diagnosis of cancer using artificial neural network and cloud ...

WORLD JOURNAL OF PHARMACY AND PHARMACEUTICAL SCIENCES Sandhu et al.

World Journal of Pharmacy and Pharmaceutical Sciences SJIF Impact Factor 2.786

SJIF Impact Factor 2.786

Volume 3, Issue 6, 1533-1548.

Review Article

ISSN 2278 – 4357

DIAGNOSIS OF CANCER USING ARTIFICIAL NEURAL NETWORK AND CLOUD COMPUTING APPROACH Ishwinder Kaur Sandhu1*, Meera Nair2, Ravindra Prasad Aharwal3 and Sardul Singh Sandhu3 1

Department of Computer Science and Engineering, Oriental College of Technology, Patel Nagar, Raisen Raod, Bhopal, M.P., India.

2

Centre for Scientific Research & Development, People’s University, Bhopal, M.P., India. 3

Department of Biological Sciences, R.D. University, Jabalpur, M.P., India

Article Received on 16 April 2014, Revised on 07 May, 2014, Accepted on 22 May 2014

ABSTRACT Diagnosis of cancer using a neural network approach engaging cloud computing has become one of the widest used technique for diagnosis of cancer. Generally, cloud computing facilitates data protection, privacy and medical record access whereas neural network judges the

*Correspondence for Author Ishwinder Kaur Sandhu Department of Computer

possibility rate of tumors in 960 of 1008 cases by using data from cancer positive patients (via tumor size, number of nodules, tumor

Science and Engineering,

hormone receptor status, etc.). User Interface Medical Services

Oriental College of

(UIMS) are the new developmental framework proposed for cancer

Technology, Patel Nagar,

diagnosis wherein neural networks and cloud computing enhances the

Raisen Raod, Bhopal, M.P., India

efficiency and accuracy for diagnosis. In the present paper, the use of neural networks in diagnosis of cancer has been described with special

reference to breast cancer detection using a cloud computing model. The paper also provides an overview of the different types of software’s used in cloud computing and how the utilization of neural network structures improves the medication and diagnosis of cancer in early stages. Keywords: Neural Networks; Artificial Neural Networks; Cloud Computing; UIMS. INTRODUCTION Neural Network (NN), a mimic of BNN (Biological Neural Network), is a massively parallel distributed processing system which is made up of highly interconnected neural computing

www.wjpps.com

Vol 3, Issue 6, 2014.

1533

Sandhu et al.

World Journal of Pharmacy and Pharmaceutical Sciences

elements that have the ability to learn and thereby acquire knowledge and make it available for use.[1] Neural Networks (NNs) are simplified imitations of the central nervous system, and are aggravated by the kind of computing performed by the human brains. Neurons are the structural entities of human brain and perform computations such as cognition, logical inference, pattern recognition and so on, where Neural networks (NNs) are simplified models of the biological nervous systems.[2] Hence the technology, which is built on such a simplified imitation of neurons is termed as Artificial Neural System (ANS) technology or Artificial Neural Networks (ANN) or simply Neural Networks and neurons. An ANN is configured for a specific application, such as pattern recognition or data classification, through a learning process where learning involves adjustments to the synaptic connections that exist between the neurons. This is well explained via a schematic model (Fig. 1) which shows the behavior of a neuron where every component depicts the analogy to the actual constituents of a biological neuron and provides a basis to Artificial Neural Networks.

Inputs

Weights

x1

W1j

x2

W2j

x3 xn

Net inputs Net j ∑

W3j Wnj

Transfer function

Activation function

β

Oj Activation

Oj Threshold

Fig. 1 Schematic of ANN Neural Networks are currently a hot research area in medicine. It has a huge application in many areas such as education, business, medical, engineering and manufacturing. Neural Network plays an important role in a decision support system. A new development

www.wjpps.com

Vol 3, Issue 6, 2014.

1534

Sandhu et al.


framework for cloud computing called User Interface Medical Services (UIMS) is formulated.[3] An attempt has been made to make use of neural networks in the medical field (carcinogenesis (pre-clinical study). In carcinogenesis, artificial neural networks have been successfully applied to the problems in both pre-clinical and post-clinical diagnosis.[4] Cancer diagnosis is carried out using neural networks and the implementation of cloud computing enhance the efficiency and accuracy of diagnosis.[3] Neural networks are important tools for cancer detection and monitoring. Kenji Suzuki,[5] investigated a pattern-recognition technique based on an artificial neural network (ANN). The ability of the physicians to effectively treat and cure cancer is directly dependent on their ability to detect cancers at their earliest stages. An initial diagnosis called early diagnosis is made based on the demographic and clinical data of the patient via which more than 30% cancer deaths are preventable. Artificial neural networks offer a completely different approach to problem solving and they are sometimes called the sixth generation of computing. Over the last two decades, a tremendous amount of research work has been conducted for automated cancer diagnosis. Gutte and Henrik,[6] developed a completely automated method based on image processing techniques and artificial neural networks for the interpretation of combined fluorodeoxyglucose (FDG) positron emission tomography (PET), computed tomography (CT) and natural products (NP) images for the diagnosis and staging of cancer. Many natural bioactive compounds such as paclitaxel,[7] Cordicepin,[8] vincristine & vinblastine etc.[9][10] are used for anticancer drugs in present time. Neural Networks Architecture For many years, artificial neural networks have been studied so to achieve human – like performance in the fields of speech and image recognition. Different Neural network architectures are widely described in the literature: 1. Feed forward Network: Feed-forward ANNs allow signals to travel only in one direction i.e. from input to output, there is no feedback (loops) i.e. the output of any layer does not affect that same layer. Feed-forward ANNs tend to be straight forward networks that associate inputs with outputs and are extensively used in pattern recognition. This type of organization is also referred to as bottom-up or top-down or three layer network. The input layer neurons receive the input signals and output layer neurons receives an output signal. Thus, it is also known as Single Layer Feed forward Network and is acyclic in nature (Figure 2).

www.wjpps.com

Vol 3, Issue 6, 2014.

1535

Sandhu et al.


Input layer

Hidden layer

Output layer

Weights, Wij 1

2

2

1

3

3

2

4

4

Inputs

Outputs

1

Fig. 2 Feed-Forward Networks

Input Layer

Hidden Layer

X1

H1

Output Layer Z1

X2

X3

Z2

Y1

Y2

H2 X4

Fig. 3 Recurrent Network 2. Recurrent Network: These networks differ from feed-forward network architecture as it contains only one feedback loop, so that activation can flow round in a loop (Fig. 3). Thus, existing only one layer with feedback connections. These networks enable temporal processing and learn sequences (e.g. perform sequence recognition/reproduction or temporal association/prediction). There could also be neurons with self-feedback links, i.e. the output of a neuron is fed back into itself as input.

www.wjpps.com

Vol 3, Issue 6, 2014.

1536

Sandhu et al.


x1

V11

V1m

z1

w11

y1

w12

Z2

V21

x2

w13

V2m

Z3 ym

V11

w1n

V1m

X1

Zn

Input layer

Hidden layer

X1: Input neurons y1: Hidden neurons zk: Output neuron v8: Input hidden layer weight wA: Output hidden layer weight

Output layer

Fig. 4 Multilayer Feed forward Network 3.Multilayer Feed-forward Network: This network is made up of multiple layers where irrespective of an input and output layer, one or more intermediary layer called as hidden layers (i.e. hidden neurons) are present. This hidden layer aid in performing an intermediary computation before directing the input to the output layers (Fig.4). Characteristics of Neural Networks (NNs) Neural Networks are developed via human cognition through biological neurons, performing similarly as

a human brain hence, the characteristic includes the ability for storing

knowledge and making it available for use whenever necessary, propensity to identify patterns, even in the presence of noise, aptitude for taking past experiences into consideration and make inferences and judgments about new situations.[11] 1. The NNs can map input patterns to their associated output patterns thereby exhibiting mapping capabilities. 2. NNs are trained with known examples of a problem thus identifying new objects previously untrained. 3. The NNs possess the capability to generalize thereby predicting new outcomes from past trends. 4. The NNs are robust systems and are fault tolerant as they can recall full pattern from incomplete, partial or noisy patterns. 5. The NNs can process information in parallels, at high speed, and in a distributed manner. www.wjpps.com

Vol 3, Issue 6, 2014.

1537

Sandhu et al.


Cloud Computing Cloud computing (‘Cloud’) is an evolving term used to describe the development of many existing technologies and approaches to computing into something different. A Cloud refers to a distinct IT environment that is designed for the purpose of remote provisioning scalable and measured IT resources. The term is generally originates as a metaphor for the Internet which is a network of networks providing remote access to a set of decentralized IT resources.[12] Cloud separates applications and information resources from the underlying infrastructure, and the mechanisms used to deliver them thereby, enhancing collaboration, agility, scaling and availability. This, in turn, cuts the cost through optimized and efficient computing. Cloud Service Models Cloud service delivery is divided among three archetypal models and various derivative combinations. The three fundamental classifications are often referred to as the “SPI Model,” where, ‘SPI’ refers to Software, Platform or Infrastructure (as a Service), respectively. Each models share similarities, but have their own distinct differences as well. These service models are Infrastructure-as-a-Service, Software-as-a-Service and Platform-as-a-Service (Fig. 5.). Software as a Service (SaaS) SaaS is a software delivery method that provides access to software and its functions remotely as a web-based service. It is generally a software delivery model in which software and associated data are centrally hosted on the cloud by independent software vendors.[13] The applications are accessible from various client devices through a thin client interface such as a web browser (e.g., web-based email) i.e. designed for end-users, delivered over the web. The consumer does not manage or control the underlying cloud infrastructure including network, servers, operating systems, storage, or even individual application capabilities, with the possible exception of limited user-specific application configuration settings.[14][15]

www.wjpps.com

Vol 3, Issue 6, 2014.

1538

Sandhu et al.


Servers

Application

Desktops

Monitoring Content

Laptops

Collaboration Communication Finance

Platform

Identity

Object Storage

Runtime

Database

Queue

Infrastructure Network Tablets

Computer

Block Storage Phones

Cloud Computing Fig. 5 Cloud computing stack Characteristics of SaaS  Web access to commercial software.  Software managed from a central location.  Software delivered in a “one too many” model.  Users are not required to handle software upgrades and patches.  Integration between different pieces of software via Application Programming Interfaces (APIs). Platform as a Service (PaaS) The capability provided to the consumer is to deploy onto the cloud infrastructure consumercreated or acquired applications created using programming languages and tools supported by the provider. The consumer does not have to manage or control the underlying cloud infrastructure including network, servers, operating systems, or storage, but has to control over the deployed applications and possibly the applications hosting environment

www.wjpps.com

Vol 3, Issue 6, 2014.

1539

Sandhu et al.


configurations. PaaS is analogous to SaaS except that, it is a platform used for creating a software and is delivered over the web. [14][15] Characteristics of PaaS  Services to develop, test, deploy, host and maintain applications in the same integrated development environment.  Web based user interface creation tools which help to create, modify, test and deploy different UI scenarios.  Multi-tenant architecture i.e. multiple concurrent users utilize the same development application.  Built in scalability of deployed software including load balancing and failover.  Integration with web services and databases via common standards.  Support for development team collaboration.  Tools to handle billing and subscription management. Infrastructure as a Service (IaaS) It is a way of delivering Cloud Computing infrastructure – servers, storage, network and operating systems-as an on-demand service. Rather than purchasing servers, software, data center space or network equipment, clients instead buy those resources as a fully outsourced service on demand. The consumer does not manage or control the underlying cloud infrastructure but has control over operating systems; storage, deployed applications, and possibly limited control of select networking components (e.g., host firewalls).[16][12] Characteristics of IaaS  Resources are distributed as a service.  Dynamic scaling.  Has a variable cost, utility pricing model.  Generally includes multiple users on a single piece of hardware. Cancer Cancer as the term defines is a class of disease characterized by an out-of-control cell growth. There are over 100 different types of cancer, and each is classified by the type of cell that is initially affected.[17] Cancer starts when cells in a part of the body start to grow out of control in comparison to the normal cell growth. Instead of dying, cancer cells continue to grow and form new, abnormal cells. This equilibrium between genesis and destruction of cells is

www.wjpps.com

Vol 3, Issue 6, 2014.

1540

Sandhu et al.


disturbed in cancer. There are two types of cancer: benign and metastatic where the benign cancer lacks the ability to invade neighboring tissues where as metastatic cancer travel to other parts of the body via the body’s bloodstream or lymph vessels and form new tumors. Over time, these tumors replace normal tissue, crowd it, or push it aside. The process of cancer spreading is called metastasis.[18] Most cancers are named from its initiation viz. lung cancer starts in the lung, and breast cancer starts in the breast. Symptoms and treatment depend on the cancer type and how advanced it is. Most treatment plans may include surgery, radiation and/or chemotherapy where as some may involve hormone therapy, biologic therapy, or stem cell transplantation.[19] Causes of Cancer There is no one single reason for cancer where an integrated approach leads to cancer in an individual the factors involved may be genetic, environmental, or constitutional characteristics of the individual. The following risk factors and mechanisms have been proposed as contributing to cancer:  Lifestyle factors.  Family history, inheritance, and genetics  Some genetic disorders.  Exposures to certain viruses. Epstein-Barr virus and HIV,  Environmental exposures.  Some forms of high-dose chemotherapy and radiation. The fundamental cause of cancer is damaged or faulty genes which are a set of instructions that tells cells to what to do. Genes are encoded within the DNA, so anything that damages DNA can increase the risk of cancer. There are mainly three main types of genes that are involved in the cell growth, and when altered (mutated) lead to certain types of cancers,[20] these are 1.Oncogenes: These genes regulate the normal growth of cells and a sudden flip of the switch may cause an uncontrollable growth of cells. 2.Tumor suppressor genes: These genes are able to recognize abnormal growth and reproduction of damaged cells, or cancer cells, and can interrupt their reproduction until the

www.wjpps.com

Vol 3, Issue 6, 2014.

1541

Sandhu et al.


defect is corrected. If the tumor suppressor genes are mutated, however, and they do not function properly, tumor growth may occur. 3.Mismatch-repair genes. These genes help recognize errors when DNA is copied to make a new cell. If the DNA does not "match" perfectly, these genes repair the mismatch and correct the error. If these genes are not working properly, however, errors in DNA can be transmitted to new cells, causing them to be damaged. Cancer Diagnosis Using Neural Networks An artificial neural network (ANN) is a computational structure model with neural structure similar to the human brain where both have many highly interconnected processing elements. Numerous studies have shown development of a system for diagnosis, prognosis and prediction of cancer using Artificial Neural Network (ANN) models.[21] Khan et al.[22] developed a method for classifying cancers into specific diagnostic categories based on their gene expression signatures using artificial neural networks (ANNs) using small, round bluecell tumors (SRBCTs) as a model. ANN research mainly aims to provide a filter between the cases and thereby distinguish the cancers hence reducing the cost of medication and helping doctors to focus on cancer-ridden patients. In addition, ANNs provide a valuable tool that could minimize the disagreement and inconsistencies in mammographic interpretation.[23] 1. Training the model Once a network has been structured for a particular application, that network is ready to be trained. To start this process the initial weights are chosen randomly. Then, the training, or learning, begins. The ANN is trained by exposing it to a set of existing data (based on the follow up history of cancer patients) where the outcome is known. Multilayer networks use a variety of learning techniques; the most popular is back–propagation (BP) algorithm.[24] It is one of the most effective approaches to machine learning algorithm information which flows from the direction of the input layer towards the output layer. There are two approaches to training - supervised and unsupervised. Supervised training involves a mechanism of providing the network with the desired output either by manually "grading" the network's performance or by providing the desired output with the inputs. Unsupervised training is where the network has to make sense of the inputs without outside help.[25]

www.wjpps.com

Vol 3, Issue 6, 2014.

1542

Sandhu et al.


Training in ANN’s is achieved via examples adjusting the connection weights in ANN’s iteratively. The number of iterations of the training algorithm and the convergence time varies depending on the weight initialization. After the repetition of the processes, for a sufficiently large number of training cycles, the network usually converges to a state where the error in the calculations is small thus implying the network to be learned to a certain target function. 2. Selection of weights Initially, the weights on all the interconnections are set at small random numbers,[26] and the network is said to be “untrained”. The weights of each neuron (node) were randomly initialized to values between -1 and +1. Weight training in ANN’s is usually formulated as minimization of an error function [xin] has the mean square error between target and actual outputs averaged over all examples, by iteratively adjusting connection weights. The weights are adjusted in such a way that each weight adjustment increases the likelihood that the network will compute. To adjust, weights are calculated and the weights are then changed such that the error is decreased (thus going downhill on the surface of the error function). The probability of successful convergence depends on the weight initialization scheme. For this reason, back propagation can only be applied on networks with differentiable activation functions. 3.Data description and Training data The data obtained from different hospitals are used for studying various cancers where, a programming language (MATLAB) is used for scientific and engineering numeric computation and visualization.[4] Cancer Diagnosis Using Cloud Computing Cloud Computing provides a new approach towards detection of cancer and signifies it as one of the major discoveries of the 21st century using ANNs. Cloud assists medical diagnosis using effective tools to provide a more consistent and reliable care. An artificial neural network detects patterns too complex to be recognized by humans and thus is utilized for diagnosis and has been recently applied to breast mass malignancy classification when evaluating by Fine Needle Aspirates (FNAs).

www.wjpps.com

Vol 3, Issue 6, 2014.

1543

Sandhu et al.

Logical Inputs

Clump Thickness

Cell Size Uniformity

Cell Shape Uniformity Single Epithelial Cell Size Normal Nucleoli

Bare Nuclei

Marginal Adhesion

Bland Chromatin

Mitoses


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Artificial Input Layer

Hidden Layer

Node 1

Node 2

Output Layer Node 3 Diagnosis Node 4

Node 5

Node 6

Fig. 6 Neural Model A cloud-based platform thus helps a health care company deliver information about breast cancer techniques and best practices to health care providers and the public.[27] Cloud-based computer program is able to find patterns in patient's genetic expression to diagnose an aggressive form of cancer, called mixed-lineage leukemia.[28] There are many companies providing medical solutions as a service, though there are many other web based solutions targeting medical and health care field such as EMR ”Electronic Medical Record”; EHR, ”Electronic Health Record”; PHR, “Personal Health Record”. Some large corporations also provide services in healthcare area such as Google and Microsoft.

www.wjpps.com

Vol 3, Issue 6, 2014.

1544

Sandhu et al.


Most of these services and applications are classified under SaaS “Software as a Service”, as there is no Platform built to target medical or health care area under the classification of PaaS. In addition, many web based applications and services, along with many open source web-based medical and healthcare solutions. Cloud services and platform provide the collaboration, sharing and also increase the productivity in team work. The Cloud Computing model performs "intelligent" tasks similar to those performed by the human brain. Neutral Network

User

Hidden Layer

Outer Layer

Cloud Service

Sigmoid Activated Function

Summation Function

Summation Function

Summation Function

Sigmoid Activated Function

Step Function

Diagnosis

Diagnosis

Fig.7. Algorithm Map The best example is the Cloud4Cancer service which identifies 99 percent of malignant breast tumors and thus can be used to alter or improve diagnostics for multiple cancers. Global Neural Network Cloud Service for breast cancer achieves a consistent diagnostic tool for the global medical community and also provides a platform to build global knowledge and improve success rates.[29] This seems to be the best approach towards diagnosis as ANNs learn by examples therefore success rate directly correlates to training set size. The neural model utilized by Cloud4Cancer is depicted in a schematic diagram (Fig. 6) and an algorithm map (Fig 7.) Thus, Global Neural Network Cloud service for breast cancer provides a ready to diagnose actual patients. These prove useful for medical diagnostics and achieve maximum perfection with an increase in samples.[29] A new framework for cloud computing called User Interface Medical Services (UIMS) is devised wherein; UIMS is an interactive decision support system (DSS) Computer Software,

www.wjpps.com

Vol 3, Issue 6, 2014.

1545

Sandhu et al.


which is designed to assist physicians and other health professionals with decision making tasks, as determining diagnosis of patient data.[30] Hence, UIMS is used to achieve clinical advice for patient care based on some number of items of patient data. In recent times, the diagnosis of cancer is carried out using ANN and the implementation in a cloud environment enhances the efficiency and accuracy of diagnosis. CONCLUSION Cloud computing service for neural networks provides inputs in different layered perceptions without maintaining any software’s. With the utilization of Cloud Computing in diagnosis the time required for diagnosis would depreciate. People would be checked for cancer quickly and painlessly thereby detecting the disease at an early stage. Thus, indicating neural networks as an effective option for cancer diagnosis so to help clinicians and oncologists in the prediction and prognosis of cancer. ACKNOWLEDGEMENT The authors wish to thank the Vice Chancellor Prof. K.N. Singh Yadav, R.D. University, Jabalpur, India and the Head of the Department of Biological Science, R.D. University, Prof. Y.K. Bansal for providing laboratory facility for this project, and IKS thanks Prof. Rachana Mishra, Head, Department of Computer Science and Engineering, Oriental College of Technology, Bhopal for constructive criticisms of the manuscript and encouragement. MN is thankful to Director HR/IT Logistics People’s Group, Bhopal, for providing requisite facilities. Conflict of Interest The authors declare there is no conflict of interest. REFERENCES 1. Kaur K. Optical Multistage Interconnection Networks Using Neural Network Approach. Master of Engineering. (2009) Thapar University. 2. Rajasekaran S, Pai GAV. Neural Networks, Fuzzy Logic, and Genetic Algorithms: Synthesis and Aplications. PHI Learning Pvt. Ltd. 2003 3. Rajkumar B, Gopikiran T, Satyanarayana S. Neural Network Design in Cloud Computing. Int J Comp Tren Techn, 2013; 4(2): 63-67.

www.wjpps.com

Vol 3, Issue 6, 2014.

1546

Sandhu et al.


4. Ganesan N, Venkatesh K, Rama MA, Palani AM. Application of Neural Networks in Diagnosing Cancer Disease Using Demographic Data. Int J Comp Appli, 2010; 1(26): 7685. 5. Suzuki K, Samuel G, Armato III, Feng L, Shusuke S and Kunio D. Massive training artificial neural network (MTANN) for reduction of false positives in computerized detection of lung nodules in low-dose computed tomography- Kurt Rossmann Laboratories for Radiologic Image Research, Department of Radiology, The University of Chicago, Chicago, Illinois 60637-Medical Physics, 2003; 30(7). 6. Gutte H, Jakobsson DO, lofsson F, Ohlsson M, Valind S, Loft A, Edenbrandt L, Kjær A. Automated interpretation of PET/CT images in patients with lung cancer”- Nuclear Medicine Communications, 2007; 28(2). 7. Stierle A, Strobel G, Stierle D. Taxol and taxane production by Taxomyces andreanae, an endophytic fungus of Pacific yew. Science 1993; 260: 214-216. 8. Tuli HS, Anil S, Sardul S S, Dharambir K. (2013) Cordycepin: A bioactive metabolite with therapeutic potential. Lif Sci, 93: 863-869. 9. Parekh J, Chanda S. Screening of aqueous and alcoholic extract of some Indian medicinal plants anti-bacterial activity, 2006; 68(6): 835-838. 10. Aharwal RP, Shukla H, Sandhu SS.. Therapeutic Potential of Catharanthus roseaus Linn. Current Trends in Biotechnol Chem Res, 2013; 3(1): 63-70. 11. Razi MA, Athappilly K. A comparative predictive analysis of neural networks (NNs), nonlinear regression and classification and regression tree (CART) models. Expert Systems with Applications, 2005; 29: 65–74. 12. http://www.whatiscloud.com/basic_concepts_and_terminology/cloud. 13. http://www.SaaS Business Model Canvas". DeckPresenter. June 2013. Retrieved 7 June 2014. 14. http://www.en.wikipedia.org/wiki/Platform_as_a_service 15. http://www. java.dzone.com/articles/what-platform service-paas. 16. http://www.diversity.net.nz/wp-content/uploads/2011/01/Moving-to-the-Clouds.pdf. 17. http://www.medicalnewstoday.com/info/cancer-oncology. 18. http://www.cancer.org/treatment/understandingyourdiagnosis/advancedcancer/advancedcancer-what-is-metastatic. 19. Nair M, Jain R, Saxena P. Reality Check: Cancer Stem Cell Route to Cancer. Cur Biotechnol, 2013; (17): 89-105. 20. http://www.cancer.stanford.edu/. www.wjpps.com

Vol 3, Issue 6, 2014.

1547

Sandhu et al.


21. Janghel RR, Shukla A, Tiwari R, Kala R. Breast cancer diagnosis using Artificial Neural Network models. Information Sciences and Interaction Sciences (ICIS), 2010; pp. 89 – 94. 22. Khan J, Wei JS, Ringnér M, Saal LH, Ladanyi M, Westermann F, Berthold F, Schwab M, Antonescu CR, Peterson C, Meltzer PS. Classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks, Nat Med. 2001; 7(6):673-9. 23. Abbass

HA.

An

evolutionary artificial neural

networks approach

for

breast

cancer diagnosis. Artif Intell Med, 2002; 25(3): 265-81. 24. Rumelhart DE, Hinton GE, Williams RJ. Learning internal representations by error propagation. In McClelland JL, Rumelhart DE and the PDP Research Group Eds, editors, Parallel Distributed Processing: Explorations in the Microstructure of Cognition., Foundations, MIT Press, Cambridge, pages 318–362, 1986. 25. http://www.psych.utoronto.ca/users/reingold/courses/ai/cache/neural3.html. 26. Carlos H, Mercedes F. Universidad Jaume I. Campus de Riu Sec, Spain, IEEE 2001. 27. http://www.baselinemag.com/cloud-computing/the-cloud-helps-in-the-fight-against cancer.html#sthash.Cksv7Sj3. dpuf. 28. http://www.dailymail.co.uk/femail/article-2335738/Is-future-cancer-diagnosis-Highschool-student-creates-program-detect-leukemia-things-transcend-humanknowledge. html. 29. http://www.cloud4cancer.appspot.com. 30. Mohana RS, Thangaraj P. Machine learning approaches in improving service level agreement-based admission control for a software-as-a-service provider in cloud. Journal of Computer Science, 2013; 9 (10): 1283-1294.

www.wjpps.com

Vol 3, Issue 6, 2014.

1548