| Visual Analysis Home | ||||||||||
| PRIMA: Patient Record Intelligent Monitoring and Analysis | ||||||||||
Prima is an example interface to a set of clinical records on bone marrow transplant patients collected over many years at Hadassah Hospital in Jerusalem, Israel. In many areas of bioinformatics and biological research, the data is of varied type. For example, it often contains both categorical (eg. strings representing medical diagnoses) and numerical (eg. white blood cell counts) values. Two of the important tasks in making these data useful and accessible are to, first, organize the data in a meaningful way so that queries can be launched against it and information returned, and second, to present the results of the queries to the user in a way which enables them to understand the data intuitively, and allows intelligent and interactive exploration of the data. WE have been developing tools to enable interactive exploration of data of varied type. We have developed a Java toolkit, called Opal, which can be used to build custom applications, of which PRIMA is an example. The toolkit includes a wide variety of visual presentations of data, some of which are particularly suited to categorical data, some which are more appropriate to numerical data; however all views are linked so that exploration can occur regardless of the underlying form of the data. In the future we forsee integration of visualization applications such as Prima with Integrated Medical Records (IMR) being developed as part of the SHAMAN project to organize, store, and query diverse medical records. For a recent paper submitted to IEEE Visualization 2002 about this work see PRIMA: A Case Study of Using Information Visualization Techniques for Patient Record Analysis. |
||||||||||
![]() |
The
PRIMA aggregate table shows the relative proportions of different classes
of patients in the different variables. In this example, patients with
a TRANSPLANT_TYPE of "autologous" have been colored green, and those with
a TRANSPLANT_TYPE of allogenic have been colored red.
A mouse dwell in the green box in the TRANSPLANT_TYPE variable line has resulted in the "pop-up" box indicating the contents of the box. |
|||||||||
![]() |
||||||||||
| A Kaplan-Meier survivability curve for the following classes of patients: green for autologous transplants, red for allogenic transplants, and black for "all others." The survivability curve displays the cummulative probability of survival to a certain time following transplantation. | ||||||||||
![]() |
A histogram showing the distribution of transplants over time. Once again, green represents autologous transplants, red represents allogenic transplants, and black represents "all others." Clearly the proportion of allogenic transplants has increased over time. | |||||||||