In 2019 National Cancer Institute (NCI) in the USA ranked lung and bronchus cancer as the second diagnosis of cancer types. It is important to mention that only a few studies have focused on lung and bronchus cancer patient's survival time by using the SEER database via Machine Learning techniques. This Thesis intends to develop a Machine Learning Approach to classify survivability (dead or survived), and in addition to classification, aims to predict the remaining lifespan for the patients who predicted would die within five years. In the first step, nine Machine Learning techniques, including Logistic Regression, Support Vector Machine, K-Nearest Neighbor, Naïve Bayes Classifier, Ensemble Max Voting, Stacking Ensemble, Random Forest, Gradient Boosting Machine, Adaboost, along with a proposed Deep Neural Network are applied to predict whether the patients would die or survive after five years. In the next step, we use another Deep Neural Network for regression for the patients who are predicted to die for actual survival prediction. The results show that the proposed Deep Neural Network outperformed other Machine learning techniques.

This book introduces a variety of advanced machine learning approaches covering the areas of neural networks, fuzzy logic, and hybrid intelligent systems for the determination and diagnosis of cancer. Moreover, the tactical solutions of machine learning have proved its vast range of significance and, provided novel solutions in the medical field for the diagnosis of disease. This book also explores the distinct deep learning approaches that are capable of yielding more accurate outcomes for the diagnosis of cancer. In addition to providing an overview of the emerging machine and deep learning approaches, it also enlightens an insight on how to evaluate the efficiency and appropriateness of such techniques and analysis of cancer data used in the cancer diagnosis. Therefore, this book focuses on the recent advancements in the machine learning and deep learning approaches used in the diagnosis of different types of cancer along with their research challenges and future directions for the targeted audience including scientists, experts, Ph.D. students, postdocs, and anyone interested in the subjects discussed.

Lung cancer is a major cause of cancer-related deaths worldwide. Machine learning techniques have shown promising results in the early detection and prediction of lung cancer. However, high-dimensional data, such as gene expression profiles, can introduce noise and decrease the classification accuracy of machine learning models. Feature selection techniques can alleviate this issue by identifying the most relevant and informative features, leading to better model performance. Optimized feature selection techniques can enhance the prediction accuracy of lung cancer using machine learning algorithms. Support vector machines, random forest, and artificial neural networks are commonly used algorithms for lung cancer prediction. By optimizing feature selection, these models can be trained with the most informative features, reducing overfitting and improving classification accuracy. Cross-validation techniques can also be used to evaluate the performance of feature selection and machine learning algorithms. The integration of optimized feature selection with machine learning techniques can provide an accurate and reliable lung cancer prediction model, which has the potential to improve early detection and precision medicine for lung cancer patients. Overall, optimized feature selection for enhancing lung cancer prediction using machine learning techniques is a promising approach to improving patient outcomes and reducing the global burden of lung cancer.

The human metabolome represents a largely unexplored area with respect to prediction of disease. I have conducted a targeted study of metabolic compounds in the human metabolome that are linked to patients with non small cell lung cancer. Through the use of machine learning techniques such as SVMs, Random Forests, and Decision Trees, I have determined models can be trained to correctly classify new patients with an F1--score above 0.95 in case vs. control classification. From these models I have produced a select subset of compounds using peak analysis as well as recursive feature extraction such that when the prediction is done on this smaller subset of compounds an F1--score of above 0.95 can still be achieved. These compounds represent potential biomarkers for future studies and clinical applications.

Lung cancer has a high incidence and mortality rate. The five-year relative survival rate for all lung cancers is 18%. Due to the high mortality and incidence rate of lung cancer worldwide, early detection is essential. Low dose Computed Tomography (CT) is a commonly used technique for screening, diagnosis, and prognosis of non-small cell lung cancer (NSCLC). The National Lung Screening Trial (NLST) compared low-dose helical computed tomography (LDCT) and standard chest radiography (CXR) for three annual screens and reported a 20% relative reduction in lung cancer mortality for LDCT compared to CXR. As such, LDCT screening for lung cancer is an effective way of mitigating lung cancer mortality and is the only imaging option for those at high risk. Lung cancer screening for high-risk patients often detects a large number of indeterminate pulmonary nodules, of which only a subset will be identified as cancer. As such, reliable and reproducible biomarkers determining which indeterminate pulmonary nodules will be identified as cancer would have significant translational implications as a therapeutic method to enhance lung cancer screening for nodule detection. Radiomics is an approach to extract high-dimensional quantitative features from medical images, which can be used individually or merged with clinical data for predictive and diagnostic analysis. Quantitative radiomics features (size, shape, and texture) extracted from lung CT scans have been shown to predict cancer incidence and prognosis. Deep learning is an emerging machine learning approach, which has been applied to the classification and analysis of various cancers and tumors. To generate generic features (blobs, edges, etc.) from an image, different convolutional kernels are applied over the input image, and then those generic feature-based images are passed through some fully connected neural layers. This category of the neural network is called a convolutional neural network (CNN), which has achieved high accuracy on image data. With the advancement of deep learning and convolutional neural networks (CNNs), deep features can be utilized to analyze lung CTs for prognosis prediction and diagnosis. In this dissertation, deep learning-based approaches were presented for lung nodule malignancy prediction. A subset of cases from the NLST was chosen as a dataset in our study. We experimented with three different pre-trained CNNs for extracting deep features and used five different classifiers. Experiments were also conducted with deep features from different color channels of a pre-trained CNN. Selected deep features were combined with radiomics features. Three CNNs were designed and trained. Combinations of features from pre-trained, CNNs trained on NLST data, and classical radiomics were used to build classifiers. The best accuracy (76.79%) was obtained using feature combinations. An area under the receiver operating characteristic curve of 0.87 was obtained using a CNN trained on an augmented NLST data cohort. After that, each of the three CNNs was trained using seven different seeds to create the initial weights. These enabled variability in the CNN models, which were combined to generate a robust, more accurate ensemble model. Augmenting images using only rotation and flipping and training with images from T0 yielded the best accuracy to predict lung cancer incidence at T2 from a separate test cohort (Accuracy = 90.29%; AUC = 0.96) based on an ensemble 21 models. From this research, five conclusions were obtained, which will be utilized in future research. First, we proposed a simple and effective CNN architecture with a small number of parameters useful for smaller (medical) datasets. Second, we showed features obtained using transfer learning with all the channels of a pre-trained CNN performed better than features extracted using any single channel and we also constructed a new feature set by fusing quantitative features with deep features, which in turn enhanced classification performance. Third, ensemble learning with deep neural networks was a compelling approach that accurately predicted lung cancer incidence at the second screening after the baseline screen, mostly two years later. Fourth, we proposed a method for deep features to have a recognizable definition via semantic or quantitative features. Fifth, deep features were dependent on the scanner parameters, and the dependency was removed using pixel size based normalization.

The LNAI 12299 constitutes the papers of the 18th International Conference on Artificial Intelligence in Medicine, AIME 2020, which will be held online in August 2020. The 42 full papers presented together with 1short papers in this volume were carefully reviewed and selected from a total of 103 submissions. The AIME 2020 goals were to present and consolidate the international state of the art of AI in biomedical research from the perspectives of theory, methodology, systems, and applications.