1-4244-2384-2/08/$20.00 ©2013 IEEE SMC 2008 The Smart Device for Healthcare Service: Iris diagnosis application David Habsara Hareva, Samuel Lukas, Natanael Oktavian Suharta Informatics, Computer Sciences, Universitas Pelita Harapan Tangerang, Indonesia dhareva@yahoo.com Abstract— We have developed an iridology application for predicting a person health through analysis of iris image. Commonly for commercial purpose, it used Desktop-PC, but we embedded it on Android-based Smartphone. The objective of this research is to find out whether healthcare application that providing direct diagnosis from real-time capturing image can be carried out through mobile phone in general. For that purpose, we developed other version of application that run on windows mobile and windows operating system. The algorithm and structure of the application is made as closely as possible with a previous application. Those applications have been tested on different environment, such as operating system, programming language, and mobile device. The device tested with regard to product maker, processor speed, screen size, storage size, and image retrieval using built-in camera. Although the results can be predicted that the used device with high specifications will be faster to process image analysis, better to obtain images, clearly in displaying images, however finding out the low limit of requirement for running the application is essential. It is found using low camera specification to take the image of iris is not recommended. Additional macro lens causes difficulty when taking a picture. Processing of iris image to predict health condition could be handled by 1 GHz processor, 512 internal memories, and 3.5 inches screen size, even it used low-end product. Keywords— Artificial Neural Network, Iridology, Canny Edge Detection, Automatic Disease Detection, Smart device I. INTRODUCTION Development of integrated technologies encompasses information, electronics, communications, and mobile enabling healthcare applications used by all levels of society, from the top to the bottom layer [1]. Access to healthcare providers are become easier, so that everyone who wishes to live a healthy lifestyle can easily reach it without has to depend on the hospital or medical doctor. We have made a prior Iridology application to diagnose a person's health through analysis of iris image. The purpose of the application is to predict the presence of organ disease, including heart, lung, spleen, and liver. Capturing iris images, analyzing iris image, and predicting a health condition of human organs are major steps of developed application. To analyze obtained iris image is using Canny Edge Detection, the output is detected iris area. This iris area is extracted by Future Texture Analysis, and then it used by Artificial Neural Networks (ANN) to predict organ condition. This one is an example of healthcare application that using a mobile technology in the form smart device as represent of Smartphone and tablet-PC. It is expected used by public communities by the use of relatively inexpensive tool that comply with their purchase capabilities. They will have benefits when it can predict their health condition quickly so early disease prevention and further medical examination can be done by their medical doctor immediately. At this year's, Smartphone sales is ranged 224.9 million, up 46.42% from last year. This shows a significant increase for Smartphone users. Both are those users who a new to Smartphone or replace with the latest Smartphone. Table 1 (a) describes five vendors with the largest percentage of sales in the second quarter (Q2) of this year is the Samsung (31.70%), Apple (14.40%), LG (5.07%), Lenovo (4.71%), ZTE (4.27%), and other (40.11%). Used mobile operating system controlled by Android at 79.01% up 14.76% from the same quarter last year, followed by a 14.13% iOS, and Microsoft 3.29%, Table 1 (b). This fact is the reason why we first build Iridology applications using the Android based. TABLE I. SALES AND MARKET SHARE 2012 AND 2013 http://www.netmarketshare.com Q2, 2012 Q2, 2013 Samsung 45.6 71.3 Apple 28.9 31.8 LG 5.8 11.4 Lenovo 4.3 10.6 ZTE 6.3 9.6 Others 62.7 90.2 Total 153.6 224.9 millions Brands (a) Smartphones vendor Q2, 2012 Q2, 2013 Android 98.6 177.8 iOS 28.9 31.8 Microsoft 4 7.4 Blacberry 7.9 6.1 Bada 4.2 0.83 Symbian 9 0.63 Others 0.86 0.47 Total 153.46 225.03 millions OS (b) Smartphone OS OS Percentage Windows 91.56% Mac OS 7.19% Linux 1.25% Others 0.00% Total 100.00% (c) Desktop OS Currently, we build other version of application that run on windows mobile and windows operating system. The reason is more than 90% of desktop users using windows as its operating system (Table I (c)). Although users of windows mobile are still relative small compared to Android or iOS, however the number of windows users on desktop is superior to others. This information may indicate that user of windows mobile increase. In the present study, we take into consideration the operating system and the device used to run the Iridology application. Types of smart devices vary greatly, not only the operating system, but also hardware, shape, internal camera, touch screen, screen size, and battery life. Our consideration is to find out what the most important feature in running the Iridology application. We hope this research will give basic