Predicting Cancer Cells Progression Via Entropy Generation Based on Ar and Arma Models Publisher



Modaresi Movahed T^{1, 2} ; Jalaly Bidgoly H³ ; Khoshgoftar Manesh MH⁴ ; Mirzaei HR⁵ Show Affiliations
Source: International Communications in Heat and Mass Transfer Published:2021

Abstract

The living cell can be described as a thermodynamic system whose structure and behavior arise from the thermodynamic processes and reactions within the cell and between the cell and environment. The origin and growth of cancer is also a thermodynamic process during which several thermodynamic variables change: cell volume, temperature, and entropy. It has been proved that cancer cells have a higher entropy generation than healthy cells. The entropy generation rate shows the intensity of proliferation, invasion, and robustness in cancer cells. Therefore, entropy dynamic modeling and this thermodynamic concept can be an efficient combined tool in cancer research. Besides, dynamic modeling of tumor volume assists in prognosticating metastasis. The examination of tumor temperature gradient confirms the applicability of thermal imaging in diagnosing cancer and analyzing its growth as a non-expensive and non-invasive tool. This paper aimed to investigate and estimate the dynamic changes in the volume, temperature gradient, and entropy generation during the growth of the cancer cells. The required experimental data for modeling the process were obtained in the laboratory, for which some 4T1 breast tumor cells were injected into several mice. The variations in tumor volume and temperature gradient were measured based on the start of injection until the death of mice. The tumor entropy generation was calculated based on the volume and temperature gradient of the tumor using a nonlinear relation. The obtained data were used to model the dynamic changes in these three variables using the AR and ARMA linear models. The observations indicated an increase in the volume, temperature gradient, and entropy generation of the cancer cells proportional to cancer growth. The developed models based on the recorded data are highly accurate and can predict the growth rate of the mentioned variables during the progression and invasion stages. The modeling results indicated that in volume modeling with AR (5), the accuracy reached 78.66%; however, with ARMA (1,1), it was 79.75%. For temperature gradient predicting, the accuracy with AR (5) was 79.7% and 79.56% with ARMA (2,2). Finally, for entropy identification, the accuracy was 78.56% with AR (5), while it was 80.93% with ARMA (1,2). As expected, the accuracy increased using the ARMA model, which is more complex than the AR. © 2021 Elsevier Ltd