Around March 1997, we got an invitation from the Department of Electronics (DOE), which is now called the Department of Information Technology (DIT) under the Ministry of Communication and Information Technology (MCIT), Govt. of India, for writing a proposal on a low cost telemedicine system. During that time the concept of telemedicine was just beginning to catch the imagination of the medical world. However, there were quite a bit of confusion and myth around it not only among laypersons, but also among many doctors and health administrators. Some of them considered it as the replacement of medical professionals, and had a negative and sarcastic attitude towards its adoption. Some thought it was a utopian sci-fi solution for the future generation. The net effect was that none took it seriously and there was a lack of enthusiasm in using this technology for improving the health care delivery to the people. In fact, when we approached the then Health Minister of the Govt. of West Bengal, we got a lukewarm response. The proposal though got a push due to the enthusiasm and energy of a doctor, Dr. V.K. Gupta. At that time he was the Head of the Dept. of Leprology of the School of Tropical Medicine (STM), Kolkata. Previously he had been collaborating with us in developing an expert system for treating leprosy. So exactly a year after we proposed to make a telemedicine system using low bandwidth channels like POTS, for treating the patients of chronic tropical diseases, such as leprosy, Malaria, Kala-azar, etc. It took almost one more year to get the sanction from the DOE and the project started in the January of 1999. We got an able partner with us, Webel Electronics System Ltd. (WECS), Kolkata in the execution of this project, and since then the partnership has been growing strong in executing different other telemedicine and related projects.
First Telemedicine System
Within a year and a half (August 2000), we developed a prototype system on a Windows-NT platform. We named the system TelemediK. As the system was meant for low-bandwidth data communication we took care of the following for reducing the requirement of bandwidth.
(i) Adoption of store-and-forward methodology for data transfer supported by organized capturing of patients’ data in a back-end RDBMS, and
(ii) Design of an online graphics communicator for data conferencing between two doctors against medical images with annotations, file transfer and text chatting.
The system went through several test runs using ordinary telephone lines between distant places, such as Kolkata and Bhubaneswar, with Kharagpur. As our Institute has centers at these two places, we could perform these experiments conveniently as and when required. Finally the system was installed in the STM, Kolkata in November, 2000. The first beta test between STM, Kolkata and IIT, Kharagpur took place in February, 2001. We found that our system can work with a data rate as low as 5 Kbps. In our design, we kept a separate telephone line for audio conversation and there was no scope for video conferencing. We made some attempts on supplementing the consultation with online video frame transfer at a very slow rate of 1 frame per second. However, it failed to clear the acceptability test from the conversing participants. In our early days, we went to several places, such as Christian Medical College, Vellore; Westbank hospital, Howrah; Mahatma Gandhi Institute of Medical Sciences, Wardha; Sikkim Manipal Instititute of Medical Sciences, Sikkim, Elitex exhibitions held at Delhi, etc., to demonstrate the system and got appreciations from different corners. At that time, a few commercially available telemedicine systems used impromptu data transfer for tele-consultation and were not bothered about their organization and archival for subsequent uses. Our system was a departure from that trend, as it used a back-end RDBMS. Gradually, other systems also started using data bases for archival and retrieval of patients’ data during telemedicine sessions.
In 2002, the first telemedicine network in West Bengal was set up with the referral center at STM, Kolkata, and two other nodal centers at the State general hospital of Habra, North 24 Parganas, and at the district hospital of Coochbehar, a place at a distance of about 500 Km. from Kolkata. The doctors at referral centers were supposed to provide their expert opinions to their peers at nodal centers. In the later versions of the software, however, we removed this distinction and two doctors at two ends are able to participate with equal roles and rights for consulting over patient’s data. WECS did excellent work in setting up the network and building the local infrastructure at those centers. They also provided necessary man power support for its daily operation. A good number of referral cases were discussed using telemedicine services during the project-period. Around 1400 patients of these two nodal centers were treated in next two years. Interestingly, by observing a large number of referral cases in dermatology from Habra, the Health department posted a dermatologist there. Naturally, this reduced the requirement of referrals. This also shows the evanescent nature of telemedicine. Through planning aided by its history of referrals, or training to a non-specialist by tele-consultation, the intensity of its use for the same purpose diminishes with time. However, as medical knowledge is dynamic and expansive, it is expected that there will be always a need for such consultation, with an improvisation on the quality of knowledge sharing.
Expansion of Network
In the year 2003 the network was expanded to include a few more hospitals, including referral centers at NRS Medical College and Hospital in Kolkata, Chittaranjan Cancer Hospital, Kolkata, Bardhaman Medical College and Hospital, Burdwan, and North Bengal Medical College and Hospital at Siliguri. The nodal centers were set up at district hospitals in Darjeeling, Purulia, Suri, Medinipur and Baharampur. This time we made several changes in our system as it used relatively higher data rates of 512 Kbps from the West Bengal State Wide Area Network (WEBSWAN) of the Govt. of West Bengal, which run on fiber optic links owned by BSNL. Some centers were also connected through ISDN links. The system was also supplemented by video conferencing functioning independently over the same link. The formal inauguration of the services took place on the 26th January, 2004, by the then chief minister of West Bengal, Mr. Buddhadev Bhattacharya.
Another stage of expansion followed soon with a new version of TelemediK (TelemediK-2005) installed in 16 hospitals (including the previous hospitals) across the state. In TelemediK-2005 we removed the distinction of referral and nodal centers, and made provisions of multi-referencing of a patient. During the project period a number of patients (about 3000 patients in six years) were treated from remote ends using this infrastructure. In particular, there were regular referral cases from Purulia, Baharampur and Raiganj. However, the usage was largely dependent on individual doctor’s enthusiasm and dedication. Many of the centers remained practically defunct since their inception. In spite of our repeated urge and suggestions on increasing the awareness of patients and doctors about the infrastructure, we failed to enthuse the officials of health department and local administrators for integrating the systems in their environment. We visited some of the centers and met the doctors to discuss their problems on using the system. We conducted a few workshops and conferences with demonstration of the system. A few meetings were also held in the Department of Health and Family Welfare in Kolkata for planning and coordination. However, they were not enough to overcome the inertia of the physicians for using the system. At one point, WECS wanted to make awareness campaign locally with hoardings and posters for drawing attention of people. Even the content of the advertisements were also provided for the approval of the Health secretary. For reasons unknown to us, the approval never came. Naturally as soon as project periods were over and systems were handed over to the State Government for their maintenance and sustenance, the telemedicine services were stopped. Even the active centers like Baharampur and Purulia stopped functioning due to lack of proper maintenance, planning and coordination from the Department. Presently, those infrastructures still exist in some of those hospitals, but they remain largely unutilized. Ironically, due to its operation during the project period, Govt. of West Bengal bagged the “National award for exemplary implementation of E-Governance initiative – 2004 on Telemedicine Project in West Bengal”.
Telemedicine in Tripura: A different experience
Our work in the hospitals of West Bengal drew attention of other state governments. In particular, the Government of Tripura showed interests in its deployment throughout the state. They were encouraged by the then undersecretary of DIT for putting forward a proposal in this regard. So in the summer of 2004 we went to Agartala and discussed a proposal with Mr. S.K. Ray, at that time the Secretary of the Govt. of Tripura for Health and Family Welfare. He was also the secretary for the Department of IT of the state. He had given us the options of connecting either district hospitals (DH) or sub-divisional hospitals (SDH) to two tertiary hospitals in Agartala, namely GB Pant Hospital (presently it is also a medical college) and the Indira Gandhi hospital, which handles child birth and gynecological diseases. In Tripura SDHs are not so equipped and they fall in the category of Primary Health Centers (PHC). From our experiences of West Bengal, we opted for connecting PHCs, as we felt they have more needs for availing telemedicine services. The district headquarters have better infrastructural support with digital connectivity through Tripura’s SWAN. For initiating a pilot program, DHs would have been convenient for us for a speedy implementation. On the other hand, many subdivisions were situated at places, where physical transportation itself through hilly terrains and bad road conditions was a nightmare. The problem got aggravated due to insurgency problem in the state reigning at that time. No one was ready to move without the security cover provided by the state. At places, even a distance of 50 Km. would take about 4 hours to pass through in those days. Apart from poor communication infrastructure, they also lacked several civic amenities. However, we took the challenge of setting up the centers in subdivisions instead of district headquarters, as we needed no new proof of concepts of our technology. We were eager to see its usage in health care services. The decision proved to be crucial for the success of the telemedicine in this state, a very different experience from what we had in the state of West Bengal.
In Tripura, four of the five proposed centers (at Amarpur, Gandachhara, Kanchanpur and Chailengta subdivisions) started functioning within a year of the sanctioning of the project. The network was inaugurated by the Health and Family Welfare Minister of the state on 9th June, 2005. We must mention about two persons, who played vital role in the speedy deployment of this network. Mr. Rajat Bhattacharya, the then Asst. Executive Engineer of the Department of Health and Family Welfare, Govt. of Tripura, took all the trouble and care for providing logistic support from the Govt. of Tripura in identifying the centers, allocating space, arranging escorts and transportation to remote interior places of the state, etc. The other person is Dr. J.N. Maiti, the Director of WECS. He with his dedicated team of young engineers, who braved all odds in installation and training of the system, supervised and monitored the execution with passion and diligence. From the very beginning the doctors at the SDHs and PHCs took the telemedicine services for consulting with specialists of two hospitals at Agartala. The local populace also enthusiastically welcomed the facility and put pressure on hospital administration, if the system remained non-operational due to failure of links or machines. With the increasing use of tele-consultation the Govt. of Tripura requested for more deployment and the DIT supported expansion of the network in three stages connecting 11 SDHs / PHCs with their two resource centers. About 14400 patients were treated in five years (from June, 2005 to October, 2010) using this network. It is heartening to see that every year the number of patients treated through telemedicine is increasing. Presently from these centers over 4000 patients per year are getting the benefit of this technology. The State Government is also proactive in the sustenance of the services. After the completion of the project period, the state government took special initiative for arranging funds for maintenance of the systems and supporting trained manpower to the telemedicine centers.
iMediK: Telemedicine over internet
TelemediK was developed as a peer to peer application. Though it was designed as an internet server, individual hospitals would require hosting this server for communicating with their peers. From this model, we moved to a more cost effective centralized server based system, which would be able to cater services to its clients over the internet (or intranet following standard application layer protocols). We named the telemedicine server iMediK, which processes clients’ requests in a four layered architecture providing much needed security to patient data from their unauthorized access. The centralized server based system reduces the cost of installation to a great extent. The client end does not require any special application software other than an internet browser. Similarly, the cost of communication link was also brought down, as it does away with the need for a dedicated link for data transfer. In the four layered architecture, a request is handled through a proxy server, which initiates a succession of internal sessions for accessing the data from data base, processing it and presenting it to the user in the form of a HTTP response. This methodology of data access prohibits their direct access and allows only controlled access through query processing at the business logic layer, which sits atop the database layer. The presentation layer on the other hand isolates the process of data formatting and has the flexibility of supporting client specific modules for presenting the result of a query. For example, we made separate modules of presentation for processing requests from clients of desktops, PDAs and mobile phones, while all of them use the same set of functions of business logic layer, which is immediately below the presentation layer in our architecture.
iMediK was installed in Medical College and Hospital, Kolkata in the year 2008. It has special modules for treating pediatric patients of HIV and drug resistant tuberculosis. iMediK also has an improved user interface, where a doctor could get the summary of patients’ record relevant to the diagnosis and treatment of a disease. The summary page is customizable and different for different diseases. It has also features of tracking patients’ visits, and presenting temporal trend of any measurable attribute through graphs and tables. HIV pediatric module also has a decision support system for drug administration and patient follow-up. Around 530 patients were treated from April 2008 to October 2009 using iMediK. During this period mostly patients were referred from the Raiganj SDH.
m-Medicine
The term m-Medicine is coined against telemedicine over mobile devices. Our research group is one of the few, who explored the development of this technology in mobile platforms at a very early stage. The TelemediK system was integrated with applications developed on WinCE based Pocket-PCs. Through these applications a user could enter patients’ data from a hand held device and also browse them, which may be available in the form of graphics, images, audio-clips, text and structured text. In our design we took care of constraints imposed by the limited memory space and small display screen of the device. For displaying medical forms, an intelligent text fragmentation algorithm was developed. The technique adapts the number of fields to be displayed in a single view by accessing the data related to screen size and resolution of the client device from HTTP request header. Further, fields are also displayed in order of their relevance, determined a priori. The relevance map is generated through a statistical analysis of number of occurrences of key-words of related documents fetched from internet using a crawler. Moreover, we have also used abbreviations for text compression and assigned color codes to highlight the normal or abnormal ranges of values. A large image is displayed using an adaptive partitioning scheme and by showing the position of a partition with reference to its thumbnail. We have also developed an interactive downloading scheme for browsing images through selection of the region of interest with varying level of compression.
While integrating with iMediK, we changed the design paradigm. Previously our approach was to develop client applications in mobile devices, which directly interact with patient databases. However, due to the security measures imposed through four layer architecture of iMediK, this direct interaction is prohibited. Moreover the advancement in mobile computing and technology makes it possible to work with the new generation of devices, which support internet browsing on LAN and WAN. In iMediK we moved all the related computation of data formatting and image partitioning to the server, in its presentation layer. This also increases the portability of the system. Like desktop computing, in this case also, we do not require any specific client installation. This is the technology we used in integrating decision support system and other modules for treating HIV pediatric patients with mobile devices. It got recognition from the World Health Organization (WHO) as one of the innovative technologies in their report entitled “Compendium of new and emerging technologies that address global health concerns 2011”.
New directions
It has been more than a decade that we are involved in the research and development of telemedicine systems. We are grateful to the DIT, Govt. of India, for their patronage and constant encouragements in our endeavor. We started with the design of telemedicine system in peer to peer network architecture and moved to a centralized server based system. We also considered other architectures of distributed systems, where a combination of nodal and centralized servers share the task of handling local and referral patients, respectively. Another interesting variation of peer to peer model is a hierarchical-referencing model, which is usually practiced in Government hospitals for patient referrals. For example, a patient from a block primary health center (BPHC) is referred to an SDH or a DH, whereas from a DH doctors refer to tertiary hospitals only. We carried out testing and performance analysis of such systems in a simulated environment. Their large scale deployment, however, requires planning at the central or state level. We hope to see such planning and deployment of this technology in the public health care system. Our experiences from the state of Tripura show that rural population of our country could immensely be benefitted from telemedicine, as it brings the quality health care at their door steps.
08/09/11