Thermographic analysis of the abdominal region of pregnant women

Abstract

Thermography’s use in the biomedical field is becoming increasingly popular and is also providing promising results in numerous medical fields such as breast screening and early detection of diabetes. Although the ultimate aim of this project would be an investigation on foetal thermography, one has to first run a preliminary study to understand the characteristics of the abdomen for both pregnant and non-pregnant subjects. The objectives of this project were to try and find a suitable acclimatisation period for the abdomen and to try and minimise the amount of time which the subject spends in front of the camera. In order to examine the temperature variations, a series of thermal images were acquired and two unforeseen circumstances were discovered. The first one is the challenge in thermal tracking which is difficult partly due to the subject’s movement and partly due to homogenous patterns across the abdomen. The best tracking performance was obtained using the summation of squared differences method. The second unanticipated situation was the fact that most studies assume an acclimatisation period of 15 to 20 minutes, even though a decrease in temperature and other variations were still present after that period. There are no reliable studies on the indication of the suitability of the acclimatisation period which consider more than 20 minutes of acclimatisation. To carry out the second objective, different mathematical models and algorithms were explored to provide proper curve fitting that satisfies both the dynamic characteristics of the data across time and the final temperature estimation. The RMSE for the dynamic characteristics is 0.1 or less whilst the RMSE for the final temperature estimation is less than 0.05, when considering between 70 and 100 minutes of data. The temperature difference between different quadrants across the abdomen shows statistical significant difference in all cases except one. This shows that most of the temperature differences are not just by chance and hence the underlying structures of the abdomen clearly affect the abdomen’s temperature. This kind of investigation on temperature changes has never been undertaken and although there is room for improvement, this dissertation shows that thermography can be a useful tool for future studies about foetal thermography.