Recent Patents on Computer Science - Volume 12, Issue 4, 2019

Background: Computer-Assisted Diagnosis (CAD) has become a common practice of use in the healthcare industry due to its improved accuracy and reliability. The CAD systems are expected to improve the quality of medical care by assisting healthcare professionals with a wide range of clinical decisions. A CAD system is a combination of Computer-Assisted Detection (CADe) and Computer-Assisted Diagnosis (CADx) system. Objective: The objective of this research article is to generate an optimized rule-set for medical diagnosis capable of providing improved accuracy. It is evident from the literature that keeping a balance between these performance parameters is a real challenge. Methods: In order to achieve the desired objective, the following two contributions have been proposed to improve diagnosis accuracy: 1) an unsupervised feature selection approach based on ACO Meta-heuristic is used to design the CADe system, and 2) an ACO assisted decision tree classifier technique is employed to make CADx system. Results: Three popular UCI (Wisconsin Breast Cancer, Pima Indian Diabetes and Liver Disorder) medical domain datasets have been used to evaluate the performance of the proposed model. The exploratory result analysis shows the efficiency of the proposed work as compared to existing work.

Background: With the expeditious development of current medical imaging technology, the availability of histopathological images has been increased in a large number. Hence, histopathological image classification and annotation have emerged as the prime research fields in the pathological diagnosis and clinical practices. Several methods are available for the automation of image classification. Methods: Recently, the bag-of-features appeared as a successful histopathological image classification method. However, all the extracted keypoints in bag-of-features are not relevant and generally have very high dimensions, which degrade the performance of a classifier. Therefore, this paper introduces a new Grey relational analysis-based bag-of-features method to search the relevant keypoints. Results: The efficacy of the proposed method has been analyzed on animal diagnostics lab histopathological image datasets having healthy and inflamed images of three organs. The average accuracy of the proposed method is 88.3%, which is the highest among other state-of-the-art methods. Conclusion: This paper introduced a new Grey relational analysis-based bag-of-features which improves the efficiency of vector quantization step of the standard bag-of-features method. The method used Grey relational analysis for similarity measure in vector quantization method of bag-offeatures. The proposed method has been validated in terms of precision, recall, G-mean, F1 score, and radar charts on three datasets, Kidney, Lung, and Spleen of ADL histopathological images.

Background: The whale optimization algorithm is one of the popular meta-heuristic algorithms which has successfully been applied in various application areas such as image analysis and data clustering. However, the slow convergence rate and chances of sticking into the local optima due to improper balance of its exploration and exploitation phases are some of its pitfalls. Therefore, in this paper, a new improved whale optimization algorithm has been proposed. Moreover, the proposed method has been used in bag-of-features method for histopathological image classification. Methods: The new algorithm, improved whale optimization algorithm, modifies the encircling phase of original whale optimization algorithm. The proposed algorithm has been used to cluster the extracted features for finding the relevant codewords to be used in the bag-of-features method for histopathological image classification. Results: The efficiency of proposed algorithm has been analyzed on 23 benchmark functions in terms of mean fitness, standard deviation values, and convergence behavior. The performance of the improved whale optimization algorithm based histopathological image classification method has been analyzed on blue histology image dataset and compared with other meta-heuristic based bagof- features methods in terms of recall, precision, F-measure, and accuracy. The experimental results validate that the proposed method outperforms the considered state-of-the-art methods and attains 12% increase in the histopathological image classification accuracy. Conclusion: In this paper, a new improved whale optimization algorithm has been proposed and applied in bag-of-features method for histopathological image classification. The results of proposed method outperform the other existing meta-heuristic methods over standard benchmark functions and histopathological image dataset.

Background: To realize a viable and resilient smart city-smart nation scenario, “peoplecentric” strategic technology components are imperative to eventually create smart outcomes for citizens. Smart health is one such domain where the government is putting incessant effort to ensure social well-being and sustainability. Contemplation of public opinion plays a very significant role in the process of government policy evaluation. The current affordable, ubiquitous generation of Web provides substantial amount of opinionated social big data which facilitates data-driven decision making. But determining the polarity of short-text, a.k.a. sentiment is hard owing to the noisy, ambiguous and heterogeneous use of natural language. Objective: A novel health governance framework using a Swarm Optimized Opinion Prediction model, SOOP Model is proposed to capture netizen views on government policies and figure out the inclination of public about the campaign. The model is investigated for the sentiment classification task on tweets pertaining to ‘Poshan Abhiyaan’, one of the latest healthcare policy, launched by the Government of India to address the issues related to malnutrition in women and children. Methods: The conventional feature extraction using TF-IDF (Term Frequency-Inverse Document Frequency) is done on the pre-processed dataset. Subsequently, Binary bat algorithm, a swarm-based optimal feature selection method is used to improve the sentiment classifier’s performance. Results: The empirical analysis validates that the proposed implementation of the meta-heuristic algorithm for feature subset selection outperforms the baseline supervised learning algorithms with an average 9.4% improvement in accuracy and approximately 39% average reduction in features. Conclusion: The proposed SOOP Model as a policy evaluation strategy within the healthcare setting empowers various stakeholders and enhances their socio-economic environment.

Background: Cervical Cancer is one of the leading causes of deaths among women in India. Accurate and early detection of cancer seems to be a fruitful approach in the diagnosis process. It will be a boon for the medical industry. Prediction of cervical cancer using all the features takes a lot of time and computational resources. Hence, reducing the features and taking only essential features into consideration is an effective solution. Objective: The aim of the research is to identify the relevant features in the classification of cancer and optimize the model. Feature selection increases the accuracy percentage of any classifier. The binary cuckoo search optimization algorithm was applied to explore the important features in the attribute list. Methods: In our research, the performance of the proposed framework has been verified via instigating it with base classifiers such as Random Forest, kernel SVM, Decision Tree and kNN and then evaluated the results with and without Binary Cuckoo Optimization (BCO). The proposed method involves cuckoo search algorithm for selection of optimal feature split points. Cuckoo Search Optimization is a nature stimulated and breeding process of the cuckoo bird’s algorithm to predict best global solution. Results: The results produced only selected features vital for prediction of cancer. In addition, its performance has been paralleled against other factors such as Accuracy, Precision, Recall and Specificity and F-measure. Conclusion: The experimental results show that Decision Tree classifier outperforms all other classifiers with an accuracy of 94.7% increased to 97% after Cuckoo Optimization.

Background: The data proliferation has been resulted in large-scale, high dimensional data and brings new challenges for feature selection in handwriting recognition problems. The practical challenges like the large variability and ambiguities present in the individual’s handwriting style demand an optimal feature selection algorithm that would be capable to enhance the recognition accuracy of handwriting recognition system with reduced training efforts and computational cost. Objective: This paper gives emphasis on the feature selection process and proposed a genetic algorithm based feature selection technique for handwritten digit recognition. Methods: A hybrid feature set of statistical and geometrical features is developed in order to get the effective feature set consist of local and global characteristics of sample digits. The method utilizes a genetic algorithm based feature selection for selecting best distinguishable features and k-nearest neighbour for evaluating the fitness of features of handwritten digit dataset. Results: The experiments are carried out on standard The Chars74K handwritten digit dataset and reported a 66% reduction in the original feature set without sacrificing the recognition accuracy. Conclusion: The experiment results show the effectiveness of the proposed approach.

Background: Breast cancer is one of the most common forms of cancers among women and the leading cause of death among them. Countries like United States, England and Canada have reported a high number of breast cancer patients every year and this number is continuously increasing due to detection at later stages. Hence, it is very important to create awareness among women and develop such algorithms which help to detect malignant cancer. Several research studies have been conducted to analyze the breast cancer data. Objective: This paper presents an effective method in predicting breast cancer and its stage and will also analyze the performance of different supervised learning algorithms such as Random Classifier, Chi2 Square test used in order to predict. The paper focuses on the three important aspects such as the feature selection, the corresponding data visualisation and finally making a prediction call on different machine learning models. Methods: The dataset used for this work is breast cancer Wisconsin data taken from UCI library. The dataset has been used to show the different 32 features which are all important and how it can be achieved using data visualisation. Secondly, after the feature selection, different machine learning models have been applied. Conclusion: The machine learning models involved are namely Support Vector Machine (SVM), KNearest Neighbour (KNN), Random Forest, Principal Component Analysis (PCA), Neural Network using Perceptron (NNP). This has been done to check which type of model is better under what conditions. At different stages several charts have been plotted and eliminated based on relative comparison. Results have shown that Random Tree classifier along with Chi2 Square proves to be an efficient one.

Background: An efficient feature selection method for Histopathological image classification plays an important role to eliminate irrelevant and redundant features. Therefore, this paper proposes a new levy flight salp swarm optimizer based feature selection method. Methods: The proposed levy flight salp swarm optimizer based feature selection method uses the levy flight steps for each follower salp to deviate them from local optima. The best solution returns the relevant and non-redundant features, which are fed to different classifiers for efficient and robust image classification. Results: The efficiency of the proposed levy flight salp swarm optimizer has been verified on 20 benchmark functions. The anticipated scheme beats the other considered meta-heuristic approaches. Furthermore, the anticipated feature selection method has shown better reduction in SURF features than other considered methods and performed well for histopathological image classification. Conclusion: This paper proposes an efficient levy flight salp Swarm Optimizer by modifying the step size of follower salp. The proposed modification reduces the chances of sticking into local optima. Furthermore, levy flight salp Swarm Optimizer has been utilized in the selection of optimum features from SURF features for the histopathological image classification. The simulation results validate that proposed method provides optimal values and high classification performance in comparison to other methods.

Background: Waste management is an essential process for the progress of any nation. The exponential growth in the urbanization and industrialization has brought the waste management issue into prime focus. The practices adopted for waste management vary across the nations as well as regions and sectors. Solid waste management encompasses a number of activities ranging from waste collection to waste recycling and waste reuse. These activities can make use of the technology for enhancing the throughput at each level. Objective: This paper provides a review of the activities involved and technology employed at each step of the solid waste management process. Methods: A phase-wise discussion of the activities involved in solid waste management cycle is provided along with the current methods in practice for each activity. The paper further provides a technical review of technology used for improving the waste management scenario. A brief discussion of the methods for reducing waste generation and increasing reuse is also presented. Finally, the paper identifies a list of challenges related to the waste management process and provides suitable suggestions for addressing the identified challenges. Conclusion: This work shall help the researchers to gain valuable insight into the challenges involved in solid waste management practices and would guide future research regarding the employment of technology for improving the efficiency of the overall waste management process.

Background: Voice over Internet Protocol (VoIP) has emerged as one of the most significant technology in the field of communication and evolved as a substitute to the conventional communication method as the Public Switched Telephone Network (PSTN). Along with the advantages such as scalability and security, VoIP has some threats such as voice quality and interference that must be dealt with. The voice quality in VoIP is degraded when transmitted over a computer network due to delay, jitter and packet loss etc. Packet loss is one of major reasons for the signal quality degradation. Objective: In this research article, Quadrature Mirror Filter Bank (QMF) has been implemented in wireless VoIP system to enhance the quality of the signals transmitted. Results: The performance has been evaluated under varying network conditions of packet loss. Conclusion: Significant improvement has been observed in the quality of VoIP signal.

Background: In today’s world it is highly difficult to manage the smart things and fulfill the communication needs without the Internet as it provides ultimate means for human to human (H2H) communication. The ‘things’ could be entities or devices that contribute for the communication. But to enhance and improve such smart communication among the things that involves nonhuman intervention, there is a need to add few more smart capabilities to the Internet. As the world of Internet is on its way of transformation into a new smart world called ‘Internet of Things’ (IoT) where the things should possess the ability of sensing, communication and control to let the things exchange information without complete intervention of humans to provide advanced and qualitative services, which is possible with the help of protocols. Objective: This paper discusses how the transition could be started and being progressed in wired and wireless systems, and how it changes the traditional means of communication among different devices and humans into a smart way. Methods: Transition here refers to how isolated things are being intermingled with each other to generate a smart protocol communication. Results and Conclusion: The aim of such transition is to improve the efficiency, flexibility, adaptability and interoperability. This paper also explores various factors that contribute to IoT.