Intelligent Medical Image Analyzer Benjamin Aribisala – Lagos State University – Nigeria (Benjamin.Aribisala@lasu.edu.ng) e-Research Summer Hackfest – Catania (Italy)

Outline Background Scientific problem Aims and benefits Use Case Workflow Implementation strategy Summary and Conclusion

Study Background Medical images are images acquired from humans or animals Medical images are used for clinical diagnosis, treatment and patient management. There are many methods of acquiring medical images. Each method has specific use, strengths or weaknesses Medical Image analysis and computer vision provides a platform for developing computational techniques for converting the raw images to meaningful information.

Scientific Problem Some of the challenges in the field of medical image analysis Imaging equipment are very expensive Image acquisition is very expensive There are limited availability of medical image analysts in Nigeria Limited collaboration efforts and ability to share expertise amongst clinical experts The advancements in internet technology, data communication and reduction in hardware cost has led to the birth of Science Gateway Science Gateway can help in solving medical image analysis problems by offering tools for Creating of image repository Making image analysis tools available Making workflow available Adapting some of the tools to make them intelligent Science Gateway is in top gear in the developed world but very sluggish in Africa In particular, there is no platform for image analysis in Africa

Project’s Aim and Benefits The aim of this project is to develop and implement an e-infrastructure which will contain a repository of medical images acquired in Africa. The infrastructure will contain image analysis tools which will allow users to be able to upload and analyze their images. BENEFITS It is our hope that such useful, rear and unique tool could lead to Improved and timely diagnosis Improved patient management Increased life expectancy Encourage collaboration Safes cost of travelling to see other experts

Use Case An e-infrastructure which will contain a repository of medical images acquired from Africa. It will also contain some tools for analyzing medical images. Users will be able to upload images and the proposed system will analyze and return outputs. The proposed system will also have the ability to advise users on how to make sense from their images. This e-infrastructure will be useful for Clinicians Researchers Postgraduate students – guidance and analyzer Users can get specific information from medical images.

Work Flow Build the repository using gLibrary and Onedata Develop a web interface to allow users access tools on the platform. Use PhP and MySQL Develop a platform for uploading of images. Use Onedata Make image processing tools available on the platform. Use MATLAB, FSL and SPM Develop technique to allow user to be able to select specific tool. Use kepler Develop a report generation module Build intelligence into the system using Expert system development approach

Scenario – Analysis of brain MRI images Medical images, e.g. Brain Images will be acquired and uploaded into the repository Users will have the opportunity to select the kind of image analysis or processing to perform on the images The web interface will allow users to select a single image processing step or a group of processing steps Some of the image analysis steps Image format conversion Brain extraction Image segmentation Image registration and extraction of information specific to certain anatomy. A report generation module will allow the user to receive specific information on the images analyzed

Implementation Strategy Install FG, Kepler, INDIGO and other related tools Install image analysis tools Develop image processing pipeline for image analysis steps Develop the proposed e-infrastructure using relevant tools Use relevant science gateway tools to create a repository for brain MRI images Develop a web interface for the proposed tool Develop the authorization and authentication modules using the relevant tools

Summary and conclusion Develop a pilot version of the medical image analyzer using brain images which will allow users to upload medical images and analyze them Future Plan Seek funding support for full implementation Include other types of medical images Include other human anatomies Opened to collaboration

Special Appreciation to Roberto Barbera and Simon Taylor Thank you! sci-gaia.eu info@sci-gaia.eu Special Appreciation to Roberto Barbera and Simon Taylor Benjamin.Aribisala@lasu.edu.ng