Fractal Analysis: An Objective Method for Identifying Atypical Nuclei in Dysplastic Lesions of the Cervix Uteri (original) (raw)
Related papers
BMC Medical Physics, 2014
Background: Fractal geometry has been the basis for the development of a diagnosis of preneoplastic and neoplastic cells that clears up the undetermination of the atypical squamous cells of undetermined significance (ASCUS). Methods: Pictures of 40 cervix cytology samples diagnosed with conventional parameters were taken. A blind study was developed in which the clinic diagnosis of 10 normal cells, 10 ASCUS, 10 L-SIL and 10 H-SIL was masked. Cellular nucleus and cytoplasm were evaluated in the generalized Box-Counting space, calculating the fractal dimension and number of spaces occupied by the frontier of each object. Further, number of pixels occupied by surface of each object was calculated. Later, the mathematical features of the measures were studied to establish differences or equalities useful for diagnostic application. Finally, the sensibility, specificity, negative likelihood ratio and diagnostic concordance with Kappa coefficient were calculated. Results: Simultaneous measures of the nuclear surface and the subtraction between the boundaries of cytoplasm and nucleus, lead to differentiate normality, L-SIL and H-SIL. Normality shows values less than or equal to 735 in nucleus surface and values greater or equal to 161 in cytoplasm-nucleus subtraction. L-SIL cells exhibit a nucleus surface with values greater than or equal to 972 and a subtraction between nucleus-cytoplasm higher to 130. L-SIL cells show cytoplasm-nucleus values less than 120. The rank between 120-130 in cytoplasm-nucleus subtraction corresponds to evolution between L-SIL and H-SIL. Sensibility and specificity values were 100%, the negative likelihood ratio was zero and Kappa coefficient was equal to 1. Conclusions: A new diagnostic methodology of clinic applicability was developed based on fractal and euclidean geometry, which is useful for evaluation of cervix cytology.
Fractal and Euclidean Geometrical Diagnosis of Cervix Cytology
Journal of Cancer Science & Therapy, 2014
Background: Conventional methods for evaluation of cervix cytology show reproducibility problems. To solve this, there was developed a diagnostic methodology based on fractal and euclidean geometry, mathematically differentiating normality, L SIL and H SIL. Objective: The aim of the present work is to confirm the clinical applicability of such diagnostic in a blind study. Methods: The clinic diagnosis of 15 normal cells, 15 ASCUS, 15 L SIL and 15 H SIL was masked. Cellular nucleus and cytoplasm were evaluated calculating fractal dimension, number of spaces occupied by the frontier and number of pixels occupied by the surface of each object. The mathematical diagnosis was established and compared with the conventional diagnosis, calculating specificity, sensibility, negative likelihood ratio and Kappa coefficient. Results: It was found that simultaneous measures of the nuclear surface and the subtraction between the frontiers of cytoplasm and nucleus, lead to differentiate normality, L SIL and H SIL. Both sensibility and specificity values were of 100 percent. Kappa coefficient was 1 and negative likelihood ratio was zero. 4 ASCUS showed mathematical measures of normality, while the remaining 11 showed values of L-SIL cells. Conclusion: The mathematical diagnostic prove to be useful for clinical evaluation of cervix cytology, differentiating normality, L SIL and H SIL, quantifying how close it is the cell to a higher severity stage, and clearing up the undetermination of the ASCUS cells.
Fractal Analysis of Cervical Intraepithelial Neoplasia
Introduction: Cervical intraepithelial neoplasias (CIN) represent precursor lesions of cervical cancer. These neoplastic lesions are traditionally subdivided into three categories CIN 1, CIN 2, and CIN 3, using microscopical criteria. The relation between grades of cervical intraepithelial neoplasia (CIN) and its fractal dimension was investigated to establish a basis for an objective diagnosis using the method proposed.
Journal of Biosciences and Medicines, 2018
Background: The concomitant use of fractal and Euclidian measurements has led to the development of new methodologies of cell evaluation, including a diagnosis of cervical cells that set up differences between normality and various degrees of lesion, to carcinoma. Aim: To confirm the diagnostic capacity of the methodology based on fractal and Euclidian geometry for the mathematical diagnosis through a blind study of normal cells and with different types of lesion, as atypia of undetermined significance (ASCUS), low grade squamous intra-epithelial lesion (LGSIL) and high grade squamous intra-epithelial lesion (HGSIL). Methods: 100 cells of Papanicolaou tests were analyzed and divided into 4 groups according to conventional parameters: 25 normal, 25 ASCUS, 25 LGSIL and 25 HGSIL. By means of the Box-counting Fractal Space, we calculated the fractal dimension and occupying spaces of the border and surface in pixels of the cell nucleus and cytoplasm. The diagnostic parameters of the previously developed methodology were applied and compared with the conventional diagnosis, setting up sensibility, specificity, negative likelihood ratio and Kappa coefficient. Results: The values of the occupation of the border and surface of the cell nucleus and cytoplasm were consistent with the values found by the diagnostic methodology previously found. The subtraction of the nucleus and cytoplasm frontiers presented values between: 189 and 482 for normality; 159
Fractal dimension and image statistics of anal intraepithelial neoplasia
Chaos, Solitons & Fractals, 2011
It is well known that human papillomaviruses (HPV) induce a variety of tumorous lesions of the skin. HPV-subtypes also cause premalignant lesions which are termed anal intraepithelial neoplasia (AIN). The clinical classification of AIN is of growing interest in clinical practice, due to increasing HPV infection rates throughout human population. The common classification approach is based on subjective inspections of histological slices of anal tissues with all the drawbacks of depending on the status and individual variances of the trained pathologists.
Automated detection of anomalies in cervix cells using image analysis and machine learning
Comparative Clinical Pathology, 2018
Usage machine-based learning image cytometry to establish the diagnosis of cervix cancer using cellular morphology classification in comparison to the conventional cytological test. The study was divided into two phases consisting of 15 samples of cervix cells. In phase1, with previous diagnosis, the samples were divided into three groups of five samples each: normal (NC), low-grade squamous intraepithelial lesion (LGSIL or LSIL), and high-grade squamous intraepithelial lesion (HGSIL or HSIL). Images of cells were analyzed to create a training set of cells with known diagnosis for machine learning purposes. With the numerical data created, the software was trained to automatically classify the three types of cells. In phase 2, 885 cells were classified without previous diagnosis. In a last step, the classification of CPA was compared to cytopathology. NC and HSIL were identified with a high sensitivity and specificity (99%, 99%) and (98%, 97%) respectively. While the sensitivity and specificity of LSIL cells were lower (78%, 79%). It is possible to extract features of cervical cells by automatically generating numerical data that allowed the program to identify and classify different cell classes, using simple and low-cost reagents and free, reproducible softwires.