REVIEW OF ELECTRONIC-NOSE TECHNOLOGIES AND ALGORITHMS TO DETECT DIABETES MELLITUS
1) Sari Ayu Wulandari, 2,*) Sutikno Madnasri, 3) Susilo, 4) Ratih Pramitasari, 5) Diana Almaas Akbar Rajah
1,4,5) Department of Electrical Engineering, Dian Nuswantoro University, Semarang, Indonesia
2,3) Advanced Composite Laboratory, Physics Department, Faculty of Mathematics and Natural Sciences, Universitas Negeri Semarang
Abstract
The new Electronic nose (enose) instrument is able to detect diseases through aroma blood, secret, stool, breath and urine. Scientists in the world develop enose for early detection of diabetes mellitus (DM). They developed a device and an enose recognition algorithm. Enose is an instrument consisting of hardware and software that contains a pattern recognition algorithm (PARC-Pattern Recognition). The instrument is able to provide decisions in realtime, sensitive to aroma changes and portable. A large number of e-nose instruments, based on a variety of technologies and operational mechanisms, are available for early detection of diabetes mellitus. Some of the common enose technologies used are polymer sensors (CP), metal-oxide semiconductors (MOS), and nano particles. The E-nose device is useful in designing portable and mobile technology as a remote monitoring tool for patients who require intensive control instulin. This article is a recent system review and enose algorithm, developed over the past decade, which has potential applications for early detection of diabetes mellitus. This article discusses a review of the development of enose for early detection of diabetes, starting from the use of the type of sensor, comparison of the number of sensors, the use of feature characterization methods, feature matrix, feature extraction methods and pattern recognition methods.
Keywords: Review, diabetes detection, enose
Topic: Theoretical and Analysis in Materials