for predicting cervical dysplasia/neoplasia case diagnoses using cytometric features derived from the digital image analysis of cell monolayers. The data base consisted of 925 cells from 27 cases diagnosed either as moderate dysplasia (n = 10), severe dysplasia (n = 5), carcinoma in situ (n = 8) or invasive carcinoma (n = 4) on both tissue biopsy and monolayer preparations. Cell features examined were cell diameter, nuclear diameter, nuclear mean optical density (OD), nuclear integrated OD (IOD), nuclear OD standard deviation, normalized IOD, nuclear texture and nuclear-cytoplasmic ratio. Features derived from cells visually classified as moderate dysplasia correctly predicted the case diagnosis of moderate dysplasia versus more severe disease for 85% of the cells. Prediction models using summary measures (mean and variance) derived from all visually classified abnormal cells within each case correctly separated all cases into their respective diagnostic categories. These findings suggest that dysplastic cells in a cytologic sample have features that collectively reflect the tissue diagnosis, regardless of the visual differences among the cells. Such information has potential use for diagnosis and possibly for prognosis.     

Morphologic characteristics of p16INK4a-positive cells in cervical cytology samples

OBJECTIVE: Overexpression of p16INK4a has been proposed as a biomarker helpful for the identification of dysplastic cervical epithelial cells on histologic slides as well as in cervical smears. Since a few nontransformed cells in the genital tract in some instances may also express p16INK4a, we evaluated whether applying established morphologic criteria for cervical dysplasia allows a distinction of dysplastic from nondysplastic p16INK4a-stained cells in cytologic samples. STUDY DESIGN: Liquid-based cytology samples were obtained from a screening population (n=50), and from patients attending a dysplasia clinic (n=40). Slides prepared from these samples were stained with the conventional Papanicolaou stain procedure. From each specimen, a second slide was prepared in parallel and immunostained for p16INK4a. Cytologic diagnoses for most patients attending the dysplasia clinic could be compared to the reported histologic diagnoses on punch biopsy samples taken from the patients at the time of colposcopy. This allowed a comparison of the cytology and p16INK4a immunostaining results with subsequent hematoxylin and eosin-based histologic diagnoses. RESULTS: Overall, in 10% of slides obtained from patients with nonsuspicious smears, few p16INK4a-positive cells were found. Using established morphologic criteria and applying these criteria on cells showing any p16INK4a immunoreactivity, p16INK4a-positive normal or metaplastic cells could be discriminated from p16INK4a-expressing dysplastic cells. In 21 of 22 cases (95%) of high grade lesions (cervical intraepithelial neoplasia 2 or higher in follow-up histology), easily recognizable p16INK4a-positive dysplastic cells could be detected, with the remaining case lacking dysplastic cells in the thin-layer slide used for p16INK4a immunostaining. CONCLUSION: Established morphologic criteria for cervical dysplasia can be readily applied to p16INK4a-immunostained cytologic specimens. Thus, p16INK4a immunostaining may help to avoid ambiguities in the interpretation of cervical cytology samples and facilitate more rapid diagnosis and possibly even automated screening of cytologic slides.     

Canonical analysis of cells in normal and abnormal cervical smears

Over 4,000 cells from 105 normal and 96 abnormal uterine cervical scrapes were prepared according to the UCLA monolayer procedure, stained by a routine Papanicolaou method and visually classified by two cytopathologists and a technologist into seven classes: parabasal, metaplastic, mild dysplasia, moderate dysplasia, severe dysplasia, carcinoma in situ and invasive carcinoma. Canonical analysis was used to correlate effects-coded class membership variables with 23 cell features derived from digital image analysis. In general, nuclear texture measures derived from linear combinations of run-length correlations along with features derived from a Markov transitional probability matrix provided the best predictors of cell class. After cells were divided into benign (moderate dysplasia or less) and malignant (severe dysplasia or worse) groups, discriminant analysis correctly classified 84% of the benign cells and 91% of the malignant cells.     

Predictive value of DNA cytometry in CIN 1 and 2. Image analysis of 193 cases

OBJECTIVE: To investigate DNA image cytometry for predicting the prognosis of cervical intraepithelial neoplasia (CIN). STUDY DESIGN: Smears from 151 women affected by CIN 1 or 2 on cytology with minimal follow-up of three years were included. Sixty-seven showed progression, with histologically confirmed carcinoma in situ or invasive cancer. Eighty-four cases showed regression of the disease, which was cytologically, histologically and colposcopically confirmed. Papanicolaou-stained smears were destained, and the Feulgen reaction was performed with consecutive image DNA cytometry of suspicious cells using an image analysis system (Cires, Zeiss, Germany). The DNA index of the greatest stemline and the number of single aneuploid cells, using 9c exceeding events, were computed. RESULTS: In the group with progression, an aneuploid DNA stemline was found in 25 smears (26.9%). In 64 cases (66.7%) more than one aneuploid event was detected. The total number of aneuploid cases in this group was 76 (81%). In the group without progression, the number of aneuploid stemlines was 2 (2%). Single aneuploid cells could be found in five cases (5%). The overall number of aneuploid cases in that group was five. The sensitivity was 74.3%, positive predictive value 85.2% and negative predictive value 77%. CONCLUSION: Aneuploidy is a marker for prospective malignancy in cervical Papanicolaou smears. DNA image cytometry, as an additional method, can be used to predict outcome in patients with CIN 1 and 2 of the cervix. DNA cytometry is not a screening method but can add further information for a treatment decision in doubtful cases.     

Retrospective DNA analyses in cervical dysplasia as related to neoplastic progression or regression

Intracellular DNA distribution was measured in cells from two groups of patients with moderate cervical dysplasia. One group consisted of patients who subsequently developed carcinoma in situ; the other consisted of patients whose lesions regressed to normality. Papanicolaou-stained slides were examined cytologically, and dysplastic cells were located. The slides were then destained and restained by means of the Feulgen DNA staining method, after which they were analyzed in a microspectrophotometer. The DNA distribution pattern of both groups was different from that of normal cells and exhibited the same characteristics observed earlier in premalignant cervical cellular atypias. There was no significant difference between the two groups. The results indicate that quantitative DNA determinations in cytomorphologically equivalent dysplastic cervical cells do not offer additional means of predicting the outcome of the lesions.     

Measurement of subvisual changes in cervical squamous metaplastic cells for detecting abnormality.

Quantitative measures of visually normal squamous metaplastic cells exfoliated from the uterine cervix were obtained to test the hypothesis that these cells, like intermediate squamous and endocervical columnar cells, show subvisual evidence of atypia in cases of bonafide squamous intraepithelial neoplasia. The cells identified as squamous metaplastic were obtained from 14 abnormal (dysplastic) and 9 diagnostically negative cases. Although the cell populations so grouped showed no statistically significant differences in overall cell size, nuclear area or nuclear/cytoplasmic ratios, there were significant differences in nuclear and cytoplasmic densitometric features and in nuclear texture features. A combination of three features (nuclear density, texture and cytoplasmic density) permitted 76% of the cells to be categorized correctly as originating in normal versus abnormal slides. It is concluded that selected quantitative features of exfoliated metaplastic cell populations may contribute to improved diagnostic accuracy in automated screening for cervical abnormalities.     

Use of amaranthus leucocarpus lectin to differentiate cervical dysplasia (CIN)

Alterations in O-glycosylation of proteins in cell surfaces can originate disorder in cellular function, as well as in cell transformation and tumoral differentiation. In this work, we investigate changes in O-glycosylation in cervical intraepithelial dysplasia (CIN) at different stages of differentiation (CIN I, CIN II, and CIN III) using lectins specific for O-glycosidically linked glycans. Twenty cases with CIN I, CIN II, and CIN III dysplasias each, and 20 normal cases were studied by lectin histochemistry and evaluated under optical microscopy. The lectins from Glycine max and Griffonia simplicifolia showed no differences in their recognition pattern among the different CIN stages and normal tissue. Dolichos Biflorus lectin recognized CIN I dysplasia. Lectin from Amaranthus leucocarpus showed increased reactivity in the presence of CIN II dysplasia, compared with CIN I and CIN III. These results suggest that subtle modifications in the O-glycosylation pattern could be considered in diagnosis or prognosis of cervical precancerous stages.     

Immunocytochemical detection of p16INK4a protein in scraped cervical cells

OBJECTIVE: To develop an immunocytochemical technique for p16INK4a protein detection in scraped cervical cells for cancer screening. STUDY DESIGN: We took duplicate cervical scrapes from each participant, the first for a Pap smear and the second for p16INK4a protein detection. From a 50-microL cell suspension prepared from the scrape rinsing, a 10-microL aliquot was dropped in a 5-mm-diameter circle on a glass slide, air dried and fixed in 0.1% formal saline (1 hour) and in 95% ethanol (10 minutes). Using the immunocytochemical technique, slides from 30 samples of each Pap diagnosis class were stained sequentially with mouse monoclonal anti-p16INK4a (primary antibody), biotinylated goat antimouse IgG (secondary antibody), horse-radish peroxidase-labelled streptavidin and 3,3'-diaminobenzidine and mixed hydrogen peroxide, then counterstained with hematoxylin. A positive sample had to contain > or = 3 immunoreactive cells. Results were confirmed by western blot analysis of lysates from the remaining 40 microL of each cervical cell suspension. RESULTS: Samples were grouped as control (normal cervical cells), mild dysplasia (ASCUS, LSIL) and high abnormality (HSIL, SCC). Using the immunocytochemical technique, > 95% of the positive (SiHa cells) but 0% of the negative controls (human embryonic lung fibroblast cells) showed immunoreactive cells. All slides displayed a clear background without mucus, and positive cells were stained in both the cytoplasm and nucleus. p16INK4a Protein was detected in 17 of 30 (56.67%) ASCUS and 10 of 30 (33.33%) LSIL and increased with the degree of abnormality to 93.33% (28 of 30) and 96.67% (29 of 30) in the HSIL and SCC group, respectively. Normal cervical cells and degenerated malignant cells were nonimmunoreactive. Western blot analysis confirmed similar positive samples in the low-abnormality group, while the whole high-abnormality group was immunoreactive. A sampling error might have caused the 2 HSIL and 1 SCC sample to be negative using our immunocytochemical technique. CONCLUSION: p16INK4a Protein detection in scraped cervical cells using the immunocytochemical technique correlated with western blot analysis and was nontraumatic and precise. It offers a significant diagnostic adjunct to the Pap test for cervical cancer screening.     

The use of an automated image cytometer for screening and quantitative assessment of cervical lesions in the British Columbia Cervical Smear Screening Programme

The use of an automated image cytometer for screeing and quantitative assessment of cervical lesions in the British Columbia Cervical Smear Screening Programme
The development of an automated device to screen cervical cytology slides for the detection of pre-invasive lesions of the cervix has been the goal of many individuals for over 30 years. The increasing sophistication of the technology of automation and increasingly powerful computer technology have enabled a number of these systems to reach the stage at which they have become a practical reality. The Department of Cancer Imaging at the British Columbia Cancer Agency has developed such a device over the past few years. This study reports the preliminary results of a trial to determine the reliability of the device for the screening and quantitative assessment of cervical cells. A training set of over 1000 cervical slides was used to train the image cytometer. A test set of 1030 slides was screened by the image cytometer and in the Cytology Screening Laboratory. At the 50% sample split the sensitivity of the image cytometer was 95% for severe dysplasia and 90% for moderate dysplasia, compared with a sensitivity of 90% for both of these lesions using conventional screening. A combination of nuclear texture features was found which can be used for the quantitative assessment of both abnormal cells and apparently normal intermediate cells.     

Cervical carcinoma progression-associated genetic alterations on chromosome 6

To identify the loci associated with progression of cervical carcinoma, chromosome 6 regions were tested for loss of heterozygosity. Detailed analysis with 28 microsatellite markers revealed a high frequency of allelic deletions for several loci of the short (6p25, 6p22, 6p21.3) and long (6q14, 6q16-21, 6q23-24, 6q25, 6q27) arms of chromosome 6. Examination of 37 microdissected carcinoma and 22 cervical dysplasia specimens revealed allelic deletions from the HLA class I-III genes (6p22-21.3) and subtelomeric locus 6p25 were found in more than 40% dysplasia specimens. With multiple microdissection of cryosections, genetic heterogeneity of squamous cervical carcinoma was analyzed, and clonal and subclonal allelic deletions from chromosome 6 were identified. Half of the tumors had clonal allelic deletion of D6S273 (6p21.3), which is in a Ly6G6D (MEGT1) intron in the HLA class III gene locus. The frequency of allelic deletions from the chromosome 6 long arm was no more than 20% in dysplasias. Allelic deletions from two loci, 6q14 and 6q16-21, were for the first time associated with invasion and metastasis in cervical carcinoma.     

细胞核内特征值的改变可引起DNA倍体的变化

目的通过测定不同DNA倍体细胞,研究细胞核内特征值的改变。方法用宫颈刷刷出宫颈细胞,经固定后,用涂片离心机制成二张玻片,一张行巴氏染色作TBS诊断,另一张行Feulgen染色做DNA定量测定。通过对宫颈细胞核图像内像素的统计,计算出细胞核内多种特征值,比较不同DNA倍体细胞内特征值的不同。结果 161873例妇女行宫颈细胞学检查,常规细胞学检查发现2454例低级别鳞状上皮内病变(low-grade squamous intraepithelial lesion,LSIL)和523例高级别鳞状上皮内病变(high-grade squamous intraepithelial lesion,HSIL);而DNA倍体分析发现3412例有3个以上>5c细胞。84%以上的LSIL和HSIL病例均可见倍体异常细胞。与2c细胞相比,4c、5c、7c及9c细胞核面积及核半径明显增大;7c、9c细胞核内平均光学密度和紧实度均值也有明显改变,而光密度方差和灰度熵无变化。结论宫颈细胞DNA倍体改变往往伴有细胞形态和DNA核内分布等特征值的改变。     

Cytomorphometric differences among individual "moderate dysplasia" cells derived from cervical intraepithelial neoplasia

Individual abnormal cervical epithelial cells can be categorized either by the lesion from which they are derived or by their unique cytologic characteristics. In building a data base for image analysis of cervical epithelial cells, categories of cells were defined according to distinct cytomorphologic characteristics, without knowledge of the lesion of origin. The three individual scorers, two cytopathologists and one senior cytotechnologist, most frequently agreed upon the cells classified as "moderate dysplasia." The measurements of digitized cells in this category had the smallest confidence intervals of any of the abnormal cell categories. For these two reasons, as well as the ubiquitous nature of "moderate dysplasia" cells in smears obtained from all patients with cervical epithelial neoplasia, cells in this category were studied in greater detail. Significant differences were noted in cell measurements among cells in this class when the cells came from patients with different grades of cervical neoplasia. The findings indicate that visually similar "moderate dysplasia" cells can be separated by digitized measurements into clusters dependent upon the parent lesion. The biologic implications are not yet clear, but such findings suggest that each disease is perhaps a committed lesion from inception. Therefore, predictability of ultimate outcome could be based on image analysis of cells derived from early cervical lesions, which would allow therapy to be performed on a more logical basis.     

P-9A COMPARISON OF THE REPORTING MORPHOLOGICAL FEATURES OF GLANDULAR NEOPLASIA FOR CONVENTIONAL AND LIQUID BASED CYTOLOGY FOR CERVICAL SAMPLES

Cervical cancer accounts for approximately 15% of cancer diagnosed in women worldwide with up to 190 000 deaths per annum. One of the major causes of cervical cancer is the infection of human papillomavirus (HPV), a DNA virus. This virus is epidermotropic; there are over 75 subtypes and subtypes 16, 18, 31 and 33 are associated with cervical intraepithelial neoplasia (CIN) and carcinomas. Since the start of the cervical screening in mid 1960s, the cervical cancer rate has decreased. There are two techniques used for slide preparation and staining: conventional cytology and liquid based cytology (LBC). Due to the differences in sample collection and preparation, certain aspects of cell morphology, architecture and patterns will present differently from each other on the slide. The study was conducted in a County Hospital. Twenty conventional slides and eight LBC slides already reported as ? Glandular neoplasia were reviewed and assessed with regards to their morphological features. Moreover, conventional slides were compared with LBC slides to determine the differences in their cell morphology, sensitivity and specificity. Furthermore, a semi-quantitative method was used and also true-positive and false-positive rates were evaluated using positive predictive value (PPV). The findings indicated that despite the differences in cell morphology there are many similarities between the two techniques. The study also showed that it was difficult to distinguish between abnormal glandular cells and abnormal squamous cells, which may end in a false positive result and over reporting of glandular neoplasia. Finally, it showed that LBC slides were easier to screen and also had a higher positive predictive value (PPV) resulting in higher sensitivity and specificity. In conclusion, the LBC technique is more accurate and conversion to this technique is the positive step in the screening program.     

Cytometric evidence that cervical intraepithelial neoplasia I and II are dysplasias rather than true neoplasias. An image analysis study of factors involved in the progression of cervical lesions.

Image analysis was performed on 40 Feulgen-stained histologic samples and 48 Feulgen-stained cytologic preparations representing normal squamous epithelium and all grades of cervical lesions (from mild dysplasia to invasive carcinoma) in order to characterize the evolutionary progressive changes in cervical epithelial proliferative disease toward malignancy. Quantitative studies included the analysis of proliferative features, differentiation features, nuclear morphology and DNA content. The data obtained on the histologic sections showed that the various features, to a different extent, detected a gradual increase in phenotypic cellular disarrangements related to the progression of the cervical lesions toward malignancy--that is, the modifications to nuclear area, perimeter, DNA content, percentage of nuclei with nucleoli, nuclear/cytoplasmic ratio and percentage of cells with no membrane positivity for soybean agglutinin lectin were progressively greater, moving from normal epithelium and mild dysplasia toward infiltrating carcinoma. In particular, all the morphologic and histochemical features appeared to parallel a diploid reduction and the appearance of aneuploidy. The simultaneous evaluation of proliferation- and differentiation-related features, together with those of nuclear DNA content, showed two main successive preneoplastic lesions: one characterized by an increase in cell turnover without alterations in its organization and another by a true neoplastic disorder. The data obtained on sequential cytologic examinations showed that individual cell changes are detectable and seem basically to be characterized by the appearance of clusters of cells with somatic characteristics not observed in previous cytologic checks. From the results of our study, the cervical intraepithelial neoplasia (CIN) concept appears to be inaccurate. In fact, only CIN III (severe dysplasia/carcinoma in situ) lesions have the morphologic and proliferative alterations of true neoplasia. In contrast, CIN I and some cases of CIN II lesions lack these characteristics and seem to be properly classified as dysplasia, thus avoiding the term neoplasia, implicit in CIN. Moreover, the multivariate study of data sets of features related to the progressive somatic changes, both in histologically and cytologically studied cases, allows us to detect the steps of progression; they are marked by the appearance of cell clusters with qualitatively different phenotypic characters when compared to the cell populations from which they presumably arise. These results seem to provide a further argument against the CIN theory, which stresses the concept that progression is related only to a gradual numerical increase in an initially established phenotype with the characteristics of malignancy.     

Clinical evaluation of the EA-rosette test in the early detection of cervical cancer

It was previously found that a negative EA-rosette test, showing EA-rosette-forming cells in a cervical cell suspension, excluded the presence of cells of invasive carcinoma (predictive value of 99.9%). This study on 2,462 patients confirmed the applicability of the EA-rosette test in screening for precancerous as well as cancerous lesions. In 98.6% of the cases of dysplasia, carcinoma in situ and invasive carcinoma, the cervical cell suspensions contained EA-rosette forming cells (the rosette test was positive). With a negative EA-rosette test, the probability of missing a specimen with class III cytology (mild/moderate dysplasia) was 1.4%, of missing one with class IV cytology (severe dysplasia/carcinoma in situ) was 0.8% and of missing one with class V cytology (invasive carcinoma) was 0.25%. The predictive value of a negative EA-rosette test was 98.6%. The false-negative rate for negative EA-rosette tests was 3.7% for invasive carcinoma, 17.5% for carcinoma in situ and severe dysplasia and 41.4% for mild to moderate dysplasia.     

Natural history of precancerous and early cancerous lesions of the uterine cervix

A prospective study of cervical dysplasia cases and control cases matched for age and parity was undertaken in search of factors related to cervical carcinogenesis. Cytologic examination of 66,736 women revealed negative findings in 28.5%, inflammation in 70.3%, dysplasia in 1.4% and carcinoma in 0.1% of the cases. Data on epidemiologic features, cytomorphologic characteristics and serologic findings of antibodies to herpes simplex virus (HSV) were collected for proven cancer patients, dysplasia cases and control subjects. Cancer patients revealed significantly elevated antibodies to HSV as compared to the controls. The analysis revealed a higher proportion of dysplasia cases with an age at consummation of marriage of less than or equal to 15 years as compared to controls, with a relative risk of 1.5 (P less than .05). Similarly, a higher proportion of women with dysplasia had HSV-II-specific antibodies as compared to control women. The relative risk was found to be 1.3, which was not statistically significant (P greater than .05). The Mantel-Haenszel summary relative risk between antibodies to HSV and the two groups (dysplasia cases and controls), adjusted for the age at consummation of marriage, worked out to be 1.38, which was also statistically not significant (P greater than .05). The overall progression rate of dysplasia to malignancy was found to be 11.7% at the end of 54 months (during a total follow-up period of 84 months). Progression to cancer was highest in severe dysplasia cases and less in mild dysplasia cases. The progression rates were also significantly higher in the group of women who revealed antibodies to HSV II. Similar differences in the progression rates were observed with regard to the age at consummation of marriage.     

Prediction of upper aerodigestive tract cancer by slide-based cytometry.

AIM: To evaluate slide-based cytometry in screening for and following up of carcinoma of the upper aerodigestive tract using swabs for a minimal-invasive approach. METHODS: Laser scanning cytometry (LSC) was used for multiparametric analysis of cells stained for cytokeratin and DNA to determine the DNA-index (DI) of the tumor cells. Histograms with 0.95 < DI < 1.05 and 1.9 < DI < 2.1 were defined as DNA euploid and any other DI as DNA aneuploid. After subsequent HE-staining, single cells were relocalized in order to document morphology. Conventional cytology was also performed on a subset of the slides. Routine histopathology of parallel biopsies served as gold standard in all cases. RESULTS: 115 swabs from 109 patients were obtained from the entire upper aerodigestive tract. 16 swabs were classified as insufficient for LSC. In the remaining 99 specimens, 1 benign lesion was misclassified as malignant, while 61 of the 75 malignant lesions were correctly identified. This corresponds to predictive values of 98.4% and 62.2% for the detection of malignant and benign samples by LSC. CONCLUSION: This pilot study demonstrates the validity of LSC screening for the identification of tumor malignancy in the upper aerodigestive tract from swab collected cytological material.     

Tissue counter analysis of histologic sections of melanoma: influence of mask size and shape, feature selection, statistical methods and tissue preparation.

BACKGROUND: Tissue counter analysis is an image analysis tool designed for the detection of structures in complex images at the macroscopic or microscopic scale. As a basic principle, small square or circular measuring masks are randomly placed across the image and image analysis parameters are obtained for each mask. Based on learning sets, statistical classification procedures are generated which facilitate an automated classification of new data sets. OBJECTIVE: To evaluate the influence of the size and shape of the measuring masks as well as the importance of feature selection, statistical procedures and technical preparation of slides on the performance of tissue counter analysis in microscopic images. As main quality measure of the final classification procedure, the percentage of elements that were correctly classified was used. STUDY DESIGN: HE-stained slides of 25 primary cutaneous melanomas were evaluated by tissue counter analysis for the recognition of melanoma elements (section area occupied by tumour cells) in contrast to other tissue elements and background elements. Circular and square measuring masks, various subsets of image analysis features and classification and regression trees compared with linear discriminant analysis as statistical alternatives were used. The percentage of elements that were correctly classified by the various classification procedures was assessed. In order to evaluate the applicability to slides obtained from different laboratories, the best procedure was automatically applied in a test set of another 50 cases of primary melanoma derived from the same laboratory as the learning set and two test sets of 20 cases each derived from two different laboratories, and the measurements of melanoma area in these cases were compared with conventional assessment of vertical tumour thickness. RESULTS: Square measuring masks were slightly superior to circular masks, and larger masks (64 or 128 pixels in diameter) were superior to smaller masks (8 to 32 pixels in diameter). As far as the subsets of image analysis features were concerned, colour features were superior to densitometric and Haralick texture features. Statistical moments of the grey level distribution were of least significance. CART (classification and regression tree) analysis turned out to be superior to linear discriminant analysis. In the best setting, 95% of melanoma tissue elements were correctly recognized. Automated measurement of melanoma area in the independent test sets yielded a correlation of r=0.846 with vertical tumour thickness (p<0.001), similar to the relationship reported for manual measurements. The test sets obtained from different laboratories yielded comparable results. CONCLUSIONS: Large, square measuring masks, colour features and CART analysis provide a useful setting for the automated measurement of melanoma tissue in tissue counter analysis, which can also be used for slides derived from different laboratories.     

Contextual analysis and intermediate cell markers enhance high-resolution cell image analysis for automated cervical smear diagnosis.

Until now, efforts to automate cervical smear diagnosis have focused on analyzing features of individual cells. In a complex specimen such as that obtained from a cervical scrape, diagnostically significant cells may not be adequately represented or may elude detection by the automated technology. An approach is needed that extracts additional quantitative information from cervical smears beyond what the cell-by-cell approach can provide. A new methodology, contextual analysis, was developed to extract global quantitative information about cells, cell clusters, and background debris. This pilot study was designed to compare the efficacy of contextual analysis with high-resolution, single cell analysis and the analysis of intermediate cell markers. Thirty-four samples prepared as monolayers and stained with the Feulgen-Thionin/Congo Red stain were measured. Contextual analysis alone was able to classify 91% of the smears correctly; single cell analysis classified 94% of the cells correctly; and the intermediate cell analysis correctly identified the smear diagnosis for 84% of the cells. When all three analysis methods were combined into a simple smear level classifier, the overall smear classification accuracy was improved over those obtained using the three methodologies alone.