scholarly journals Markovian analysis of cervical cell images.

1976 ◽  
Vol 24 (1) ◽  
pp. 138-144 ◽  
Author(s):  
N J Pressman
Keyword(s):  
Transition Probabilities ◽  
Classification Performance ◽  
Error Rates ◽  
Classification Error ◽  
Correct Classification ◽  
Cell Nuclei ◽  
Step Size ◽  
Cervical Cell ◽  
Texture Parameters ◽  
Markovian Analysis

Markovian analysis is a method to measure optical texture based on gray-level transition probabilities in digitized images. Experiments are described that investigate that classification performance of parameters generated by Markovian analysis. Results using Markov texture parameters show that the selection of a Markov step size strongly affects classification error rates and the number of parameters required to achieve the maximum correct classification rates. Markov texture parameters are shown to achieve high rates of correct classification in discriminating images of normal from abnormal cervical cell nuclei.

scholarly journals Texture analysis of cervical cell nuclei by segmentation of chromatin patterns.

1979 ◽  
Vol 27 (1) ◽  
pp. 199-203 ◽  
Author(s):  
A W Smeulders ◽  
L Leyte-Veldstra ◽  
J S Ploem ◽  
C J Cornelisse
Keyword(s):  
Texture Analysis ◽  
Single Cell Analysis ◽  
Automated Classification ◽  
Cell Nuclei ◽  
Step Size ◽  
Data Set ◽  
Chromatin Pattern ◽  
Cervical Cell ◽  
Texture Parameters ◽  
Visual Properties

Texture parameters of the nuclear chromatin pattern can contribute to the automated classification of specimens on the basis of single cell analysis in cervical cytology. Current texture parameters are abstract and therefore hamper understanding. In this paper texture parameters are described that can be derived from the chromatin pattern after segmentation of the nuclear image. These texture parameters are more directly related to the visual properties of the chromatin pattern. The image segmentation procedure is based on a region grow algorithm which specifically isolates high chromatin density. The texture analysis method has been tested on a data set of images of 112 cervical nuclei on photographic negatives digitized with a step size of 0.125 micron. The preliminary results of a classification trial indicate that these visually interpretable parameters have promising discriminatory power for the distinction between negative and positive specimens.

scholarly journals Estimating false positive and false negative error rates in cervical cell classification.

1977 ◽  
Vol 25 (7) ◽  
pp. 696-701 ◽  
Author(s):  
L H Oliver ◽  
R S Poulsen ◽  
G T Toussaint
Keyword(s):  
False Negative ◽  
Recognition System ◽  
Classification Performance ◽  
Error Rates ◽  
Cell Classification ◽  
Cervical Cell ◽  
Performance Curves ◽  
False Negative Error ◽  
A Cell ◽  
Cervical Cells

The performance of a cell recognition system on unknown data is often estimated in terms of its error rates on a test set. This paper investigates methods for producing estimates of error rates in cervical cell classification. Classification performance curves calculated using these methods are given for several classification schemes used to classify 1500 cervical cells.

scholarly journals AUTOMATED CLASSIFICATION OF NORMAL AND ABNORMAL LEUKOCYTES

1974 ◽  
Vol 22 (7) ◽  
pp. 697-706 ◽  
Author(s):  
J. F. BRENNER ◽  
P. W. NEURATH ◽  
W. D. SELLES ◽  
T. F. NECHELES ◽  
E. S. GELSEMA ◽  
...  
Keyword(s):  
Automated System ◽  
Classification Error ◽  
Correct Classification ◽  
Automated Classification ◽  
Texture Parameters ◽  
Blood Smears ◽  
Definition Of ◽  
Peripheral Blood Smears ◽  
Optimal Set

The development of an automated system for counting and classifying normal and abnormal leukocytes in peripheral blood smears is described. General requirements are discussed and the results of a simulation experiment are presented. A sample of 1572 leukocytes, divided equally among 17 types, was photographed and analyzed using computerized pattern recognition techniques. Various geometrical, color and texture parameters were extracted from the cell images and an optimal set of 20 were used in several computerized classification runs. Training on one-half of the sample and classifying the other half resulted in an over-all correct classification of between 67 and 77% depending on the definition of classification error. When only normal cells are considered, correct classification is obtained for 9l.5% of the cells.

scholarly journals Clinical Value of RNA Sequencing–Based Classifiers for Prediction of the Five Conventional Breast Cancer Biomarkers: A Report From the Population-Based Multicenter Sweden Cancerome Analysis Network—Breast Initiative

2018 ◽  
pp. 1-18 ◽  
Author(s):  
Christian Brueffer ◽  
Johan Vallon-Christersson ◽  
Dorthe Grabau ◽  
Anna Ehinger ◽  
Jari Häkkinen ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Endocrine Therapy ◽  
Single Gene ◽  
Growth Factor Receptor ◽  
Population Based ◽  
Classification Performance ◽  
Error Rates ◽  
Classification Error ◽  
Rna Seq ◽  
Clinical Value

Purpose In early breast cancer (BC), five conventional biomarkers—estrogen receptor (ER), progesterone receptor (PgR), human epidermal growth factor receptor 2 (HER2), Ki67, and Nottingham histologic grade (NHG)—are used to determine prognosis and treatment. We aimed to develop classifiers for these biomarkers that were based on tumor mRNA sequencing (RNA-seq), compare classification performance, and test whether such predictors could add value for risk stratification. Methods In total, 3,678 patients with BC were studied. For 405 tumors, a comprehensive multi-rater histopathologic evaluation was performed. Using RNA-seq data, single-gene classifiers and multigene classifiers (MGCs) were trained on consensus histopathology labels. Trained classifiers were tested on a prospective population-based series of 3,273 BCs that included a median follow-up of 52 months (Sweden Cancerome Analysis Network—Breast [SCAN-B], ClinicalTrials.gov identifier: NCT02306096), and results were evaluated by agreement statistics and Kaplan-Meier and Cox survival analyses. Results Pathologist concordance was high for ER, PgR, and HER2 (average κ, 0.920, 0.891, and 0.899, respectively) but moderate for Ki67 and NHG (average κ, 0.734 and 0.581). Concordance between RNA-seq classifiers and histopathology for the independent cohort of 3,273 was similar to interpathologist concordance. Patients with discordant classifications, predicted as hormone responsive by histopathology but non–hormone responsive by MGC, had significantly inferior overall survival compared with patients who had concordant results. This extended to patients who received no adjuvant therapy (hazard ratio [HR], 3.19; 95% CI, 1.19 to 8.57), or endocrine therapy alone (HR, 2.64; 95% CI, 1.55 to 4.51). For cases identified as hormone responsive by histopathology and who received endocrine therapy alone, the MGC hormone-responsive classifier remained significant after multivariable adjustment (HR, 2.45; 95% CI, 1.39 to 4.34). Conclusion Classification error rates for RNA-seq–based classifiers for the five key BC biomarkers generally were equivalent to conventional histopathology. However, RNA-seq classifiers provided added clinical value in particular for tumors determined by histopathology to be hormone responsive but by RNA-seq to be hormone insensitive.

scholarly journals Multilayer Learning Network for Modulation Classification Assisted with Frequency Offset Cancellation in Satellite to Ground Link

2018 ◽  
Vol 2018 ◽  
pp. 1-13 ◽  
Author(s):  
Guan Qing Yang
Keyword(s):  
Carrier Frequency ◽  
Carrier Frequency Offset ◽  
Frequency Offset ◽  
Training Sequence ◽  
Classification Performance ◽  
Modulation Classification ◽  
Correct Classification ◽  
Offset Cancellation ◽  
Learning Network ◽  
Hidden Layer

A multilayer learning network assisted with frequency offset cancellation is proposed for modulation classification in satellite to ground link. Carrier frequency offset greatly reduces modulation classification performance. It is necessary to cancel frequency offset before modulation classification. Frequency offset cancellation weights are established through multilayer learning network based on MSE criterion. Then the weight and hidden layer of multilayer learning network are also established for modulation classification. The hidden layers and weight are trained and tuned to combat the interference introduced by frequency offset. Compared with current modulation classification algorithm, the proposed multilayer learning network greatly improves the Probability of Correct Classification (PCC). It has been proven that the proposed multilayer learning network assisted with frequency offset has higher performance for modulation classification within the same training sequence.

scholarly journals A New Nearest Centroid Neighbor Classifier Based on K Local Means Using Harmonic Mean Distance

Information ◽  
2018 ◽  
Vol 9 (9) ◽  
pp. 234 ◽  
Author(s):  
Sumet Mehta ◽  
Xiangjun Shen ◽  
Jiangping Gou ◽  
Dejiao Niu
Keyword(s):  
Small Sample Size ◽  
Classification Performance ◽  
Error Rates ◽  
Small Sample ◽  
Harmonic Mean ◽  
Local Means ◽  
Local Mean ◽  
Real World Datasets ◽  
Nearest Neighbour Classifier ◽  
Query Sample

The K-nearest neighbour classifier is very effective and simple non-parametric technique in pattern classification; however, it only considers the distance closeness, but not the geometricalplacement of the k neighbors. Also, its classification performance is highly influenced by the neighborhood size k and existing outliers. In this paper, we propose a new local mean based k-harmonic nearest centroid neighbor (LMKHNCN) classifier in orderto consider both distance-based proximity, as well as spatial distribution of k neighbors. In our method, firstly the k nearest centroid neighbors in each class are found which are used to find k different local mean vectors, and then employed to compute their harmonic mean distance to the query sample. Lastly, the query sample is assigned to the class with minimum harmonic mean distance. The experimental results based on twenty-six real-world datasets shows that the proposed LMKHNCN classifier achieves lower error rates, particularly in small sample-size situations, and that it is less sensitive to parameter k when compared to therelated four KNN-based classifiers.

Abstract 343: Prediction of Lifesaving Interventions Using Changes in Vital Signs from Prehospital Scene to Emergency Department

Circulation ◽  
2014 ◽  
Vol 130 (suppl_2) ◽  
Author(s):  
Jorge H Mena Munoz ◽  
Ashley Petersen ◽  
Francis X Guyette
Keyword(s):  
Emergency Department ◽  
Decision Trees ◽  
Recursive Partitioning ◽  
Vital Signs ◽  
Large Change ◽  
Predictive Performance ◽  
Classification Performance ◽  
Secondary Outcome ◽  
Classification Error ◽  
Trauma Patients

Objective: We investigate whether changes in vital signs between the prehospital scene and emergency department (ED) can be used to develop triage tools to predict the need for life-saving interventions (LSI) and survival in trauma patients. Methods: We analyzed a prospective cohort with any prehospital systolic blood pressure (SBP) ≤ 90 mmHg or Glasgow Coma Scale ≤ 8 who were admitted to an ED at 11 sites of the Resuscitation Outcomes Consortium. The primary outcome was the need for in-hospital LSI (e.g. invasive airway management, invasive bleeding control, blood transfusion, craniotomy, cardiopulmonary resuscitation). Secondary outcome was survival to hospital discharge. Changes in heart rate (HR), SBP, shock index (SI), and respiratory rate (RR) from first prehospital assessment to first ED assessment were considered as predictors in addition to sex, age, mechanism of injury, trauma center level, duration of transport, type of transport, and prehospital fluid volume. Decision trees for each outcome were developed using binary recursive partitioning with predictive performance measured using sensitivity, specificity, and classification error. Results: 5625 subjects were included in our analysis with 49% in need of LSI and 21% dying prior to discharge. Patients needing an LSI tended to either: (1) have an increasing SI (delta ≥ 0.22), (2) have a decreasing SI (delta < 0.22) and >500 mL prehospital fluids, or (3) have a decreasing SI (delta < 0.22), ≤500 mL prehospital fluids, and large change in RR (delta ≥ 9.5 or delta < -7.5). Those surviving to discharge tended to either: (1) have a decreasing SI (delta < 0.57) and a HR that did not decrease greatly (delta > -47) or (2) have an increase in SI (0.57 ≤ delta < 1) and a declining RR (delta < 5). LSI tree had a sensitivity of 58.7% and specificity of 63.3%. Survival tree had sensitivity of 96.2% and specificity of 21.3%. Conclusion: Though the decision trees were constructed with the best data in terms of initial triage and early secondary triage, the classification performance was limited. This highlights the difficulties of developing vital sign based triage tools to predict the need for LSI and survival.

scholarly journals From the slit-island method to the Ising model: Analysis of irregular grayscale objects

2014 ◽  
Vol 24 (1) ◽  
pp. 49-63 ◽  
Author(s):  
Przemysław Mazurek ◽  
Dorota Oszutowsk A-M Ażurek
Keyword(s):  
Image Analysis ◽  
Fractal Dimension ◽  
Ising Model ◽  
Image Data ◽  
Input Image ◽  
Cell Nuclei ◽  
Cervical Cell ◽  
Atypical Cells ◽  
Model Context ◽  
Slit Island Method

Abstract The Slit Island Method (SIM) is a technique for the estimation of the fractal dimension of an object by determining the area- perimeter relations for successive slits. The SIM could be applied for image analysis of irregular grayscale objects and their classification using the fractal dimension. It is known that this technique is not functional in some cases. It is emphasized in this paper that for specific objects a negative or an infinite fractal dimension could be obtained. The transformation of the input image data from unipolar to bipolar gives a possibility of reformulated image analysis using the Ising model context. The polynomial approximation of the obtained area-perimeter curve allows object classification. The proposed technique is applied to the images of cervical cell nuclei (Papanicolaou smears) for the preclassification of the correct and atypical cells.

Sign in / Sign up

Export Citation Format

Share Document