Developing Smart Telehealth System in Indonesia: Progress and Challenge

BTele-Health Systemackground

Indonesia have a high mortality rate caused by heart and cardio vascular diseases. One of the major factor that caused this issue is the lack of medical check up especially for heart monitoring. It is caused by the limited number of medical instrumentation e.g. ECG in hospitals and public health centers. Another factor is the small number of cardiologist in Indonesia. There are only 365 cardiologists across the country, which is a very small number compared to the 250 million of Indonesia population. Furthermore, they are not distributed evenly in all provinces, but centered in Jakarta and other major cities. Meanwhile, Indonesia also has similar problem in health area. High number of fetal and mother mortality becomes aTele-Health System serious problem. One of the major factor is the lack of fetal growth monitoring. It is caused by the limited number of USG device and

Obstetricians in
Indonesia. Therefore, fetal growth monitoring is difficult.

Methods

The Tele-ECG system has three main components: an ECG sensor, a smartphone, and server. ECG sensor is used to acquire heartbeat signal from patient. ere are baseline wandring removal (BWR), beat segmentation, and wavelet for dimensionality reduction. Afterwards, the signal can be classified to predict the patient’s condition automatically. Tele-USG System has 2 main components, they are the smartphone and server. In Tele-USG system, we have not developed hardware yet, due to the complexity of the USG sensor. In this system we used ultrasound image captured from conventional USG devices. Soware installed in the patient’s smartphone is used to monitor fetal growth.

Results

We have been developing classfier algorithm named ANGLVQ to classify hwRT heart beat signal. The performance of the algorithm is above 95%. Next, the signal is sent to server to be verified by cardiologists. To provide fast transmission, we first compressed the signal. For compression, we used 2D SPIHT compression. e error of the compression is relatively low, which is 3,46% for 24 compression ratio. e soware can compute HC, BPD, AC, FL, and HL from head, abdomen, femur and humerus organs. For this automated computation we have been developing Hough Transform based curve approximation. Fetal head and abdomen can be approximated by ellipse curve, whereas fetal femur and humerus was approximated by line curve. The approximation algorithm has less than 10% error.