Role of Digital Image Processing

in Telemedicine
T. Kalaiselvi1, K. Somasundaram2 and P. Kumarashankar3
Department of Computer Science and Applications,
Gandhigram Rural Institute, Deemed University, Gandhigram624302, Tamil Nadu, India
e-mail: 1kalaivpd@gmail.com, 2somasundaramk@yahoo.com, 3pkumarashankar@gmail.com
AbstractIn this paper we have discussed about the
growth of telemedicine and teleradiology in India. We
have also discussed about the possibility of saving storage
space and bandwidth while sending MRI scan images over
the network for telemedicine purposes. We have come up
with an index to compare the optimized storage space
usage and bandwidth while transmitting the images. We
have done this using the datasets used by our previous
work. This article is a continuation of our previous work
in further enhancing the role the image processing in
teleradiology. We have also discussed about the various
telemedicine projects in India and the various
organizations, both government and private entities that
have involved in taking telemedicine to the remote areas of
India. We have briefly discussed the usage and benefits of
cloud computing in the field of telemedicine/teleradiology.
Keywords: Digital Image Processing, MRI head scans,
Teleradiology,
Telemedicine,
Medical
Imaging,
Telemedicine in India, Cloud Computing in Teleradiology.

I. INTRODUCTION
Telemedicine is the practice of giving clinical care
to people at a distance who dont have the privilege of
getting specialized care in their place. Its a boon to the
humanity by modern science and technology.
Telemedicine can be categorized like: Collecting the
data from the patients and sending to medical experts
for offline assessment; Monitor the health conditions of
the patients remotely using technological devices;
Interactive services providing interaction between the
medical officer and the patients through phone or online
communications.
Teleradiology is the process of transmitting the
clinical images from one location to another in want of
expert opinion from other radiologists and physicians.
The growth of radiologist population does not meet the
growth of imaging procedures and hence teleradiology
plays a vital role at this juncture. In situations where
sub specialists are needed, teleradiology ensures that
there is 24/7 availability. This makes use of specialized
software to transmit the medical images through
internet and other networks. The latest addition to this
list is Cloud computing. Image processing and
compression are often used in teleradiology. These are
used to ensure that only slices of interest are sent to the
specialists instead of sending all the information thus
saving the cost and bandwidth. The growth of
technology
in
the
medical
imaging
and
telecommunications has created a revolution in field of
teleradiology.

II. TELEMEDICINE IN INDIA

India is a fore runner in the field of information
technology. Now it is emerging as a popular hub for
medical tourism given the fact that various treatments in
India are much cheaper when compared to other
countries in the world. Considering all these facts India
can emerge as a cloud computing hub for providing
services in the medical sector particularly in the field of
teleradiology where the requirements for specialists is
abundant.
Considering the geographical complexity of India
and its population distribution, it is a herculean task to
provide health care to most of its people who live in
remote villages. Having most of its medical specialists
working in metros, there are only very little options to
provide health care to remote areas. There are various
government agencies and NGOs which setup occasional
health care programmes throughout the country which
do very little to the Indian community.
At this juncture telemedicine is what seems to be a
boon to the Indian health care industry. Given the
advancements of technology in telecommunications in
India, this seems to be a doable job in the next 15 to 20
years. As a first step of providing health care to remote
areas, the Union government has introduced ambulance
services at door step which is well implemented in most
of the states. Some of the states like Gujarat have come
up with plans to setup call centres with paramedics to
suggest primary actions to be taken in case of an
emergency. The Apollo Telemedicine Networking
Foundation has tied up with the Government of Gujarat
to create mobile telemedicine facilities keeping in mind
the disaster caused due to the natural calamities [1]. The
telemedicine units setup as part of this initiative make
use of the Ku band satellite connectivity provided by
ISRO.
The government of Tamilnadu has setup
telemedicine at Primary health Centres which are linked
to the nearest medical college hospitals under the RCH
programme. The Meenakshi Mission Hospital and
Research Centre in Madurai has also introduced
telemedicine to cater the needs of the rural mass in
collaboration with Mahasemam Trust and supported by
Direct Relief International, USA, and Indian Space
Research Organization. Sri Ramachandra Medical
College & Research Institute, Chennai is one of the

Role of Digital Image Processing in Telemedicine 173

forerunners in providing telemedicine facilities starting

from as early as 1997 [2]. It is one of the super specialty
hospitals to provide health care across the country.
There are a number of telemedicine centres that have
been setup by various hospitals, research institutes and
other organizations across the country.
III. HANDLING ELECTRONIC PATIENT REPORTS
All the information of the patients is sent through
insecure networks and hence there is a high risk of
using this personal information for inappropriate and
illegal purposes. This is a great challenge to the
telemedicine field as the patients medical history is not
only available to the physician, but also to other
individuals like technical staff etc. Any breach in data
security will not only affect the patient but also the
telemedicine practitioners as well. Considering the
several advantages of telemedicine, such security issues
should be addressed to ensure that the data is as secure
as paper records or even better than that.
Telemedicine becomes successful only if it takes
less time to transfer data between both the parties.
Especially, in case of Magnetic Resonance Imaging
(MRI)scans the data volume is huge and data clarity is
very much on the focus. Bandwidth is considered to be
a barrier as bulk amount of images are to be sent
through the networks which take plenty of time,
resources and cost. At the same time we have to ensure
that the Region of Interest should be identified
automatically. The data hiding in the MRI volumes
could be done taking various factors into consideration
like implementing firewallsat both the ends, allowing
access to data by allocating user ids and passwords by
rolesSay a technician who is in the position of
sending/receiving the data can have access to send the
files from appropriate folders from the sender
side/download the data to appropriate locations in the
receiving end. But s/he need not view those data by
opening them; the consulting physician can have access
to view those data and write reports based on that data;
a patient can have access to view his details but not to
edit his details.
The MRI volumes should contain some patient
information so as to relate them to the corresponding
patient. The quality of the image should not be
compromised to include this information in the images.
The patient information could be encrypted and sent
along with the image. The encrypted EPR should not
occupy more space as the space available outside the
region of interest is very less.
The data of the patients are captured electronically
instead of traditional paper records. These records are
saved in the computer systems and could be shared
across geographical areas so as to get opinions from
various medical officers. This process has so many
benefits and at the same time it has some practical
difficulties in some countries like India where multiple

languages are spoken. Multi lingual support for forms is

needed in these scenarios in order to provide service to
a larger group.
IV. CLOUD COMPUTING AND CLUSTERING
In teleradiology when information has to be kept
confidential, it is mandatory that there is restricted
access and at the same time we have to ensure that
necessary details are available to the telemedicine
application users. Seeing it from the perspective of
using computer clouds, the telemedicine service
providers have access to the details of many patients.
This is an advantage for the service providers, as they
can group similar cases together in thus reducing the
wait time for expertsA specialist can identify a
common treatment procedure for a group of similar
cases, or a specialist can access the treatment procedure
of a similar case that has occurred in the past. Thus
specialists do not have to spend more time in designing
a new treatment procedure. This grouping of similar
cases is called as Clustering [6] and there are various
algorithms available to do the clustering. This way we
can build a repository of various patients details under
various categories, grouping the similar cases under
same categories and indexing them. In this way we can
refer to a similar case in the repository just by passing a
query to the system.
V. SAVING STORAGE SPACE AND BANDW IDTH
W hile analyzing the brain tumours [3],[4], actual
brain portion is identified and the Slices of Interest are
grouped for expert assessments. In cases where remote
analysis of the medical images is inevitable, we have to
send those images through internet to the experts
located in remote areas. As discussed above, bandwidth
becomes a major factor in determining what has to be
sent and what neednt be sent. Referring to the dataset
[5], we have done a small assessment to calculate the
space needed to store and transmit the slices of interest.
Assuming, in the total set of the images, we will be
transmitting the actual slices of interest in addition to
10% of the leading and trailing images (a minimum of
one and a maximum three at each end). The images
taken for comparison are grey images with a resolution
of 256x256pixels [5].
Using appropriate brain extraction algorithms
(BEA)will ensure that exact data is not missed and
correct slices are chosen for assessments. The existing
algorithm extracts the brain [3], [4] and identifies the
slices of interest that contains the abnormal portion [5].
Once this is identified, we need to send them for
assessment. Table 1 gives the details as to how much
space is saved which is directly proportional to the
effectiveness of the algorithm that we have chosen to
automatically identify the slices of interest. The above
table can be used as an index with an acceptable range
falling within 10%.

174 National Conference on Signal and Image Processing (NCSIP-2012)

TABLE 1
S. No

Volume

Total Slices

1
2
3
4
5
6
7
8

v01
v02
v03
v04
v05
v06
v07
v08

54
43
56
29
24
24
27
24

Actual
Slices
21
9
3
12
6
11

Transmitted
Slices
25
11
5
14
8
13

VI. CONCLUSION
Currently we are in the process of developing a
simple algorithm that could be used to encrypt the EPR
that is sent along with MRI scan images. Teleradiology
has a high potential as a business. Various other factors
like, internet connectivity, scan centres and other
logistics have a say in its success.

Total Size
(256x256)
3670016
1900544
1572864
1572864
1769472
1572864
[3]

[4]

[5]

REFERENCES
[1]
[2]

Apollo Telemedicine Networking FoundationKey Projects,

Available: http://www.telemedicineindia.com/KeyProjects.htm
Sri Ramachandra Medical CenterTelemedicine, Available:
http://www.sriramachandra.edu.in/telemedicine.htm

[6]

Transmitted Size
(256x256)
1638400
720896
327680
917504
524288
851968

% of Space
Saved
44
38
21
58
29
54

K. Somasundaram and T. Kalaiselvi, Automatic Brain

Extraction Methods for T1 Magnetic Resonance Images using
Region Labeling and Morphological Operations, Computers in
Biology and Medicine, Vol. 41, pp. 716725, 2011.
K. Somasundaram and T. Kalaiselvi, Fully Automatic Brain
Extraction Algorithm for Axial T2-W eighted Magnetic
Resonance Images, Computers in Biology and Medicine, Vol.
40, pp. 811822, 2010.
K. Somasundaram and T. Kalaiselvi, Fully Automatic Method
to Identify Abnormal MRI Head Scans using Fuzzy
Classification and Fuzzy Symmetric Measure, International
Journal on Graphics, Vision and Image Processing
(ICGST-GVIP), Vol. 10, No. 3, pp.19, 2010.
K. Somasundaram and T. Kalaiselvi, Segmentation of Brain
Portion from MRI of Head Scans using K-Means Cluster,
Accepted in International Journal of Computational Intelligence
and Informatics, 2011.