Exercise Test Database Analysis Software

Janne Kallio^ba, Rami Lehtinen^af, Jari Viik^b, Tiit Kööbi^a, Kari Niemelä^d,
Terho Lehtimäki^e, Väinö Turjanmaa^ag

^aDepartment of Clinical Physiology, Tampere University Hospital,
^bRagnar Granit Institute, Tampere University of Technology,
^cDivision of Cardiology, Dept. of Internal Medicine, Tampere University Hospital,
^dDepartment of Clinical Chemistry, Tampere University Hospital,
^eTampere Polytechnic, Tampere, Finland

Correspondence: Msc. Janne Kallio, Ragnar Granit Institute, Tampere University of Technology, P.O. Box 692, FIN-33101 Tampere, Finland. E-mail: janne.kallio@tut.fi, phone +358 405239293

1.        Introduction

The digital environment brings new possibilities to ECG and exercise test information analysis. There is a considerable amount of information gathered from a patient during operations in the hospital. For utilization in research projects, it means a need for highly sophisticated databases and software. [Kallio, 2001][Viik 2000]

Exercise test database has been gathered in the Tampere University Hospital since October 2^nd, 2001. There are 1147 patients in the database as of January 17th, 2003. The information consists of all the information measured or asked from the patient. There are measurements during the test, ECG-signals, doctors valuations based on the test, and patient characteristics of the patient lifestyle.

 For researchers, this database offers possibilities to find a proper patient group and analyze their ECG-signals. To help this task an analysis software has been developed. With the software the user can choose the desired group of patients based on factors that are collected during the test. This way, the researcher can focus on any material that database contains.

For hospital administration the database offers the possibility to monitor the tests that are performed in research laboratories. Statistics from the daily work can be easily displayed and used on outlining the necessary resources.

2.        Material and Methods

2.1. Database

The database contains information captured from the patient before, during, and after the test. Before the test, the patient fills out a questioner regarding his life style. Questions include for example, alcohol consumption and amount of weekly exercise. During the test the nurse and doctor uses GETID software to save necessary information from the patient. There are approximately 100-200 information items about the test filled during the test, Such as blood pressure and wmax6. The doctor also makes his valuation of patient condition and inputs it to the GETID software. From GETID all the information goes to research database. During the test there is also a 14-lead ECG recorded from the patient with Cardiosoft^â software (GE Marguette Hellige). The ECG filenames are stored to the database and some ECG characters are exported to the database. After the test there is a blood sample taken from the patient from which DNA information is captured. There will also be angiography information gathered to the database afterwards. The description of the database is shown in Fig 1.

The variety of information that is saved to the database is connected to each other by a unique exercise test code. There is new code given to the patient even if he/she has taken part of the test before. Afterwards, there can be additional information inserted to the database about the patients if required.

Figure 1. Description of database. There are several tables that include different information about the patient. Every patient has a unique code, which can be found in every table and this way identifies this information.

2.2. Analysis Software for the Database

The purpose of the software is to give the researcher a possibility to find the necessary patient group. The researcher can mark the limits of different characters, and this way give the user a possibility to use proper patient information. Information can then be exported as in comma-separated values (CSV) if necessary. Most of the statistical software can import this CSV format. Graphical user interface with the characteristics of test information that the user can adjust to find the proper research group is shown in figure 2. There is a similar graphical interface for electrocardiography parameters and information of the patients’ behavior.

Figure 2. Graphical Interface for database. Each factor can be adjusted.

ECG data files can also be listed based on test information, ECG parameters, and characteristics of the patients’ behavior. This gives the researcher the possibility to filter the necessary ECG data files based on patient information. Since ECG data is in a raw format, it is possible to test the performance of different kinds of signal processing algorithms.

3.        Results

The software gives the user an efficient way of adjusting the selection of patients. After the selection process, the researcher can export this information to the statistical software or utilize the ECG files for the given patient group. Selection of the proper patient group is done with a visual interface. It is also possible for researchers to try and test the meaning of different factors for patients.

For hospital administration the software offers information straight from the database. This helps the supervision of the exercise test laboratory and makes it easier to target the resource effectively.

4.        Discussion

Exercise test database analysis software offers the researcher an easy to use solution for selection of a particular patient group. This information isn’t real-time since the information is not connected to the intra/internet. Further development goals could be to make the database available over the intranet. This would help the administration to supervise the exercise test labs in real time.

For hospital administration, a basic interface in internet would make it possible to make statistical reports of an exercise test. With statistical reports the hospital administration has the possibility to monitor the exercise test procedures and the personnel work.

Acknowledgements

The financial support of the Ragnar Granit Foundation, Medical Research Fund of Tampere University Hospital and Finnish Cultural Foundation are gratefully acknowledged.

References

Viik J, Vänttinen H, Malmivuo J. ECG variable cine: computer program for presentation of temporal changes in ECG variables over different number of ECG leads. Comput Methods Programs Biomed.2000.

Kallio J. Software for Acquisition of Exercise Test Database, Master of Science Thesis, 2001.