Search PubMed⌕ Search

Biomedical subjects

S S Cross

Publications and source records attributed to S S Cross.

At least 19 recordsLinked to original sources
PubMed1996

Application of the fuzzy ARTMAP neural network model to medical pattern classification tasks.

This paper presents research into the application of the fuzzy ARTMAP neural network model to medical pattern classification tasks. A number of domains, both diagnostic and prognostic, are considered. Each such domain highlights a particularly useful aspect of the model. The first coronary care patient prognosis, demonstrates the ARTMAP voting strategy involving 'pooled' decision-making using a number of networks, each of which has learned a slightly different mapping of input features to pattern classes. The second domain, breast cancer diagnosis, demonstrates the model's symbolic rule extraction capabilities which support the validation and explanation of a network's predictions. The final domain, diagnosis of acute myocardial infarction, demonstrates a novel category pruning technique allowing the performance of a trained network to be altered so as to favour predictions of one class over another (e.g. trading sensitivity for specificity or vice versa). It also introduces a 'cascaded' variant of the voting strategy intended to allow identification of a subset of cases which the network has a very high certainty of classifying correctly.

Artificial Intelligence↗
PubMed1996

Funeral directors, mortuaries and necropsies: implications for necropsy consent rates and the prevention of infection.

AIM: To evaluate the attitudes and experiences of funeral directors in relation to necropsies. METHODS: All 1631 members of the National Association of Funeral Directors were surveyed by postal questionnaire about the purposes of necropsies, the technical and administrative problems associated with necropsied cases and their relations with relatives, mortuaries and pathology departments. RESULTS: In total, 123 funeral directors completed the questionnaire (8% response rate). Workload, proportion of cases necropsied and type of mortuary did not influence answers. Necropsies were considered important for the assessment of treatment outcome, identification of inherited disease and junior pathologist training, but not for medical audit. There was strong support for more education about necropsies. The areas of necropsy practice most frequently discussed with relatives related to concerns about funeral delay and the involvement of the coroner or equivalent authority. Funeral directors occasionally counselled relatives for or against giving necropsy consent. The commonest technical problems associated with necropsies were difficulties in embalming, leakage of body fluids and scalpel penetration in visible areas. Few administrative problems were reported; the commonest was inflexibility in body collection times. There was strong support for a national code of practice to cover relations between funeral directors and mortuaries despite general satisfaction with relations with local pathology departments. CONCLUSIONS: Although the relation among the funeral profession, mortuaries and pathology departments is largely satisfactory, a national code of practice for funeral directors and mortuaries is desirable.

Attitude↗
PubMed1996

Kappa statistics as indicators of quality assurance in histopathology and cytopathology.

Kappa statistics are widely-used to assess performance in quality assurance schemes. Low values, however, are difficult to interpret, especially when confidence intervals have not been calculated. A model of a dichotomous decision in pathology (benignancy or malignancy in fine needle aspirates of the breast) was used to calculate kappa statistics (with confidence limits) for increasing false positive rates. It was found that the level at which the upper 95% confidence interval for the kappa statistic fell below 1 was an insensitive method of detecting unsatisfactory performance as at that level the false positive rate was unacceptably high (> 1%) for all populations of specimens less than 800 in number. Either large populations of samples are required in quality assurance schemes which use kappa statistics (which may well be impractical) or other methods of assessing performance, possibly with weighted outcomes, are required.

Biopsy, Needle↗
PubMed1995

Fractal and integer-dimensional geometric analysis of pigmented skin lesions.

Accurate in vivo diagnosis of pigmented skin lesions is required to identify and excise malignant melanomas but to avoid unnecessary excision of benign lesions; the published rates of clinical diagnostic accuracy are about 65%. This study investigates whether fractal geometric analysis of pigmented skin lesions can improve the rate of diagnostic accuracy. Forty-two pigmented skin lesions (15 malignant melanomas, 21 melanocytic naevi, and 6 basal cell papillomas) on patients attending a dermatology clinic were photographed, excised, and sent for histopathological examination. The fractal dimension of the boundary of the lesions was measured using a box-counting method implemented on a microcomputer-based image analysis system. Euclidean geometric parameters were also measured. The fractal dimension of all the lesions was greater than the topological dimension (one), indicating that there is a fractal element to their structure. Using all measured parameters together, multivariate linear discriminant analysis produced a confusion matrix in which 45% of the lesions were assigned to the correct diagnostic group with a kappa statistic of 0.33. There was no significant difference between the fractal dimension of melanocytic naevi and that of malignant melanomas (p = 0.18). Although pigmented skin lesions have a fractal element to their structure, the fractal dimension of their boundaries is not a useful morphometric discriminant between the diagnostic groups of malignant melanomas and benign melanocytic naevi.

Algorithms↗
PubMed1995

Have declining clinical necropsy rates reduced the contribution of necropsy to medical research?

AIMS: To examine trends in necropsy based research output for a period of 27 years during which there has been a progressive decline in clinical necropsy rates. METHODS: The numbers of necropsy based research papers published between 1966 and 1993 were determined using the CD-Plus Medline computed literature database. RESULTS: The number of necropsy based research papers containing necropsy or a synonym in the title increased by 220% between 1966 and 1993. When papers including necropsy or a synonym in the abstract, but not in the title, were included, the proportion of all indexed papers increased from 0.35% in 1975, when abstracts were first included, to 0.53% in 1993. Analysis of the subject material indicated that necropsy based research has constantly reflected trends and advances in clinical medicine. Neuroscience related research represented the largest subject category which may reflect the difficulties in obtaining human tissue from sources other than necropsy. CONCLUSIONS: The modern necropsy continues to provide valuable information for all clinical and laboratory based disciplines. The decline in clinical necropsy rates would not yet appear to have undermined the contribution of the necropsy to research.

Autopsy↗
PubMed1994

Nm23 'anti-metastatic' gene product expression in colorectal carcinoma.

Nm23 is a putative metastasis suppressor gene and alterations in this gene have been reported in colorectal carcinomas suggestive of a role for nm23 in the dissemination of these tumours. In this study we used an affinity purified polyclonal antibody, ab-11, on formalin-fixed, paraffin-embedded sections of colorectal carcinomas from 46 patients in a three-stage avidin-biotin complex immunoperoxidase technique. Follow-up of these patients was until time of death or for 5 years, with a mean time of 31.2 months. Two observers scored the staining results from 1 to 3 according to the proportion of tumour cells positive. The association of nm23 staining with survival, sex, age, vascular invasion, and Dukes' stage was determined using Cox's regression model. The association of death from colorectal cancer and nm23 status reached marginal significance in this study (P = 0.0417). Moreover, there is some suggestion of a protective effect from nm23 as the relative risk of dying from colorectal cancer for each increment of nm23 positivity is 0.573 (95 per cent confidence limits 0.34-0.98).

Adenocarcinoma↗
PubMed1994

Fractal geometric analysis of colorectal polyps.

Colorectal polyps have a subjectively self-similar structure which suggests that these structures may have fractal elements and that the fractal dimension may be a useful morphometric discriminant. The fractal dimensions of images from haematoxylin and eosin-stained sections of 359 colorectal polyps (214 tubulovillous adenomas, 41 'pure' tubular adenomas, 29 'pure' villous adenomas, 68 metaplastic polyps, and 7 inflammatory polyps) were measured using a box-counting method implemented on a microcomputer-based image analysis system. Results were assessed using polychotomous logistic regression, confusion matrices, and kappa statistics. All examined polyps were shown to have a fractal structure in the range of scales examined. The fractal dimension was significantly different between different diagnostic categories (P < 0.0001) and was a useful discriminant between these categories (kappa statistic 0.60 for the confusion matrix with size as the other variable). The fractal dimension did not shown any significant correlation with the grade of epithelial dysplasia (P > 0.05). This study shows that colorectal polyps have a fractal structure over a defined range of magnification and Euclidean morphometric measurements will be invalid outside precisely defined conditions of resolution and magnification. The fractal dimension is a better way of quantitating the polyp shape and is a useful morphometric discriminant between diagnostic categories.

Adenoma↗
PubMed1994

Fractal geometric analysis of material from molar and non-molar pregnancies.

Histological sections from 25 non-molar pregnancies, nine partial hydatidiform moles, and 16 complete hydatidiform moles were examined (diagnosis was taken as the consensus of seven experienced histopathologists) and the fractal dimension was measured using a box-counting method implemented on a microcomputer-based image analysis system. The fractal dimensions of the different diagnostic categories were normally distributed with a mean of 1.50 for non-molar pregnancies, 1.44 for partial moles, and 1.42 for complete moles. All the measured fractal dimensions were greater than the topological dimension (1), demonstrating that the specimens had a fractal element to their structure. There was a significant difference between the fractal dimensions of non-molar pregnancies and complete moles (P = 0.0005), but not between partial moles and non-molar pregnancies (P = 0.0823) or complete and partial moles (P = 0.4400). Using the fractal dimension to predict the histopathological diagnosis assigned 56 per cent of the cases to the correct category with a kappa statistic of 0.26, so the fractal dimension, used alone, is not a useful morphometric discriminant in the diagnosis of molar and non-molar pregnancy.

Female↗
PubMed1994

Chaos and antichaos in pathology.

Deterministic chaos is a pattern of fluctuations that may seem to be stochastic (caused by random external forces) but is actually produced, in a determined manner, by the action of nonlinear dynamic processes. The existence of this phenomenon has important implications for pathologists because it means that experimental data with apparently random fluctuations may be caused by a deterministic process. The patterns noted in deterministic chaos are very sensitive to variation in initial conditions, so if chaotic processes do occur in pathology then methods of predicting biological behavior (prognosis) of lesions, such as tumors, will have to be modified. This review shows the properties of deterministic chaos using a simple mathematical model of tumor growth and describes the methods that may be used to analyze experimental data for the presence of such chaos illustrating these with some clinical data. Established applications in medicine are reviewed and antichaotic mechanisms are discussed.

Humans↗
PubMed1994

The application of fractal geometric analysis to microscopic images.

Fractal geometry is a relatively new tool for the quantitative microscopist that is a more valid way of measuring dimensions of complex irregular objects than the integer-dimensional geometries (such as Euclidean geometry). This review discusses the theory of fractal geometry using the classic examples of the Von Koch curve, the Cantor set and the Sierpinski gasket. The problems of describing the dimensions of these objects are discussed and the concept of fractal dimensionality is introduced. Methods for measuring fractal dimensions are discussed, including their implementation on microcomputer-based image analysis systems . The advantages and problems of fractal geometric analysis are discussed and current applications in the field of microscopy are reviewed.

Animals↗
PubMed1994

Measuring fractal dimensions. Sensitivity to edge-processing functions.

The fractal dimension is a useful tool in quantitative histology and cytology, and its measurement is easily implemented on computerized image analysis systems. However, the optimal conditions for capture of images and the effect of image-processing functions on the measurement of the fractal dimension have not been reported. Edge-processing functions were applied to images of Euclidean (square) and fractal (Koch island, renal angiogram) objects. The fractal dimension of processed images was measured using implementation of the box-counting method, and the area of thresholded image was also recorded. The method was shown to be accurate, with errors of < 1.5% for objects with known fractal dimensions, and highly reproducible, with a reliability coefficient of 0.972 (95% confidence limits of 0.868-0.987). The fractal dimension of the fractal images showed a marked (> 15%) reduction when a binary noise reduction function was applied with the minimum neighbors limit set above 3. In contrast, the fractal dimension of the Euclidean square was unchanged by this function. The reduction in fractal dimension was due to the erosion of complex convolutions at the edge of the fractal objects. Edge-processing functions should be avoided when manipulating images of fractal objects.

Angiography↗
PubMed1993

Trabecular bone does not have a fractal structure on light microscopic examination.

The fractal dimension of the boundary of trabecular bone in 62 biopsies was measured on histological sections using a box-counting method implemented on a microcomputer image analysis system. The calculated fractal dimension had a mean value of 0.99 with a normal distribution. Since this value is not greater than the topological dimension, trabecular bone, when examined by light microscopy, does not have a fractal structure. Conventional Euclidean dimensions will continue to be the most useful measurements in bone histomorphometry.

Adult↗
PubMed1993

Quantitation of the renal arterial tree by fractal analysis.

To determine whether the renal arterial system has a fractal structure, the fractal dimension of renal angiograms from 52 necropsy cases was measured using an implementation of the box-counting method on an image analysis system. The method was validated using objects with known fractal dimensions. The method was accurate with errors of less than 1.5 per cent and reproducible with initial values within 1.2 per cent of the mean of ten sets of measurements (reliability coefficient 0.968, 95 per cent confidence limits 0.911-0.984). In the 36 satisfactory angiograms the mean fractal dimension was 1.61 (SD 0.06), which was significantly greater than the topological dimension of 1 (P < 0.0001), indicating that the renal arterial tree has a fractal structure. There was no significant relationship between age (P = 0.494), sex (P = 0.136), or systolic (P = 0.069) or diastolic (P = 0.990) blood pressure, but two congenitally abnormal kidneys (hypoplastic dysplasia and renal artery stenosis) had fractal dimensions at the lower end of the normal range (third percentile). Since the renal arterial tree has a fractal structure, Euclidean geometric measurements, such as area and boundary length, are invalid outside precisely defined conditions of magnification and resolution.

Adult↗
PubMed1992

Combined hepatoblastoma and yolk sac tumor of the liver.

The authors report a liver tumor that occurred in a 6-month-old boy in which areas of yolk sac and hepatoblastoma were identified. To the best knowledge of the authors, this morphologic pattern has not been reported previously. Theories of histogenesis are discussed.

Carcinoma, Hepatocellular↗
PubMed1992

The fractal dimension may be a useful morphometric discriminant in histopathology.

The hypothesis is presented that fractal geometry, and the fractal dimension in particular, is likely to be a more useful method of image analysis in histopathology than the use of regular Euclidean geometry. The concept of the fractal dimension is described and it is suggested that many human tumours have a fractal structure over a wide range of scales. It is postulated that the fractal dimension will be useful in discriminating between different diagnostic categories and may also allow more precise prognostic stratification within single diagnostic groups.

Humans↗