Application of GIS in healthcare using asymmetric isochrones and different digitized planar...

Application of GIS in healthcare using asymmetric isochrones and different digitized planar curves


Suvra Prakash Mandal
Suvra Prakash Mandal
Student, Post Graduate Diploma in Computer Aided Management
Add. Annexe 237, IIM Calcutta, Joka 700104
E-Mail: [email protected]

Ayan Nandy
Student, Doctoral Fellow Program
Add. Annexe 201, IIM Calcutta, Joka 700104
E-Mail: [email protected]
Tel.No. : 033-8681607, Fax No. : 033-4678307

The purpose of this paper in to apply GIS in the domain of health care to cater to the following

i) Locations of new hospitals and ambulance depots in an administrative unit (state, district, subdivision etc.) in such a way that within t minutes a patient can be transported to the nearest hospital from any population conglomerate (village or municipal ward). The value of t will be decided by the decision-maker.
ii) suggesting a trajectory for medication and vaccination camps in a region affected by an epidemic in such a way that maximum people can be served in a limited time. In the ‘View’ of an administrative unit where we need to investigate the location for new hospitals, we digitize all the population conglomerates in a ‘theme’ named vertices and all existing roads between different pairs of population conglomerates in a ‘theme’ named edges and each edge is ordered. For each directed edge, time taken to travel from one vertex to other is saved in the spatial database. Using Dijkstra’s algorithm we find the minimum time path from each vertex to every other vertex in the zone of our consideration. The travel time between a pair of points is not necessarily symmetric. We define buffer zones about each vertex as the zone within the t minute asymmetric isochrone around it, i.e., the zone containing all vertices from where it takes less than t minutes to reach the vertex around which the buffer zone is drawn. Thus in polynomial time we can draw buffer zones using asymmetric t minute isochrones about every existing hospital. In the uncovered zones (zones not in buffers around existing hospitals), locations for new hospitals are proposed. We have suggested a heuristic, which is an extension of our earlier work[1]. In every hospital domain (new or old) we suggest locations for ambulance depots, which are also extensions of our earlier work[1]. In urban areas with limited social cohesion, GPS device-driven emergency health services to older people can be designed. Another area for application of GIS in healthcare is epidemiology. In case of an epidemic we digitize the affected villages and draw a convex hull around it. Then we draw a buffer zone of k kilometers around the convex hull. In our earlier work[1], we suggested an Archimedes spiral trajectory of the medication and vaccination camps from the epicenter of the epidemic to the periphery of the buffer zone. Here we compare it with logarithmic spirals, hypocycloid, concentric circles and deltoid trajectories depending on the contiguousness of the disease.

As a result of using our heuristic we can identify the locations for new hospitals in an administrative unit. The results of GIS application in the epidemic affected zone are:

1) For a zone with uniform density of cases of the disease, a combination of Archimedes spiral and logarithmic spiral trajectory is identified, the Archimedes spiral being for the affected zone (which needs medication) and the logarithmic spiral for the buffer zone (which needs vaccination)
2) For a zone where the density in not uniform and where we need to employ more than two teams, we suggest Hypocycloidal paths for denser zones. The hypocycloids will be drawn around concentric circles centered at the epicenter of the epidemic as defined in our earlier work [1].

[1] Nandy A., Mandal S. P., Implementing GIS in healthcare; Xpressions 2002, Xavier Institute of Management, Bhubaneswar.