Cancer therapy optimization based on multiple model adaptive control. (February 2019)

• Record Type: Journal Article
• Authors: Teles, Francisco F.; Lemos, João M.
• Is Part Of: Biomedical signal processing and control. Volume 48(2019)
• Page Start: 255
• Page End: 264
• Publication Date: 2019-02
• Subjects: Cancer therapy design -- Tumor growth model -- Anti-angiogenesis -- Immunotherapy -- Multiple model adaptive control -- Model clustering; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2018.09.016
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Cancer therapy optimization through multiple model adaptive control. Model clustering for multiple model adaptive control. Innovative model combining tumor growth, anti-angiogenesis and immunotherapy. Abstract: During the last years several clinical decision support systems have been developed, some of which clearly improved the results obtained with standard clinical practice. However, this kind of decision computing process has not been quite explored in cancer treatment. In this work a control system that designs an optimal therapy based on adaptive control methods, aiming to allow the eradication of a metastatic renal cell carcinoma as quickly and efficiently as possible, and with lower associated toxicity, is developed. In order to do so, a new mathematical model describing the growth of this kind of tumor is developed, taking into account the effects of two of the most promising therapies: anti-angiogenesis and immunotherapy. Additionally, models describing pharmacodynamical aspects of the organism are also included. The therapy is designed through multiple model adaptive control. Together with a system of selection and aggregation of key classes of models, it allows to deal with the uncertainty associated with the patient, namely his intra- and inter-patient variability. The simulation results show that the approach proposed presents robustness in terms of stability and performance. The reference tracking errors for the simulations are around 3%, which allows a tumor eradication in less than a year and a half with mild and moderate toxicity levels. Therefore, the tool developed allows to contribute with new perspectives in the creation of decision support systems for cancer therapy, thus enhancing the medical choices.