Spatio- temporal signals, signals that take both time and space into account, are difficult to analyze and therefore rarely used in the medical field. But as spatio-temporal signals are becoming more frequently used, the need for models that can analyze the measured information is increasing.
“The goal is user friendly systems that both doctors and other medical personnel can use and interpret correctly.”
This says Jun Yu, associate professor at Centre of Biostochastics, SLU in Umeå and in charge of the project Development of Bio stochastic Methods for Analysis of Spatio- temporal Signals. The result of the project will serve as a base for other projects and will be used in fields like muscle physiology and characterization of cancer tumours. However, Yu’s work is focused on biostochastic method development.
The term stochastic means taking elements of chance into consideration. According to Yu, a big part of his work is separating noise from significant data.
“Noise, is any form of disturbance or false data” he explains, “in theory, there is no noise, but when measuring in reality it is impossible to avoid.”
Earlier and ongoing research, both within muscle physiology and the characteristics of different cancer tumours, has to some extent been limited by the methods of measuring not being accurate and reliable enough. A development of stochastic models and statistical methods enabling a noticeably higher accuracy of the spatio-temporal signals from different measuring instruments could be used in deciding on individual radiation doses for cancer treatment. Jun Yu says the project is strongly linked to other projects in research and development within radio physics and medical technology. At the moment, three people are working with Yu in the project and the interest as well as the need for what they are doing is big.
Yu describes the practical part of his work with finding new methods.
“First we normally look at existing literature” he says, “since there is no use trying something that has already failed. Then we look at the features of the current data and try to use the knowledge we have to adjust the method after the needs.”
Next step is for the method to be evaluated, for which a combination of simulation studies and real data is used.
“It’s a method of verification” Yu explains, “if it proves successful we apply the method to real data but the challenge is getting the simulated data as close to reality as possible.”
The difficulties the project is facing are mainly those complications coming from collected data, but there is constant progress and the market is eagerly waiting. Today there are several advanced methods for analysis, providing the users with more information than is possible to handle manually. These methods depend on the mathematical statistical models Yu and his co-workers are working on to be able to reach a breakthrough and be used in practice. Thus, the future rests in the hands of the project.
CMTF, c/o Tillämpad fysik och elektronik
Umeå universitet, 901 87 Umeå
Tel: +46 (0)90-786 96 38.
E-post: Britt Andersson.
