Search

The static and dynamic characteristics of carbonate reservoirs are very challenging to sedimentologists, geologists, petrophysicists, and reservoir engineers. The multi-scale and multi-discipline nature of carbonate characterization complicates any global rock typing workflow that aims to satisfy all disciplines and datasets. The heterogeneity of pore space and pore network systems obscures the understanding of the interaction between rock, pore and fluid, and hence there is no simple relationship between pore type distribution and depositional environment for carbonate reservoirs. The challenge is to accurately assign the complex pore system properties to their sedimentological rock container for vertical and lateral predictability. In this thesis, a novel Ternary Rock Typing (TRT) application and workflow combines depositional facies, wireline logs, routine core analysis, capillary pressure data and relative permeability curves from three parameters representing rock, pore and fluid interaction. The three parameters proposed are porosity, permeability and irreducible water saturation represented in the shape of a 3D ellipsoid. The complex interaction of rock, pore and fluid affect the location, shape, orientation and relative position of the three-parameter ellipsoid. The quality control of derived rock types by saturation height functions becomes an integral part of defining a Ternary Rock Type (TRT). A Ternary Rock Typing (TRT) methodology using TRT software was developed as part of this thesis was tested against a synthetic data set. It was then tested on 1D and 3D domains using 6 cored wells from an actual carbonate field. The data included sedimentological core description, wireline logs, routine core analysis, and SCAL data. The utilization of TRT is subdivided into several applications: rock typing scheme, concept, workflow, plot and software. All of the previous applications can be adapted to work with TRT or accompanied by any of the other rock typing methods. The TRT proved to perform well in 1D and 3D static modelling environments where it proved to predict the oil in place for the limestone and dolostone of the highest storage and flow capacity rock types compared to the reference based model. The TRT technique can be used as a raw unguided solely data driven approach (rTRT) or as a guided "sedimentologically" data driven approach (gTRT). The guided TRT (gTRT) uses seed points from the centre of ellipsoids representing "sedimentological facies" for the ternary rock typing algorithm to predict the likelihood for rock typing clusters and thus paves a link between sedimentological facies, wireline logs, RCA and SCAL data. The TRT was also tested against two simulation models, the first of which is an active water drive and the second is an injector/producer scheme with no water drive. The TRT model also proved to accurately model barriers, baffles and flow units. Thus, it gave the best fluid movements compared to other rock typing schemes. This affects the production and completion strategy. It also affects water breakthrough, fingering, channelling, and by passed oil. The impact of changing rock typing scheme on static and dynamic models is seldom quantified, as it creates too great a number of realizations often not feasible to dynamically simulate. A subset is usually chosen for the simulation testing. The TRT workflow disregards rock typing schemes that are not satisfying the static and dynamic interrelationships, and hence minimizes the effort of simulating unnecessary models.

Machine learning (ML) and deep learning (DL) approaches have been used as indispensable tools in modern artificial intelligence-based computer-aided diagnostic (AIbased CAD) systems that can provide non-invasive, early, and accurate diagnosis of a given medical condition. These AI-based CAD systems have proven themselves to be reproducible and have the generalization ability to diagnose new unseen cases with several diseases and medical conditions in different organs (e.g., kidneys, prostate, brain, liver, lung, breast, and bladder). In this dissertation, we will focus on the role of such AI-based CAD systems in early diagnosis of two kidney diseases, namely: acute rejection (AR) post kidney transplantation and renal cancer (RC). A new renal computer-assisted diagnostic (Renal-CAD) system was developed to precisely diagnose AR post kidney transplantation at an early stage. The developed Renal-CAD system perform the following main steps: (1) auto-segmentation of the renal allograft from surrounding tissues from diffusion weighted magnetic resonance imaging (DW-MRI) and blood oxygen level-dependent MRI (BOLD-MRI), (2) extraction of image markers, namely: voxel-wise apparent diffusion coefficients (ADCs) are calculated from DW-MRI scans at 11 different low and high b-values and then represented as cumulative distribution functions (CDFs) and extraction of the transverse relaxation rate (R2*) values from the segmented kidneys using BOLD-MRI scans at different echotimes, (3) integration of multimodal image markers with the associated clinical biomarkers, serum creatinine (SCr) and creatinine clearance (CrCl), and (4) diagnosing renal allograft status as nonrejection (NR) or AR by utilizing these integrated biomarkers and the developed deep learning classification model built on stacked auto-encoders (SAEs). Using a leaveone- subject-out cross-validation approach along with SAEs on a total of 30 patients with transplanted kidney (AR = 10 and NR = 20), the Renal-CAD system demonstrated 93.3% accuracy, 90.0% sensitivity, and 95.0% specificity in differentiating AR from NR. Robustness of the Renal-CAD system was also confirmed by the area under the curve value of 0.92. Using a stratified 10-fold cross-validation approach, the Renal-CAD system demonstrated its reproduciblity and robustness with a diagnostic accuracy of 86.7%, sensitivity of 80.0%, specificity of 90.0%, and AUC of 0.88. In addition, a new renal cancer CAD (RC-CAD) system for precise diagnosis of RC at an early stage was developed, which incorporates the following main steps: (1) estimating the morphological features by applying a new parametric spherical harmonic technique, (2) extracting appearance-based features, namely: first order textural features are calculated and second order textural features are extracted after constructing the graylevel co-occurrence matrix (GLCM), (3) estimating the functional features by constructing wash-in/wash-out slopes to quantify the enhancement variations across different contrast enhanced computed tomography (CE-CT) phases, (4) integrating all the aforementioned features and modeling a two-stage multilayer perceptron artificial neural network (MLPANN) classifier to classify the renal tumor as benign or malignant and identify the malignancy subtype. On a total of 140 RC patients (malignant = 70 patients (ccRCC = 40 and nccRCC = 30) and benign angiomyolipoma tumors = 70), the developed RC-CAD system was validated using a leave-one-subject-out cross-validation approach. The developed RC-CAD system achieved a sensitivity of 95.3% ± 2.0%, a specificity of 99.9% ± 0.4%, and Dice similarity coefficient of 0.98 ± 0.01 in differentiating malignant from benign renal tumors, as well as an overall accuracy of 89.6% ± 5.0% in the sub-typing of RCC. The diagnostic abilities of the developed RC-CAD system were further validated using a randomly stratified 10-fold cross-validation approach. The results obtained using the proposed MLP-ANN classification model outperformed other machine learning classifiers (e.g., support vector machine, random forests, and relational functional gradient boosting) as well as other different approaches from the literature. In summary, machine and deep learning approaches have shown potential abilities to be utilized to build AI-based CAD systems. This is evidenced by the promising diagnostic performance obtained by both Renal-CAD and RC-CAD systems. For the Renal- CAD, the integration of functional markers extracted from multimodal MRIs with clinical biomarkers using SAEs classification model, potentially improved the final diagnostic results evidenced by high accuracy, sensitivity, and specificity. The developed Renal-CAD demonstrated high feasibility and efficacy for early, accurate, and non-invasive identification of AR. For the RC-CAD, integrating morphological, textural, and functional features extracted from CE-CT images using a MLP-ANN classification model eventually enhanced the final results in terms of accuracy, sensitivity, and specificity, making the proposed RC-CAD a reliable noninvasive diagnostic tool for RC. The early and accurate diagnosis of AR or RC will help physicians to provide early intervention with the appropriate treatment plan to prolong the life span of the diseased kidney, increase the survival chance of the patient, and thus improve the healthcare outcome in the U.S. and worldwide.

Early diagnosis of acute renal transplant rejection (ARTR) is of immense importance for appropriate therapeutic treatment administration. Although the current diagnostic technique is based on renal biopsy, it is not preferred due to its invasiveness, recovery time (1-2 weeks), and potential for complications, e.g., bleeding and/or infection. In this thesis, a computer-aided diagnostic (CAD) system for early detection of ARTR from 4D (3D + b-value) diffusion-weighted (DW) MRI data is developed. The CAD process starts from a 3D B-spline-based data alignment (to handle local deviations due to breathing and heart beat) and kidney tissue segmentation with an evolving geometric (level-set-based) deformable model. The latter is guided by a voxel-wise stochastic speed function, which follows from a joint kidney-background Markov-Gibbs random field model accounting for an adaptive kidney shape prior and for on-going visual kidney-background appearances. A cumulative empirical distribution of apparent diffusion coefficient (ADC) at different b-values of the segmented DW-MRI is considered a discriminatory transplant status feature. Finally, a classifier based on deep learning of a non-negative constrained stacked auto-encoder is employed to distinguish between rejected and non-rejected renal transplants. In the “leave-one-subject-out” experiments on 53 subjects, 98% of the subjects were correctly classified (namely, 36 out of 37 rejected transplants and 16 out of 16 nonrejected ones). Additionally, a four-fold cross-validation experiment was performed, and an average accuracy of 96% was obtained. These experimental results hold promise of the proposed CAD system as a reliable non-invasive diagnostic tool.