Sakthi A and Dr.M.Rajaram 45 Enhanced Image Processing with Digitized Microscopic Images a Foresight in Indian Context Sakthi A and Dr.M.Rajaram Abstract: The healthcare environment is generally information rich but knowledge poor.[1] There is a wealth of data available within the healthcare systems but they lack effective analysis tools to discover hidden relationships and trends in data, one such stream is automating microscopic stained slide analysis. Technically Advanced countries are extensively introducing modern medical devices to diagnose disease from patient s blood or tissue. But as a matter of fact still classical microscope is used in most of the laboratories and Hospitals of India and many other backward countries. [2] Manually performed test results with classical microscope is inaccurate, cumbersome, and proof less and determination is crucial for the clinicians, demanding for highly specialized personnel who are not always available. This paper presents an innovative approach to capture images seen from any classical microscope with the aid of qualitative wireless camera connected with computer, and these images are exposed to the Service Oriented Architecture (SOA) server [3] in the Internet by invoking web services. A component Knowledge Base (KB) has the repository of various cell patterns, and image processing rules. These inputs are used as benchmark to process the queued up images from Clients. Diagnostic Expert (DE) has the experts inputs from the medical domain. Clients can also query the SOA for FAQ, suggestions, expert s comments. Once the image is processed once again web services return the accurate results to the intended clients I. Introduction Many diagnostic centers, healthcare service providers are widely spread across the country. But the quality and accuracy of diagnostic results are not up to the mark and the Quality of Service (QoS) is not satisfactory. Quite a number of service providers do not have advanced medical equipments and they fail to provide accurate laboratory test results, the act of false results sometime deviate the treatment procedures to a worse situation. Manual Microscopic diagnosis is one such area where due to human errors the actual results are either not found or missed out. Presently this happens in many small scale laboratories, clinics, and even in middle level hospitals. Manual observation of microscopic slides such as counting of blood cells, identifying peculiar diseases, analysis of complex disease cell patterns do require special skill and experience. It is important that only the results provided by the lab incharge or the investigator plays a pivotal role in taking decision towards treatment. But always the results observed manually are not accurate as such it is not completely reliable. Sakthi A is a Research scholar, Anna University of Tech, Coimbatore, and Dr.M.Rajaram is a Professor / EEE, GCT, Coimbatore, Emails: sakthijesma@gmail.com, rajaramgct@rediffmail.com Today countries like India and under developed countries are highly in demand of basic amenities that are essential for clinical services; it might be very difficult to modernize every service related entities abruptly. But we can think of providing a feasible and cost effective solution to improve this plight and by the efficient solution we can mend the road for the upcoming advancements. Already considerable amount of effort is done on the microscopes but the major flaw with the previous works are that they don t produce results based on pattern matching but just provide features to capture slides of images, or as just a video Our solution highlights on cost effective solution by making use of the available resources but with the minimal tweak. Healthcare centers can embed a wireless camera to their existing classical microscope and install the RESult PROvider (RESPRO an open source tool that is available locally for image processing and communicates with server). Once computer receives the images of stained slides they are constantly captured with the help RESPRO either as video clips or as images. These captured images are enhanced; fine tuned and compressed for image analysis. Then these images are sent to SOA, where KB provides the mechanism to analyze the image against the expected tests with help of available benchmarked cell patterns. When image is processed server provides the accurate results and specification criteria taken for producing test results to the client. In critical cases SOA can be queried for the suggestion on the test results provided, in such case Diagnosis Expert (DE) provide more information and suggestion from the repository of previous cases and can provide accurate information on how to precede with the diagnosis results. II. Backgrounds and Motive The distinction between e-healthcare and traditional healthcare is very clear. The later can be termed as a doctor centered approach where the doctor gives consultation after analyzing the available information, such as past history and constraints of the patient based on his knowledge and experience. Hence he is the prime source of advice to the patient regarding his health problem, future care and medical prescription. Whereas e-healthcare is a consumer-centered model of health care in which stakeholders collaborate utilizing information and communication technologies (ICTs) including Internet technologies to manage one s up to date
Sakthi A and Dr.M.Rajaram 46 health information at the finger tips and obtain the relevant treatment with in the right time and in the right place [4]. It is possible to provide the best e-healthcare services in spite of service centers like middle level Indian based hospitals by facilitating proper health records and verifying and updating diagnoses and services related information to patients. One such area is to improve the laboratory test results. By eradicating manual analysis into automation III. Related Works Researchers are very keen at minimizing the effort made into the observation process of the microscopic based investigations. The growth of microscopy is unimaginable in the modern times. Today microscopes are even able to capture the complete information without human intervention. [5] But the cost is too high to acquire such a costly device in the economically deprived healthcare centers. As a matter of fact solutions only addresses mediocre and creamy layer of the society. It is true that previous researchers have found considerable amount of solution only by modernizing the complete scenario while not giving prior importance to the aforementioned problem. In the field of autoimmune diseases existing researches have been done that follows the similar classical methods of human observations, where the availability of accurately performed and correctly reported laboratory determinations is crucial for the clinicians, demanding for highly specialized variability that limits the reproducibility of the method [6, 7].personnel that are not always available. Moreover, the readings are subjected to inter observer so in autoimmune diseases, Indirect Immunofluorescence (IIF) represents the recommended method for detection of antinuclear autoantibodies (ANA). IIF diagnosis requires both the estimation of fluorescence intensity and the description of staining pattern, demanding for highly specialized personnel, In the staining pattern recognition of IIF wells, since several cells constitute each well, They Have developed a Multiple Expert System (MES) devised to classify the pattern of individual cells. The whole well staining pattern is computed on the strength of the recognition of its cells Similar approaches like finding the segregation of cancer cells are also have been initiated in the history of cancer ailments [8]. Microscopes modernizing have simplified the tasks of observation, but it is not used democratically by everyone IV. Work Flow To implement this model, we propose Software as a [9] Service (SaaS) driven client-server architecture model, where the Server acts as service provider and client requests for services. The idea behind the proposal is to make the cell pattern image analysis method in common hosted in internet given free of cost as Service Oriented Architecture (SOA), SOA process the image and provide results as per the diagnosis expected by clients. Diagnostic comments from its history of records is also provided, Experts can define and update the image pattern rules, recent information, comments based on diagnosis, solution for critical cases are provided based comment for any type of biological test conducted services being accessed from clients from any part of the world. Figure 1 explains the flow in detail. Process flow at the Client Blood or Tissue sample is collected from patient It is either stained in the slides or observed as it is A wireless camera that is being fixed to the microscope is connected to the web based software Slide is exposed to the microscope that is attached with the wireless camera Slide is exposed to the microscope that is attached with the wireless camera Wireless camera communicates with the desktop application RESPRO (RESult PROvider (RESPRO) Upon receiving the image from wireless camera it is being fine tuned by RESPRO to the required format for PACS (picture Archiving Communication Server) and exposed for image processing RESPRO process the images from the desktop as well as synchronizes with SOA through web service on need basis for additional functionalities for any peculiar services Once the result is ready, the satisfaction of the result is verified, for any doubts SOA can be queried with regard to the result DE returns the remarks for the abnormal cases if any Process flow at the Server SOA (Service Oriented Architecture) Engine is hosted in the web site for anybody to access it free of cost KB is one of the entities has the repository of image patterns with respect to slide tests, tissue patterns. KB can be update periodically to keep the patterns up to date KB process the image received from client as per the image pattern analysis and test expected by client SOA Web Service Function takes the parameters on the service call such as test name, samples of images and the other specifications if any based on that it produces the results to the client DE is another entity has the repository of diagnosis based results, exceptional cases, and remarks for every test result provided to the client DE also maintains a database of clients and patient record, on any query received from client will be answered based on the information available in the repository PACS (Picture Archiving and Communication Server) is the another entity which manages the images in standard manner When image are manipulated via PACS, it has been well fine tuned. When processing takes places KB receives the image from PACS and operates on it Industry experts, Doctors, web sites, Expert systems, Healthcare software agents can update SOA
Sakthi A and Dr.M.Rajaram 47 functionality as per the recent standards so as to keep the SOA up to date Image retrieval is the technique to find similar images from an image archive by their textual and visual contents and image registration is the establishment of correspondences between images or between image and physical space. [12] The ultimate goal of medical image retrieval system, whether content based image retrieval or text based is to deliver the similar images compared to a query image in a most effective and efficient way. 5.4 Imaging System Modern standards such as Digital Imaging and Communication in Medicine (DICOM) and Picture Archival and Communication Systems (PACS) make it relatively easy to store and transport these images and increase interoperability in SOA architecture as well as in RESPRO software. Figure 1: Service Oriented Architecture V. Tool Architecture and Technology There are many image techniques are used in our framework 5.1 Pattern Recognition Pattern recognition is "the act of taking in raw data and taking an action based on the category of the pattern" [10] Pattern recognition aims to classify data patterns based either on prior knowledge or on statistical information extracted from the patterns. 5.2 Lossless Compression [11] Every communication and archiving of the image results is done using the various lossless compression techniques. The goal of lossless image compression is to generate an absolutely equivalent but shorter representation than the original image. Advantage of lossless compression is that the original image can be recovered - there can therefore be no subsequent claim that important information was lost as a result of the compression process, which could be crucial. In medical images, quality loss can also affect diagnostic accuracy. 5.4 Service Oriented Architecture The design of a Service Oriented Architecture-based platform for medical image processing in assisted diagnosis. Service oriented architecture (SOA) improves the reusability and maintainability of distributed systems. In service oriented architectures, the most important element is the service, a resource provided to remote clients via a service contract. SOA-based systems proven to be a good solution to address problems like large storage and expensive computational requirements faced in medical image processing applications. We propose a component-based platform which is not tied to a specific programming language or a specific technology. Our platform provides a repository that contains components used to create services The term Service Oriented Architecture, SOA for short, contains some important notions. We have the following definitions for these notions Architecture is a formal description of a system, defining its purpose, functions, externally visible properties, and interfaces. It also includes the description of the system s internal components and their relationships, along with the principles governing its design, operation, and evolution. A service is a software component that can be accessed via a network to provide functionality service requester. The term service-oriented architecture refers to a style of building reliable distributed systems that deliver functionality as services, with the additional emphasis on loose coupling between interacting services. Services The service is the core element in SOA. A service is defined as a mechanism to enable access to one or more capabilities, where the access is provided using a prescribed interface and is exercised consistent with constraints and policies as specified by the service description. 5.3 Image Retrieval and Registration
Sakthi A and Dr.M.Rajaram 48 An optional Header element containing generic information. A required Body element containing the request /response data. An optional Fault element that provides information about errors occurred while processing the request. In figure 4 is represented a sample SOAP request message for adding two numbers. The service response is shown in figure5. Note that the HTTP header is not represented in this example and only the required SOAP elements are used. Figure: 2 Web Service Consumer Provider Model Web Services (WS) Standard A web service is a special case of service, processing XML data and using communication protocols like SOAP (Simple Object Access Protocol) and HTTP. Web services provide a well-defined interface that is described by an XML-based document called the Web Service Description Language (WSDL) document (WSDL contract). This document contains the operations (methods) that the service supports, including data type information, and binding information for locating and communicating with the Web service operations. Figure: 4. SOAP Request In Figure 3 depicts the web services architecture. The UDDI (Universal Description, Discovery and Integration) plays the role of a service directory Figure: 5. SOAP Response VI. Experiment & Analysis Figure: 3 Web service Architecture Having collected feedbacks and inputs from more than twenty six hospitals of Tamil Nadu, Kerala, Andhra and Karnataka, and visited the government hospitals in Tamil Nadu and having observed the laboratories, conducted interviews with physicians, we generally got an input that computer s interference with the existing situation can lead to a great change. The human error probabilities notified in the Figure 6 explains under various circumstances how the error ratio can occur. This is more sufficient to look for an automation against the manual observation of slides The communication protocol used with Web services is SOAP. SOAP is a protocol for exchanging XML-based messages over computer networks, normally using HTTP/HTTPS. A SOAP message is an ordinary XML document containing the following elements: A required Envelope element identifying the SOAP message.
Sakthi A and Dr.M.Rajaram 49 with the minimized cost people at lower level can get the best results VIII. Future Direction The service oriented architecture based microscopic image analysis can be further extended to the legally authorized government agencies and fit their need into making it used by many. The ultimate motivation behind this is to reduce the cost while providing best of accurate test results at any health care centre. Along with that even advanced devices such as Scan, X-ray other type of patient images to be incorporated in the centralized agent in order to possibly provide accurate and reliable results from the centralized forum of experts thus the best e-healthcare services can be provided even at the economically lowered groups of Indian like context. Figure 6: Human Error Probabilities Chart 1 described in the following section explains the few healthcare providers satisfaction level on classical microscope usage and the proposal for the automated test environment. The survey was conducted at various levels, right from observation and conducting interviews with the laboratory assistants and chief, who conduct various tests from almost 600 to 750 patients on daily basis Acknowledgements The author thanks Dr.N.Devarajan,Prof/EEE,GCT for his valuable guidance and support towards the outcome of this paper and Dr.P.Uma Maheswari,Prof/CSE,ITC and also Dr.Rajendran Assistant Professor of Medicine, Madras Medical College for sharing his valuable inputs and Chief Nurse Mrs.Visalatchi Chennai, for her constant support in providing conducive environment to study and actualize the solution References Chart 1: Ratio of Automation proposal VII. Conclusions It is crystal clear that SOA based image processing for classical microscopes will enhance the Quality of Service (QoS) to the patients who are deprived of wealth. When SOA s web service and RESPRO software are given free of cost and healthcare agencies will consume it for maximized improved results generation. The beauty of the framework is that one need not invest in anything.it is enough that one makes use of available resources for better improved situation with the available computers to connect with the existing microscope with the minimal investment of updating their environment to suit their needs into the proposed framework methodologies. This can be of spectacular use in hospitals, laboratories, research scholars, education agencies to make the best use of existing microscopes and integrate them across the global community to make best diagnosis of laboratory results thus [1] Sophia Rahaman, Seena Biju, Data Mining Facilitates e- Patients through e-healthcare: An Empirical Study, International Conference on New Trends in Information and Service Science-2009. [2] Paolo Soda, Early Experiences in the Staining Pattern Classification of HEp-2 Slides, Twentieth IEEE International Symposium on Computer-Based Medical Systems (CBMS'07) [3] V. Todica, M. F. Vaida, SOA-Based Medical Image Processing Platform, 2008 IEEE [4] Tom Closson, Charting A Course For ehealth, Ontario Hospital ehealth Council March 2004. [5] http://www.amtimaging.com/ [6] A. Rigon, P. Soda, D. Zennaro, G. Iannello, and A. Afeltra, Indirect immunofluorescence (IIF) in autoimmune diseases: Assessment of digital images for diagnostic purpose, Cytometry, Accepted for Publication, February 2007. [7] U. Sack, S. Knoechner, H. Warschkau, U. Pigla, and M. K. F. Emmerich, Computer-assisted classification of HEp-2 immunofluorescence patterns in autoimmune diagnostics, Autoimmunity Reviews, vol. 2, pp. 298 304, 2003. [8] The Extraction of Intracisternal A-Particles from a Mouse Plasma-Cell Tumor1 - (CANCER RESEARCH 28, 2137-2148, October 1968) [9] http://www.saas.com the complete reference for the Software as a Service [10] http://en.wikipedia.org/wiki/patternrecognition [11] Robina Asraf, Muhammad Akbar, Diagnostically Lossless Compression-2 of Medical images, Proceedings of the first Distributed diagnosis and Home Healthcare (D2H2) Conference Arlington,Virginia, USA, 2006 [12] Md. Mahmudur Rahman, Tongyuan Wang, Bipin C. Desai, Medical Image Retrieval and Registration: Towards Computer Assisted Diagnostic Approach, Proceedings of the IDEAS Workshop on Medical Information Systems: The Digital Hospital (IDEAS-DH 04)