Health is a fundamental human right although more than one billion people are unreached in terms of quality healthcare services. Insufficient healthcare facilities and unavailability of medical experts in rural areas are the two major reasons that kept the rural people unreached to healthcare services in developing countries, like Bangladesh. According to the World Health Organization (WHO) statistics, the doctor to population ratio is 1: 1500 in urban areas and 1:15000 in rural areas of Bangladesh. This scenario can be dramatically changed if we can simply convey medical tips using ICT infrastructure to the targeted unreached community. Recent development of Information and Communication Technologies (ICT) of the digital divide has been reduced and these technologies have the great potential to address contemporary global health problems. Telemedicine refers to the use of information and communication technologies to distribute information and or expertise necessary for healthcare services provision, collaboration and or delivery among geographically separated participants including physicians and patients. Telemedicine can be the key for providing good health care facilities to the target unreached community especially low resource countries, like ours (Bangladesh).
In this research work, we have developed an intelligent telemedicine system based on Smart phone. The price of smart phone has reduced drastically in recent time and the number of users is increasing in a rapid rate. Recent study shows that even the relatively poor populations at rural areas are using smart phones. Our Smartphone based telemedicine system, therefore has a great potential to deliver the health care services for rural population of Bangladesh at very reduced cost and less hassle, as it requires a very little movement or out of home town.
DICOT (Digital Imaging and Communication for Telemedicine) is the name of the machine which has been built by telemedicine working group of Bangladesh, and in use at some telemedicine centers. One of the major problems of the DICOT users is long cumbersome registration process. In this thesis work we have developed android apps which will provide opportunity of health care at home. After logging in apps, user can update their basic medical records and can choose and get a confirmation of appointment of specialist doctors.
6
Different medical records can be updated like body temperature, glucometer, ECG, personal information and others. After updating with database then a confirmation message of assigned appointed day and time will be sent to the patient through android app. So, patients need not experience a long cumbersome process of registration. A website of database has been developed using HTML, CSS and PHP.
We have also developed a Telecardiology system which has been designed and implemented in this work. The Raw analog type ECG signal is amplified and filtered by band pass filter. Analog signal is digitized using Arduino board and then, interface between Arduino and smart phone sends this signal to Smartphone. Digitized value of the filtered ECG signal is stored in SD storage card of Smart phone. Using Bluetooth or existing telecommunication network digitized ECG signal can be sent to other Smart phone or to Server.
During transmission of signal it generally gets corrupted by random noise or white Gaussian noise of the existing telecommunication network(s) and even some data points may be lost. Adaptive filter with three different algorithms have been used in MATLAB platform for denosing i.e. removing noise from the ECG signal. In this research work we have used three algorithms named as LMS (Least Mean Square), NLMS (Normalized Mean Square), and RLS (Recursive Least Square) and tested their performances to reduce the noise from ECG signal.
We have taken 250 mV amplitude ECG signal from MIT-BIH database and 5mV (2 % of original ECG signal), 10 mV (4% of original ECG) 15mV (6% of original ECG), 20 mV (8% of original ECG signal) and 25mV (10% of original ECG signal) of random noise and white Gaussian noise is added with ECG signal and Adaptive filter with three different algorithms have been tested to reduce the noise that is added during transmission through the telemedicine system. Normalized mean square error was calculated. For highest amplitude random noise, 25 mV (10% of original ECG signal) added ECG signal, we have got normalized mean square error for LMS, NLMS and RLS adaptive filters are respectively 3.5566×10-4, 2.8322×10-4, 1.5938×10-5.For the case of 25 mV amplitude Gaussian Noise we have found simulation result of normalized mean square error for LMS, NLMS and RLS adaptive filters respectively 4.2407×10-4, 2.459×10-4 and 7.0148×10-5. The errors are very less in all of the cases and we found RLS Filter performed
7
the best amongst the three FILTERS mentioned above in our MATLAB simulation for denosing the ECG signal.
We have used Cubic Spline Interpolation for regaining missing data point of ECG signal. We have taken 5000 data points of ECG signal from MIT-BIH database. In our simulation 11 data points (From 689 to 699 of original data points of ECG), 201 data points (from 800 to 1000 of original data points of ECG), 300 data points (From 1600 to 1900 of original data points of ECG), 500 data points (From 2000 to 5000) and 6 data points (From 4095 to 5000) are made zero and cubic spline interpolation function was called and it could regain the original data points of ECG signal. The Normalized Mean Square Error was calculated and it was found respectively .0909,
.0050, .0033, .0020 and .1667. In all of the cases normalized mean square error is very less and so Cubic Spline Interpolation could be a good solution for regaining missing data points of original ECG signal.
DICOT (Digital Imaging and Communication for Telemedicine) machine is in use to send different medical report like X-ray, mammography, skin image and others, and Gray Scale image is sent without compression.
To increase the efficiency and reduce the BANDWITH requirements, we have developed a DCT based image compression technique. We have used five medical gray scale images of File size 110 KB, 51.1 KB, 118 KB, 62.5KB and 62 KB and after using compression technique we have got 92.5 KB, 38.1 KB, 113 KB, 44kB and 39 kB size of compressed image file. The compression ratio of file sizes becomes 15.9%, 25.4%, 4.23%, 29.6% and 37.097% respectively. File sizes are reduced to maximum 37.097 % without significant loss in image quality or medical information contained in it. Our result suggests that file size can be reduced in an efficient way using DCT and image reconstruction is possible without any loss of medical information, though some not very important fine details are lost.