A method of solution for a class of inverse problems involving measurement errors and its application to medical microwave radiometry

Retrieval of temperature-vs.-depth profiles in a biological tissue structure from multifrequency microwave radiometric measurement data constitutes a typical inverse problem in which the data involve relatively large measurement errors. The authors have developed a method of solution for a class of problems of this type. The method gives solutions in terms of the confidence interval and level. It also has a built-in capability of assessing the degree of fit of solutions to unknown actual source distributions. An agar phantom experiment and computer simulation based on a five-band (1-4-GHz) radiometry were performed to test the method, and the results are presented.<>