Early diagnosis of breast cancer by gene expression profiles

Breast cancer is the second leading cause of cancer death in women worldwide. Nevertheless, there is evidence that early detection and treatment can increase the survival rate of breast cancer patients. This paper presents an intelligent decision support system (IDSS) for breast cancer diagnosis by using gene expression profiles. The proposed system first extracts significant features from the input patterns by using information gain and then employs deep genetic algorithm for feature reduction as well as for breast cancer diagnosis. The proposed system is evaluated by considering a benchmark microarray dataset and compared with the most recent systems. The results show that the proposed IDSS outperforms other systems in terms of diagnosis time and accuracy. The proposed system produces 100 % classification accuracy. In addition, the proposed system reduces the required memory space.     

Proteome analysis of substantia nigra and striatal tissue in the mouse MPTP model of Parkinson's disease

The dopaminergic neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) replicates many of the pathological hallmarks of Parkinson's disease (PD) in mice via selective destruction of dopamine neurons of the substantia nigra and striatum. Although MPTP has been widely used to study downstream effects following the degeneration of dopaminergic neurons, the underlying mechanisms of MPTP action remain poorly understood. To determine the underlying mechanisms of MPTP action at the protein level, a 2-DE-based proteomics approach was used to evaluate the changes in protein expression in substantia nigra and striatal tissue in C57BL/6 mice after MPTP administration. We identified nine proteins that were markedly altered and are likely to be involved in mitochondrial function, heat shock protein activity, and which contribute enzyme activities for energy metabolism and protein degradation.     

基于C-AdaBoost模型的乳腺癌预测研究

机器学习和深度学习技术可用于解决医学分类预测中的许多问题,其中一些分类算法的预测精度较高,而另一些算法的精度有限。提出了基于C-AdaBoost模型的集成学习算法,对乳腺癌疾病进行预测,发现了判断乳腺癌是否复发、乳腺癌肿瘤是否为良性的最优特征组合。通过逐步回归方法对现有特征进行二次选取,并结合C-AdaBoost模型使得预测效果更优。大量实验表明,基于C-AdaBoost模型的算法的预测准确率比SVM、Naive Bayes、RandomForest以及传统的集成学习模型等机器学习分类器的准确率最多可提高19.5%,从而可以更好地帮助医生进行临床决策。     

Robust predictive model for evaluating breast cancer survivability

ObjectiveMany machine learning models have aided medical specialists in diagnosis and prognosis for breast cancer. Accuracy has been regarded as a primary measurement for the performance evaluation of the models, but stability which indicates the robustness of the performance to model parameter variation also becomes essential. A stable model is in practice of benefit to the medical specialists who may have little expertise in model tuning. The main purpose of this work is to address the importance of the stability of a model and to suggest one of such models.MethodsA comparative study of three prominent machine learning models was carried out for the prognosis of breast-cancer survivability: support vector machines, artificial neural networks, and semi-supervised learning models.MaterialThe surveillance, epidemiology, and end results database for breast cancer was used, which is known as the most comprehensive source of information on cancer incidence in the United States.ResultsThe best performance was obtained from the semi-supervised learning model. It showed good overall accuracy and stability under model parameter variation. The sharpening procedure enhanced the stability of the model via the noise-reduction.ConclusionWe suggest that semi-supervised learning model is a good candidate that medical professionals readily employ without consuming the time and effort for parameter searching for a specific model. The ease of use and faster time to results of the predictive model will eventually lead to the accurate and less-invasive prognosis for breast cancer patients.     

A probabilistic model for breast cancer survival data

The aim of the present study is the development of a probabilistic model to analyze breast cancer survival data. A finite-state discrete-time Markov chain model has been formulated for the analysis of follow-up probability and mortality data for 780 breast cancer patients.

The proposed stochastic model can also be used in comparing the transition in order to estimate treatment effects.     

Using evolvable genetic cellular automata to model breast cancer

Cancer is an evolutionary process. Mutated cells undergo selection for abnormal growth and survival creating a tumor. We model this process with cellular automata that use a simplified genetic regulatory network simulation to control cell behavior and predict cancer etiology. Our genetic model gives us the ability to relate genetic mutation to cancerous outcomes. The simulation uses known histological morphology, cell types, and stochastic behavior to specifically model ductal carcinoma in situ (DCIS), a common form of non-invasive breast cancer. Using this model we examine the effects of hereditary predisposition on DCIS incidence and aggressiveness. Results show that we are able to reproduce in vivo pathological features to hereditary forms of breast cancer: earlier incidence and increased aggressiveness. We also show that a contributing factor to the different pathology of hereditary breast cancer results from the ability of progenitor cells to pass cancerous mutations on to offspring.     

A deep ensemble network model for classifying and predicting breast cancer

Breast cancer is one of the leading causes of death among women worldwide. In most cases, the misinterpretation of medical diagnosis plays a vital role in increased fatality rates due to breast cancer. Breast cancer can be diagnosed by classifying tumors. There are two different types of tumors, such as malignant and benign tumors. Identifying the type of tumor is a tedious task, even for experts. Hence, an automated diagnosis is necessary. The role of machine learning in medical diagnosis is eminent as it provides more accurate results in classifying and predicting diseases. In this paper, we propose a deep ensemble network (DEN) method for classifying and predicting breast cancer. This method uses a stacked convolutional neural network, artificial neural network and recurrent neural network as the base classifiers in the ensemble. The random forest algorithm is used as the meta-learner for providing the final prediction. Experimental results show that the proposed DEN technique outperforms all the existing approaches in terms of accuracy, sensitivity, specificity, F-score and area under the curve (AUC) measures. The analysis of variance test proves that the proposed DEN model is statistically more significant than the other existing classification models; thus, the proposed approach may aid in the early detection and diagnosis of breast cancer in women, hence aiding in the development of early treatment techniques to increase survival rate.     

Characterisation of tumoral markers correlated with ErbB2 (HER2/Neu) overexpression and metastasis in breast cancer

The receptor tyrosine kinase ErbB2 (HER2/neu) is overexpressed in ?30% of breast cancers and is associated with poor prognosis and an increased likelihood of metastasis. Clinical treatments such as trastuzumab are effective in less than 35% of women diagnosed as ErbB2‐positive, highlighting the necessity of searching for novel targets and alternative therapies. Herein, a proteomic screening strategy combining quantitative‐based gel electrophoresis and MS was used to compare the protein expression of 48 normal human breast and tumour tissues differing in ErbB2 expression and lymph node status. The aim was to identify proteins associated with the aggressive phenotype of ErbB2‐positive breast cancer which could be potential biomarkers of the disease as well as therapy targets. In total, 177 protein isoforms (107 gene products) differentially expressed between tissue groups were identified. Immunohistochemical staining of a tissue‐microarray was used for validation of selected protein candidates. We found that expression of HSP90α, laminin and GSTP1 significantly correlated with ErbB2 expression, while others such as AGR2, NM23H1 and Annexin 2 were overexpressed in greater than 40% of tumours. Finally, knocking‐down the expression by RNA interference of three candidates, AGR2, Transgelin2 and NM23H1 resulted in an enhanced invasive capacity of MDA‐MB435 cells. These data support the involvement of these targets in tumour progression and identify them as novel biomarkers of the disease.     

An optimal model of information diffusion principles to risk and decision analysis of breast cancer morbidity

Information diffusion methods (IDMs) often deal with “small sample problems” and distribute the information of one data point to its neighbors in fuzzy information processing. By considering new criteria, we establish an optimal model for parameters of IDMs to risk and decision analysis of fatal disease. We further illustrate a specific process and a successful application of IDMs by a more reasonable morbidity surface from one-dimensional and two-dimensional case studies of breast cancer analysis in Yanpu District, Shanghai.     

A new deep convolutional neural network model for classifying breast cancer histopathological images and the hyperparameter optimisation of the proposed model

The Journal of Supercomputing - Deep learning algorithms have yielded remarkable results in medical diagnosis and image analysis, besides their contribution to improvements in a number of fields...     

A review of computer assisted detection/diagnosis (CAD) in breast thermography for breast cancer detection

Breast cancer is the leading type of cancer diagnosed in women. For years human limitations in interpreting the thermograms possessed a considerable challenge, but with the introduction of computer assisted detection/diagnosis (CAD), this problem has been addressed. This review paper compares different approaches based on neural networks and fuzzy systems which have been implemented in different CAD designs. The greatest improvement in CAD systems was achieved with a combination of fuzzy logic and artificial neural networks in the form of FALCON-AART complementary learning fuzzy neural network (CLFNN). With a CAD design based on FALCON-AART, it was possible to achieve an overall accuracy of near 90%. This confirms that CAD systems are indeed a valuable addition to the efforts for the diagnosis of breast cancer. Lower cost and high performance of new infrared systems combined with accurate CAD designs can promote the use of thermography in many breast cancer centres worldwide.     

Screening of breast cancer from thermogram images by edge detection aided deep transfer learning model

Multimedia Tools and Applications - Breast cancer, the most common invasive cancer, causes deaths of thousands of women in the world every year. Early detection of the same is a remedy to lessen...     

Semi-supervised model applied to the prediction of the response to preoperative chemotherapy for breast cancer

Breast cancer is the second most frequent one, and the first one affecting the women. The standard treatment has three main stages: a preoperative chemotherapy followed by a surgery operation, then an post-operatory chemotherapy. Because the response to the preoperative chemotherapy is correlated to a good prognosis, and because the clinical and biological information do not yield to efficient predictions of the response, a lot of research effort is being devoted to the design of predictors relying on the measurement of genes’ expression levels. In the present paper, we report our works for designing genomic predictors of the response to the preoperative chemotherapy, making use of a semi-supervised machine learning approach. The method is based on margin geometric information of patterns of low density areas, computed on a labeled dataset and on an unlabeled one.     

A hybrid neural network/genetic algorithm applied to breast cancer detection and recurrence

Genetic algorithms (GAs) and neural networks (NNs) are both inspired by computation in biological systems and many attempts have been made to combine the two methodologies to boost the NNs performance. This paper deals with the evolutionary training of a feedforward NN for both breast cancer detection and recurrence. A multi‐layer perceptron (MLP) has been designed for this purpose, using a GA routine to set weights, and a Java implementation of this hybrid model has been made. Four databases concerning cancer detection and recurrence have been used, two databases containing numerical attributes only, one database containing ordinal (categorical) attributes solely and one database with mixed attributes. In comparison to some standard NNs, the performance of this approach using the same databases is shown to be superior. Moreover, this hybrid MLP/GA model is very flexible in terms of providing accurate classification, even with different types of attributes, which is usually found in medical studies.     

Label-free Her2 detection and dissociation constant assessment in diluted human serum using a longitudinal extension mode of a piezoelectric microcantilever sensor

We have investigated real-time, label-free, in situ detection of human epidermal growth factor receptor 2 (Her2) in diluted serum using the first longitudinal extension mode of a lead zirconate-lead titanate (PZT)/glass piezoelectric microcantilever sensor (PEMS) with H3 single-chain variable fragment (scFv) immobilized on the 3-mercaptopropyltrimethoxysilane (MPS) insulation layer of the PEMS surface. We showed that with the longitudinal extension mode, the PZT/glass PEMS consisting of a 1 mm long and 127 μm thick PZT layer bonded with a 75 μm thick glass layer with a 1.8 mm long glass tip could detect Her2 at a concentration of 6–60 ng/ml (or 0.06–0.6 nM) in diluted human serum, about 100 times lower than the concentration limit obtained using the lower-frequency flexural mode of a similar PZT/glass PEMS. We further showed that with the longitudinal mode, the PZT/glass PEMS determined the equilibrium H3–Her2 dissociation constant K_d to be 3.3 ± 0.3 × 10⁻⁸ M consistent with the value, 3.2 ± 0.28 × 10⁻⁸ M deduced by the surface plasmon resonance method (BIAcore).