EE4001 Projects (Academic Year 05/06)
A NFC-based intelligent diabetes management system
A fuzzy classifier for cardiologic auscultation
A fuzzy approach for assessing significance of missense mutations in disease genes
 
Adaptive Type-2 Fuzzy Controller
SPAM filtering using fuzzy theory

A NFC-based intelligent diabetes management system
Seah Yew Keng and Tan Ting Feng
1st Prize, Philips Young Innovators Challenge 2006

Evolving from a combination of contactless identification and networking technologies, Near Field Communication (NFC) is a wireless connectivity technology that enables convenient short-range communication between electronic devices. It offers the ultimate in convenience for connecting all types of consumer devices by enabling rapid and easy communications. In our increasingly complex and connected word, NFC provides a solution for controlling data.

This project aims at using Philips NFC chip solutions to develop an intelligent system for helping a diabetic patient to manage the illness. Diabetes is a disorder in which there is too much sugar (glucose) in the blood. At this moment, no effective short-term treatment had been found for diabetes. Diabetic patients have to manage their diet, medicine and sport lifelong. A well balanced diet is essential to the management of diabetes. Good dietary control aims to provide general adequate nutrition for health and well-being, control blood glucose levels within normal levels, maintain desirable body weight and individualize each meal plan. In addition, proper use of anti-diabetes tablets are important. Patients have to know the name and dosage of their anti-diabetes tablets, do not take more or less than the doctor ordered and take it at the same time each day. Furthermore, appropriate amount of sport had been found to be useful for diabetic patients. However, too much sport will bring problems such as low blood sugar. Overall, a management system that takes good care of diet, medicine and sport will be much helpful for diabetic patients.

It is envisaged that the intelligent diabetes management system will comprise of a wearable device, as well as a data management and analysis program. The wearable device should provide the following functions :

1) a pedometer for monitoring the patient's activities
2) serve as a conduit for transmitting information such food intake and blood sugar level to the PC-based program
3) remind user to take medication and to measure blood sugar level

The data analysis program will reside in a PC. It's main function is to collate and analyze the information collected via the wearable device. Based on the data, it should provide health management advice and reports for transmission to the medical team. Two students are needed for this project. One student will focus on the interface programming part on the wearable device side, and the other student will focus on the data analysis part on the PC side.

Back to Top

 
A fuzzy classifier for cardiologic auscultation
Foo Chek Liang

A heart murmur is an extra heart sound in addition to the two heart sounds normally heard with each heartbeat. Innocent heart murmurs are of no clinical consequence but pathologic heart murmurs indicate congenital heart disease is present. When a heart murmur is detected during auscultation (listening to the heart through a stethoscope), the clinician has to assess the murmur qualitatively to differentiate between innocent and pathelogic heart murmurs. Such auscultatory skills can only be accumulated by exposure to a sizeable murmur population and maintained through constant practice. Hence, doubt as to the nature of the heart murmur may still exist after auscultation in many cases and patients are further tested using advanced imaging techniques that are expensive. A more screening device will help to reduce unnecessary testing, thereby helping to reduce health care costs. This project aims at developing a fuzzy classifier to assist a clinician in determining whether the heart murmur is innocent or pathelogic.

Back to Top

 
A fuzzy approach for assessing significance of missense mutations in disease genes
Lee Kunfeng, Daniel

Identification of deleterious mutations within candidate gene provides a means for detecting disease. One approach is to try to infer structural change asociated with mutations and draw conclusions regarding their deleterious character. However, the significance of any base or amino acid change within a gene is unknown until detailed structural and functional analysis has been carried out. Another, potentially rapid, way of finding functionally important sites within a gene is to identify evolutionarily conserved regions. Mutations affecting such sites are assumed to be deleterious for the carrier. This project seeks to develop a fuzzy approach for predicting whether specific mutation is deleterious given sequence data from a set of homologues.

Back to Top

 
Adaptive type-2 fuzzy controller
Kwa Lip Oon
Membership functions of standard (type-1) fuzzy sets are usually represented by the Guassian function. The mean and standard deviation of the Guassian function are frequently determined by 'learning' algorithms. One example is the Adaptive Network Based Fuzzy Inferencing System (ANFIS), which has successfully been applied to a variety of applications. Recently, there is more and more interest in type-2 fuzzy sets because they present a better representation of the 'fuzziness'. Type-2 fuzzy logic systems also provide better modelling accuracy with a smaller rule base. In view of the advantages offered by type-2 fuzzy sets, this project seeks to extend the learning algorithms for adapting a type-1 fuzzy controller to its type-2 counterpart.

Back to Top

 
SPAM filtering using fuzzy theory
Fun Seong Ngee

SPAM is an unsolicited automated email. It is any message, regardless of it's content, that is sent to multiple recepient who do not specifically requested the mail. The amount of such unsolicited commercial email or "junk mail" showing up in an email box is growing. SPAM filtering is a method to combat this problem. The role of the filter is to distinguish SPAM from HAM, the innocent mails, without any false positive (false positive mail is an innocent mail which is considered as Spam even though it is not). This project aims at utilising fuzzy set theory to implement a "Spam Filter".

Back to Top

 
Current Projects

Click on poster for details