Title of project: Intelligent Mobile Disgnosis System using Incremental Learning Algorithm:m PADI and mTADI
Researchers:
Azizi Ab Aziz
Azham Hussain
Mohdnor Basri Shafe
This project is based on the development of an intelligent diagnosis system using an incremental learning algorithm via smart mobile devices. Universiti Utara Malaysia's system utilizes the specific knowledge of previously experienced, concrete problem situations (cases) gathered from previous selected observation.
A new problem is solved by finding a similar past case and reusing it in a new problem situation. By having an incremental learning ability, the system sustained learning, through retaining new experiences for future problems. In addition, this project harnesses the capability of mobile devices in order to serve the nomadic nature of most decision makers (expecially the agricultural and medical domains).
The ability of mobile devices to synchronize with computers enables the easy creation of portable databases. Currently, we have successfully developed two fully functional systems to serve that purpose, namely mPADI and mTADI.
The first system, mPADI is an intelligent mobile system which has the ability to diagnose 6 types of paddy diseases that are commonly found in paddy fields. mPADI was tested using real paddy disease cases gathered from selected locations in the northern regions of Malaysia. The experimental result showed more than 93.5% diagnosis accuracy, thus providing an alternative means of decision making support for the agricultural officer in the field, using mobile devices.
The 2nd project is, the mobile Intelligent Thalassaemia Diagnosis System (mTADI). Thalassaemia is a genetic blood disorder where the blood cells are unable to carry sufficient oxygen supply to the body’s organs. It is estimated that more than 2000 Malaysians are suffering from this disease, while 3 to 5 % were diagnosed as carriers and this number is increasing. mTADI was developed and tested using Thalassaemia cases collected from the Pathology Department of Alor Setar General Hospital in Malaysia. It has with more than 90 % accuracy rate. This adds confidence to the ability of this system as an alternative method to support medical practitioners in Thalassaemia screening decision making.
One of the most important contributions of this project is the implementation of mobile intelligent diagnosis for medical and agricultural cases and providing the basic framework to develop similar systems in different scopes and domains. Thus it attests to the merit of relying on mobile diagnosis model as a worthy tool for decision making process, based on incremental learning model generated.
