Experimental method and statistical analysis to fit tumor growth model using SPECT/CT imaging: a preclinical study

Abstract

Background: Over the last decade, several theoretical tumor-models have been developed to describe tumor growth. Oncology imaging is performed using various modalities including computed tomography (CT), magnetic resonance imaging (MRI), single-photon emission computed tomography (SPECT) and fluorodeoxyglucose-positron emission tomography (FDG-PET). Our goal is to extract useful, otherwise hidden, quantitative biophysical parameters (such as growth-rate, tumor-necrotic-factor, etc.) from these serial images of tumors by fitting mathematical models to images. These biophysical features are intrinsic to the tumor types and specific to the study-subject, and expected to add valuable information on the tumor containment or spread and help treatment plans. Thus, fitting realistic but practical models and assessing parameter-errors and degree of fit is important. Methods:We implemented an existing theoretical ode-compartment model and variants and applied them for the first time, in vivo. We developed an inversion algorithm to fit the models for tumor growth for simulated as well as in vivo experimental data. Serial SPECT/CT scans of mice breast-tumors were acquired, and SPECT data was used to segment the proliferating-layers of tumors.Results: Results of noisy data simulation and inversion show that 5 out of 7 parameters were recovered to within 4.3% error. In particular, tumor "growth-rate" parameter was recovered to 0.07% error. For model fitting to in vivo mice-tumors, regression analysis on the P-layer volume showed R 2 of 0.99 for logistic andGompertzian while surface area model yielded R weights of the models (giving their relative probability of being the best Kullback-Leibler (K-L) model among the set of candidate models) were ~0, 0.43 and 0.57 for surface-area, logistic and Gompertzian models.Conclusions: Model-fitting to mice tumor studies demonstrates feasibility of applying the models to in vivo imaging data to extract features. Akaike information criterion (AIC) evaluations show Gompertzian or logistic growth model fits in vivo breast-tumors better than surface-area based growth model. IntroductionOncology imaging is performed using various modalities including computed tomography (CT), magnetic resonance imaging (MRI), single-photon emission computed tomography (SPECT) and fluorodeoxyglucose-positron emission tomography (FDG-PET).Quantification of cancer from images is important for assessment of disease progress and treatment. Tumors grow in a specific way and theoretical tumor-models of increasing complexity have been developed that may be potentially used to find (otherwise concealed) biophysical information from serial images of tumors. If accurate, tumor growthrate, cell-motility (diffusion), tumor-necrotic-factor, can be consistently and reliably extracted from oncological images, they may add valuable information on the tumor containment/spread and help treatment plan. For example, assessing the growth-rate accurately will help in treatment planning whether in surgery or in determining the numbe...

A multi-scale model for determining the effects of pathophysiology and metabolic disorders on tumor growth

“…The total number of cells linearly increases during ATG, plateaus during angiogenesis (ATG-to-VTG), and then exponentially increases during VTG. All these four tumor population dynamics agree with experimental observations and computational analyses 19,30,55,93 and are typical of solid tumors initiated far from primary-vessels 43,56 .…”

3D printing in medicine: current applications and future directions

“…Technical developments in medical imaging techniques have led to significant improvements in the diagnostic performance of less-invasive imaging modalities such as computed tomography (CT), magnetic resonance imaging (MRI), nuclear medicine and ultrasound. Quantitative analysis of these imaging modalities allows for detection and diagnosis of various diseases with high accuracy (1)(2)(3)(4)(5)(6)(7)(8)(9)(10). Despite promising results available in the literature, traditional two-dimensional (2D) and three-dimensional (3D) visualization tools are still limited to a 2D screen, which affect realistic visualization of anatomical structures and pathologies of 3D datasets, and this is particularly apparent when dealing with complex pathologies.…”

The value of the computer-aided diagnosis system for thyroid lesions based on computed tomography images

“…The threshold of the discriminant function is placed at the midpoint between the mean values of two classes. SVM is a classical supervised learning method that performs classification tasks by constructing hyper-plane in multidimensional space (29,30). SVM constructs the optimal separation hyper-plane to data that is linearly inseparable, by mapping the data into a high-dimensional feature space in which they can be separated linearly.…”