Biomedical subjects

M M Pahlplatz

Publications and source records attributed to M M Pahlplatz.

At least 19 recordsLinked to original sources

PubMed2001

Improved prediction of metastasis in tongue carcinomas, combining vascular and nuclear tumor parameters.

BACKGROUND: Predicting the presence of metastasis, based on tumor or tumor-related characteristics is of utmost importance. The authors studied the significance of tumor DNA features and tumor-related angiogenesis to predict the occurrence of metastasis in squamous cell carcinomas (SCCs) of the tongue. METHODS: Paraplast blocks from resection specimens of 20 metastasized and 20 nonmetastasized SCCs of the tongue with a minimum follow-up of 24 months were used. Tissue sections were stained with anti-CD34 monoclonal antibodies for vessel visualization, and according to Feulgen to stain DNA. Using image analysis, data from both stainings were computed for each of the 40 carcinomas. A logistic regression model to predict the presence of metastasis, based on vascular and nuclear morphology features, was developed. RESULTS: The intratumor variation of chromatin condensation and the percentage vessels smaller than 5 microm in diameter were selected for the model. The model correctly predicted metastasis in 90% of patients and excluded metastasis correctly in 75% of nonmetastasized tumors. Taking into account the prevalence of metastasis in SCC of the tongue of between 30% and 60%, this means a predictive value for a negative outcome of between 95% and 83%. CONCLUSIONS: The proposed model shows an improvement of predictive values compared with previous models with single parameters. Therefore, a multiparameter model appears to predict the multiparameter process of metastasis better.

Carcinoma, Squamous Cell↗

PubMed2000

Hue-saturation-density (HSD) model for stain recognition in digital images from transmitted light microscopy.

BACKGROUND: Transmitted light microscopy is used in pathology to examine stained tissues. Digital image analysis is gaining importance as a means to quantify alterations in tissues. A prerequisite for accurate and reproducible quantification is the possibility to recognise stains in a standardised manner, independently of variations in the staining density. METHODS: The usefulness of three colour models was studied using data from computer simulations and experimental data from an immuno-doublestained tissue section. Direct use of the three intensities obtained by a colour camera results in the red-green-blue (RGB) model. By decoupling the intensity from the RGB data, the hue-saturation-intensity (HSI) model is obtained. However, the major part of the variation in perceived intensities in transmitted light microscopy is caused by variations in staining density. Therefore, the hue-saturation-density (HSD) transform was defined as the RGB to HSI transform, applied to optical density values rather than intensities for the individual RGB channels. RESULTS: In the RGB model, the mixture of chromatic and intensity information hampers standardisation of stain recognition. In the HSI model, mixtures of stains that could be distinguished from other stains in the RGB model could not be separated. The HSD model enabled all possible distinctions in a two-dimensional, standardised data space. CONCLUSIONS: In the RGB model, standardised recognition is only possible by using complex and time-consuming algorithms. The HSI model is not suitable for stain recognition in transmitted light microscopy. The newly derived HSD model was found superior to the existing models for this purpose.

Blood Vessels↗

PubMed1998

An improved procedure to quantify tumour vascularity using true colour image analysis. Comparison with the manual hot-spot procedure in a human melanoma xenograft model.

In a number of recent papers, the degree of tumour vascularization has been described as a promising new prognostic factor. Methods for the assessment of vascular density involve immunohistochemical staining of the vasculature, followed by counting the number of vessel profiles in the angiogenic hot spot. One of the problems of this procedure is the selection of the angiogenic hot spot, which has been described as being subject to inter-observer variation. In this study, the value of true colour image analysis in reducing inter-observer variation has been assessed. Highly (MV3) and poorly (M14) vascularized human melanoma xenografts were used to evaluate the image analysis procedure, and the image analysis results were compared with results from the conventional manual hot-spot procedure. Assessment by image analysis was performed on measurement fields covering the entire tumour tissue specimens rather than on a single hot-spot field. Also, by selecting the most densely vascularized area from all fields assessed by the semi-automatic procedure, it was possible to objectify the hot spot selection (automated hot-spot procedure). Manual assessment showed a good correlation between two independent observers for MV3 xenografts (r = 0.74, P = 0.014), but a poor correlation for M14 xenographs (r = 0.4, P > 0.05). Automated assessment by different operators showed good correlations for both MV3 xenografts (r = 0.99, P < 0.001) and M14 xenografts (r = 0.80, P = 0.006). It is concluded that although both manual vessel counting and semi-automated image analysis can differentiate between the level of vascularization in the two types of xenograft (P < 0.001 for both methods), the automated method is favourable in that it showed no significant inter-observer effects. In M14 xenografts, the manual hot-spot vessel densities did not correlate well with the automated hot-spot densities (r = 0.27, P > 0.05), indicating that selection of angiogenic hot spots in this tumour type is indeed subject to observer bias. The automated hot-spot vessel densities were a reliable indicator of overall tumour vessel density in both tumour types. Image analysis allows analysis of vessel subclasses based on morphological criteria such as vessel profile area or diameter. In the model system used, the discrimination between MV3 and M14 xenografts was further enhanced by selectively examining vessels with diameters between 6 and 9 microns (P < 0.0005). In conclusion, image analysis appears to offer an objective and more reproducible method to quantify tumour vascularity than manual counting of vessel profiles in the hot spot. Analysis of subclasses of vessels may further enhance the value of vessel density measurements in discriminating between tumour types differing in biological behaviour.

Animals↗

PubMed1998

An image analysis study on nuclear morphology in metastasized and non-metastasized squamous cell carcinomas of the tongue.

In a retrospective case-control study on 46 metastasized and 34 non-metastasized primary tongue carcinomas, the nuclear morphology and chromatin pattern were assessed in 3 microns thick, formalin-fixed, paraffin-embedded, and Feulgen-stained tissue sections of surgical resection specimens, by means of high-resolution computer-assisted image analysis. The aim of this study was to disclose differences in karyometric features, such as nuclear size-, shape-, and chromatin-pattern features, between these groups, with a view to developing a discriminant function that can predict the occurrence of metastasis for the individual patient. In addition, the lymph node metastases of 31 patients and the normal tongue epithelium of 21 patients were also assessed, to study the possible differences between these two groups and primary tumours. In the metastasized tumours, the chromatin was significantly more condensed (P = 0.01) and exhibited significantly less variation in chromatin condensation (P < 0.001) than in the group of non-metastasized carcinomas. Comparison of lymph node metastases with their primary tumours disclosed only minor differences in chromatin pattern. These findings suggest that only minor genetic differences exist between primary tongue carcinomas and their metastases. Tumour cells of tongue carcinomas showed highly significant differences from cells of normal tongue mucosa for most karyometric features. Logistic regression analysis resulted in a classifier, based on the circularity of the nucleus (CIRC) and the standard deviation of the chromatin condensation (SD COND), to predict the occurrence of lymph node metastases. After cross-validation, the percentages of correct classifications in the group of metastasized and non-metastasized tumours were 72 and 62 per cent, respectively. These results are comparable to the classification results obtained from a classifier based on the clinical T-stage, but our karyometric classification results show a much more equal distribution between the sensitivity and specificity. Karyometric features appeared to be more appropriate to predict metastases than biomarkers such as p53, bcl-2, and Ki-67.

Carcinoma, Squamous Cell↗

PubMed1998

DNA-cytometry of progressive and regressive cervical intraepithelial neoplasia.

A retrospective analysis was performed on archival cervical smears from a group of 56 women with cervical intraepithelial neoplasia (CIN), who had received follow-up by cytology only. Automated image cytometry of Feulgen-stained DNA was used to determine the differences between progressive and regressive lesions. The first group of 30 smears was from women who had developed cancer after initial smears with dysplastic changes (progressive group). The second group of 26 smears with dysplastic changes had shown regression to normal (regressive group). The goal of the study was to determine if differences in cytometric features existed between the progressive and regressive groups. CIN categories I, II and III were represented in both groups, and measurements were pooled across diagnostic categories. Images of up to 700 intermediate cells were obtained from each slide, and cells were scanned exhaustively for the detection of diagnostic cells. Discriminant function analysis was performed for both intermediate and diagnostic cells. The most significant differences between the groups were found for diagnostic cells, with a cell classification accuracy of 82%. Intermediate cells could be classified with 60% accuracy. Cytometric features which afforded the best discrimination were characteristic of the chromatin organization in diagnostic cells (nuclear texture). Slide classification was performed by thresholding the number of cells which exhibited progression associated changes (PAC) in chromatin configuration, with an accuracy of 93 and 73% for diagnostic and intermediate cells, respectively. These results indicate that regardless of the extent of nuclear atypia as reflected in the CIN category, features of chromatin organization can potentially be used to predict the malignant or progressive potential of CIN lesions.

Adult↗

PubMed1996

MIB1, a promising marker for the classification of cervical intraepithelial neoplasia.

Formalin-fixed and paraffin-embedded tissue specimens of normal and dysplastic cervical epithelia (five CIN1, seven CIN2, five CIN3, and five normal) were assessed by an immunoperoxidase technique, using the monoclonal antibody MIB1, regonizing a formalin-fixation-resistant epitope on the cell proliferation-associated Ki-67 antigen. An image analysis system was used to measure four parameters associated with proliferative activity: the Ki-67 labelling index (LI), the number of Ki-67-positive nuclei per unit length of basement membrane, and the maximum value and 90th percentile of the relative distances of Ki-67-positive nuclei from the basement membrane. All these four proliferation-related parameters were highly correlated with the grade of dysplastic change in the epithelium (0.90 < r < 0.97, p < 0.0001). The best correlation was found for the 90th percentile of the relative distance and with this parameter all CIN lesions could be correctly classified. The means and standard deviations of the Ki-67 LIs in normal epithelium, CIN1, CIN2, and CIN3 lesions were 0.07 +/- 0.03, 0.16 +/- 0.03, 0.25 +/- 0.06, and 0.39 +/- 0.06, respectively. These findings support the theory that CIN involves a progressive dysfunction of the proliferative activity of cervical epithelial cells. Image analysis of MIB1 is a promising alternative method for the objective, reproducible, and reliable classification of dysplastic changes in cervical epithelium.

Antibodies, Monoclonal↗

PubMed1995

A model for evaluation of in situ hybridization spot-count distributions in tissue sections.

The interpretation of in situ hybridization (ISH) spot-count distributions, obtained from evaluation of ISH signals in tissue sections, is complicated by the unknown impact of nuclear truncation and of the localization of ISH spots within the nuclei. In this study, a mathematical model was developed to investigate the effects of nuclear truncation and of the distribution of ISH spots within the nucleus on the ISH spot-count distribution in tissue sections. In this model, it was assumed that nuclei are spherical and of constant diameter and that ISH spots have negligible size and are distributed randomly within the nucleus ("volume model") or along the nuclear membrane ("surface model"). A minimal nuclear profile diameter was introduced in order to study the effect of rejecting small nuclear fragments for spot-count evaluation. Given the section thickness, the nuclear size, the minimal nuclear profile diameter, and the true number of ISH spots per nucleus and their spatial distribution within the nucleus, the model predicts the proportion of nuclei observable in the section with a specific number of ISH spots. A program that performs the model calculations was developed for PC and is available upon request. For section thickness greater than 50% of the nuclear diameter, the main effect of increasing section thickness on spot-count distributions was the increase of the proportion of nuclei with the true chromosome copy number of spots. For lower section thickness, the total distribution shifted towards lower spot frequencies. The influence of the minimal profile diameter was most notable for values close to the nuclear diameter. The effect of the localization of ISH spots within the nucleus was shown to be prominent, especially for sections with thickness smaller than the nuclear diameter. Good correspondence between model-predicted distributions and measured distributions was obtained using the volume model and taking into account only large nuclear profiles.

Animals↗

PubMed1994

Practical aspects of fixatives in high resolution nuclear image analysis.

Three different methods of fixation (ethanol/Carbowax, formaldehyde, and Carnoy) and four different protocols (without Böhm post-fixation on an uncoated slide, and with Böhm post-fixation on Poly-L-Lysine coated slide, an uncoated slide and a previously Papanicolaou stained slide) were evaluated for their application in high resolution image analysis of Feulgen stained nuclei. The aim of the study was to find a combination with the best reproducibility and the least variance under normal laboratory conditions. Care was taken not to exclude any "normal" laboratory variability. The combinations were evaluated for densitometric, geometric, as well as texture features. Selected features were determined on a CAS-100 using the Cell Measurement Program (Cell Analysis Systems, Inc. Lombard, IL). Diploid and tetraploid rat liver nuclei were used. Ethanol/Carbowax fixation-with Böhm post-fixation proved most stable. This fixation method also gave feature values for previously Papanicolaou stained slides that were comparable to direct Feulgen stained nuclei. Acceptable results were achieved with all three fixatives and the various combinations if one adhered strictly to protocol. In routine practice this usually does not happen. Therefore Ethanol/Carbowax fixation with Böhm post-fixation was considered most suited for routine determination of feature values on the CAS-100.

Acetates↗

PubMed1992

DNA changes in progressive cervical intraepithelial neoplasia.

Cervical intra-epithelial neoplasm (CIN) is treated as a progressive lesion, even though most CIN will not progress to invasive cancer if left untreated. This study focussed on DNA-cytometric analysis of cytologic smears of patients who had developed invasive cancer after initial smears showing CIN. The first part of the study aimed at describing the DNA-cytometric changes in these progressive ('malignant') CIN lesions. In the second part a cluster analysis was performed on 'malignant' CIN III lesions and CIN III lesions, with 'unknown' malignant potential. The results indicated that 'malignant' CIN lesions developed high DNA-index (DI) values during malignant transformation, as demonstrated by increasing mean DI values, a high percentage of DNA-aneuploidy and 2.5c Exceeding Rates. Furthermore, a trend-like pattern of texture feature values occurred in 'malignant' CIN lesions with increasing severity. These findings provide objective quantitative confirmation of the evolution of nuclear changes during malignant transformation. Cluster analysis showed that it was possible, using a set of four cytometric features, to subdivide the 'unknown' CIN III lesions into a cluster of lesions with feature values similar to the vast majority of the 'malignant' CIN III lesions, and a second cluster of lesions with feature values dissimilar to 'malignant' CIN III. It is argued that the profile of 'malignant' CIN has become clearer and that the results of this study may serve as a basis for a more objective cytopathologic subdivision of premalignant CIN. It may be justified to follow up patients whose lesions do not yet fit this 'malignant' profile. Not treating the non-progressive lesion group will avoid putting these patients at risk.

Aneuploidy↗

PubMed1991

Cytophotometric analysis of corresponding cytological and histological cervical intraepithelial neoplasia grade III specimens.

Cytophotometric analysis of cervical intraepithelial neoplasia grade III (CIN III) was performed in 22 cytological smears (CS) and in 22 corresponding cytospin specimens retrieved from selected areas of paraffin-embedded tissues (PEC). The average time interval between cytological and histological diagnosis was 6 weeks. CIN III nuclei in CS and PEC specimen were Thionin-Feulgen stained and digitized. Beside the visual classification of DNA ploidy patterns, the 2.5c and 5c exceeding rates and the specimen mean and standard deviation values of 21 photometric features were also analyzed. It was shown that, although there was a significant correlation between DNA ploidy patterns in corresponding PEC and CS specimen, the DNA patterns were dissimilar in eight of 22 cases. The DNA index, as represented by 2.5c and 5c exceeding rates, was significantly higher in the CS specimen. High-resolution cytophotometric analysis of cell nuclei in CS and PEC specimens showed significant differences for a large number of nuclear photometric features. These findings can possibly be explained by differences in selection of CIN III cells from CS and PEC specimens and by differences between fixation procedures as used for the two techniques. It was concluded that cytophotometric data of CS and PEC specimens representing CIN III lesions should not be regarded as interchangeable.

Cytophotometry↗

PubMed1990

DNA ploidy and cytophotometric analysis of cervical intraepithelial neoplasia grade III and invasive squamous cell carcinoma.

Cervical intraepithelial neoplasia grade III (CIN III) and squamous cell carcinoma (INV) were examined using DNA ploidy and cytophotometric analysis. Based on hysterectomy, exconisation, and biopsy material from 69 patients in two age categories, analysis was performed in nuclei isolated from selected areas of paraffin-embedded tissue. High percentages of DNA-diploidy in INV lesions were found mainly in the group of patients age 45 years or younger. CIN III lesions in women age 46 or older demonstrated high percentages of DNA-aneuploidy. DNA-polyploidy was most frequent in CIN III lesions in the younger age category. The results of cytophotometric analysis indicated that the overall mean values of 16 nuclear photometric features discriminated significantly between the whole groups of CIN III (n = 37) and INV (n = 32). On an individual patient level, however, the mean feature values showed a large overlap. Based on the results of a stepwise linear discriminant analysis of patient mean values, a combination of geometrical and run-length texture features was used to discriminate between CIN III and INV lesions. The correct classification rate was highest in the category of patients in the older age category. The results of this study indicate age related differences in CIN III and invasive squamous cell carcinoma, and they may be of help in assessing cytophotometric features in the study of progressive and non-progressive CIN lesions.

Adult↗

PubMed1990

Cytophotometric analysis of cervical intraepithelial neoplasia grade III, with and without synchronous invasive squamous cell carcinoma.

Cytophotometric analysis was performed in nuclei retrieved from paraffin-embedded cervical tissue from 57 cases of CIN III. CIN III lesions of patients without invasive squamous cell carcinoma (N = 37) were regarded to represent a mixture of progressive and nonprogressive lesions. The CIN III lesions of patients with a synchronous invasive squamous cell carcinoma (N = 20) were regarded as representing truly progressive precursor lesions (CIN.INV). Twenty-one photometric features describing geometrical, density, and texture characteristics were extracted from the digitized nuclear images. Statistical analysis of cytophotometric data indicated significant differences between the group of CIN III lesions and CIN.INV lesions. A cluster analysis, using one co-occurrency texture feature (S-HOMOG), one density feature (S-DI), and two geometrical features (S-AREA and M-CIRC), showed that two clusters (C1 and C2) were present in the total group of CIN III and CIN.INV lesions. The vast majority of CIN.INV lesions was member of one and the same cluster C1. The CIN III group appeared to consist of a mixture of two clusters, 54% C1 and 46% C2 lesions. Of patients 45 years or younger, the majority (62%) of CIN III lesions had feature values, corresponding with those of cluster C1, and as such possibly with a potentially progressive course. In patients older than 45 years the percentage of CIN III lesions with C1 feature values was 27%.

Adult↗

PubMed1989

Image and flow DNA cytometry of small cell carcinoma of the lung.

Both image and flow DNA cytometry were performed in isolated nuclei from paraffin-embedded tumor tissue of patients with small cell carcinoma of the lung (SCCL). In 14 patients tissue was obtained by surgery from the primary tumor. From 14 patients tissue was taken by autopsy. From two patients tissue obtained by both surgery and later autopsy were available. From the autopsy patients tissue was taken only from the primary tumor (n = 6), from a metastasis (n = 1) and from the primary tumor and distant metastases (n = 7). Twelve of the tumors obtained by surgery were diploid, and two multiploid (two stem lines present). This was found both with image and flow cytometry. The group of patients could clearly be subdivided in short survivors (less than 9 months, n = 6) and long survivors (greater than 16 months, n = 8); since in both groups one multiploid and the remainder diploid cases were present, ploidy did not seem to be a good prognosticator for survival. In most (n = 26) of the tissues measured from the autopsy patients, again, a good correlation between image and flow DNA cytometry was obtained, the histograms being either (near) diploid or multiploid. In six cases, however, flow cytometry showed multiploidy whereas image showed aneuploidy (one single peak clearly deviating from diploidy). This discrepancy is caused because normal diploid (nonneoplastic) cells in the preparations could not be discarded from the flow cytometry measurements. Using the image cytometry data of the primary tumors, five diploid, three aneuploid, and four multiploid tumors were found. In five of the seven patients of whom tissue was obtained from the primary tumor and multiple metastases, differences between the histograms were found, mostly showing two malignant cell populations in one tissue and only one of them in another. Of one of the two patients of whom tissue was obtained by surgery and later autopsy, a change in histogram pattern was observed. It is concluded that although there is a high similarity between image and flow DNA cytometry, for an optimal interpretation of the histogram pattern, image measurements are more reliable. Ploidy determination does not seem to be of use in prediction of survival, and care should be taken in interpreting DNA histograms of metastases in SCCL patients because of the variability in histogram pattern.

Adult↗

PubMed1988

DNA ploidy patterns in cervical intraepithelial neoplasia grade III, with and without synchronous invasive squamous cell carcinoma. Measurements in nuclei isolated from paraffin-embedded tissue.

This study presents the results of cytophotometric (CPM) and flow cytometric (FCM) DNA ploidy measurements in cervical intraepithelial neoplasias grade III (CIN III) with and without synchronous invasive squamous cell carcinoma. Hysterectomy and biopsy material from 21 patients 35 years of age or younger and from 18 patients age 50 years or older was studied. The DNA analysis was performed in nuclei isolated from specific areas of paraffin-embedded tissue. There were significant differences in the distribution of DNA patterns between the two age groups. About 80% of CIN III lesions in women 50 years of age or older, with or without a coexisting invasive cancer were aneuploid. In the group of younger women a diploid DNA pattern was found in about 60% of CIN III with concomitant invasive cancer. In the absence of an invasive cancer, CIN III lesions were mostly polyploid. The DNA pattern of invasive cancers was generally identical with the adjacent CIN, thus suggesting that the two lesions were related. Although the prognostic value of DNA ploidy measurements in cervical intraepithelial lesions in women in these two age groups has to be further evaluated, these results are at considerable variance with previously published data on DNA values in CIN and invasive carcinoma. In four CIN III lesions without invasive cancer, in women of the group of 35 years of age or younger, human papilloma virus common antigen could be demonstrated by immunochemical procedure. In three of these cases a polyploid DNA pattern was present; the fourth case showed a bimodal aneuploid pattern.

Adult↗

PubMed1988

Quantitative description of classic and variant small cell lung cancer cell lines by nuclear image cytometry.

Six small cell lung cancer (SCLC) cell lines were examined using nuclear image analysis to find features characteristic of the classic and the variant type of SCLC. On the basis of their biochemical and biological properties three of these cell lines have been shown to represent the classic types, and three represent the variant type of SCLC. Using a combination of the image-derived run length, density, and geometric features, it was possible to distinguish between the classic and variant SCLC cell lines. The results of this study may be of help in assessing photometric features for the separation of the classic and variant subtypes of SCLC in solid tumors. Because of differences in treatment and prognosis between these two subtypes, such a separation may be of clinical value.

Carcinoma, Small Cell↗

PubMed1988

DNA cytometry of pure dysgerminomas of the ovary.

The value of DNA cytometry for predicting the malignant potential of pure dysgerminomas of the ovary was investigated. Feulgen-thionin DNA image cytometry and propidium iodide DNA flow cytometry were performed in isolated nuclei from paraffin-embedded tissue of 25 dysgerminomas. Nineteen were in clinical stage Ia1, and six were in stage III (two in IIIa and four in IIIb). Eight patients showed recurrences during a 5-year follow-up period and one patient died of the disease. The DNA index (DI; the modal DNA value compared with that of normal cells) of the tumor cells was computed from the image and flow DNA histograms. Comparable DI values, ranging from 1.29 and 3.23, were found with both types of cytometry. No correlation was found with the clinical stage or the recurrence state. Furthermore, it was striking that, although values were found strongly deviating from the normal diploid content (DI = 1.0) suggesting an unfavorable prognosis, the survival rate was relatively high. It can be concluded that prediction of the clinical course of pure dysgerminomas by DNA cytometry does not seem feasible.

Cell Nucleus↗

PubMed1987

Carcinoma in situ specimen classification based on intermediate cell measurements.

In this study the possibility of classifying carcinoma in situ and normal specimens by measuring normal-appearing intermediate cells was explored. Twenty-five histologically verified carcinoma in situ specimens and 99 normal specimens, matched with the abnormal specimens for age and use or nonuse of an oral hormonal contraceptive, were examined. The smears were monolayer preparations stained with Thionin-Feulgen Congo red. Twenty-one nuclear features were measured. A discrimination among the experimental groups could be made on the basis of the relationship between two features, area and average optical density (AOD). A regression of AOD on area for each smear was performed. The correlation, coefficient of variation, slope, intercept, as well as the mean of the AOD level and the age of the subject were used in a discriminant analysis. This resulted in a smear classification with a false-positive rate of 14% and a missed-positive rate of 32%. When contraceptive use was taken into account the overall classification was improved with a false-positive rate of 12% and a missed-positive rate of 20%.

Adult↗

PubMed1987

Evaluation of DNA and nuclear protein features for their use in describing normal and malignant endometrium.

Different feature sets (geometric, densitometric, and textural) derived from DNA and nuclear protein staining were evaluated for their use in describing atrophic, secretory, and proliferative endometrium, and well-differentiated stage I and moderately differentiated stage I adenocarcinomas of the endometrium. It was found that the pattern of significant differences among these groups varied between feature sets, while remaining consistent within a set of features. The DNA density and run-length features were not very effective in describing group mean differences, whereas the co-occurrence features revealed significant differences among most groups. The protein run-length features were the only ones that consistently showed a difference between proliferative endometrium and well-differentiated adenocarcinomas. Analyses repeated on only cells in the G0/G1 DNA region improved the differentiation between moderately differentiated adenocarcinomas and the other groups. It was concluded that the use of DNA and nuclear protein texture features are effective in describing group differences that cannot be described by DNA content only.

Adenocarcinoma↗