The recent boom of Machine Learning Network Architectures like Generative Adversarial Networks (GAN), Deep Convolution Generative Adversarial Networks (DCGAN), Self Attention Generative Adversarial Networks (SAGAN), Context Conditional Generative Adversarial Networks (CCGAN) and the development of high-performance computing for big data analysis has the potential to be highly beneficial in many domains and fittingly in the early detection of chronic diseases. The clinical heterogeneity of one such chronic auto-immune disease like Systemic Lupus Erythematosus (SLE), also known as Lupus, makes it difficult for medical diagnostics. One major concern is a limited dataset that is available for diagnostics. In this research, we demonstrate the application of Generative Adversarial Networks for data augmentation and improving the error rates of Convolution Neural Networks (CNN). Limited Lupus dataset of 30 typical ’butterfly rash’ images is used as a model to decrease the error rates of a widely accepted CNN architecture like Le-Net. For the Lupus dataset, it can be seen that there is a 73.22% decrease in the error rates of Le-Net. Therefore such an approach can be extended to most recent Neural Network classifiers like ResNet. Additionally, a human perceptual study reveals that the artificial images generated from CCGAN are preferred to closely resemble real Lupus images over the artificial images generated from SAGAN and DCGAN by 45 Amazon MTurk participants. These participants are identified as ’healthcare professionals’ in the Amazon MTurk platform. This research aims to help reduce the time in detection and treatment of Lupus which usually takes 6 to 9 months from its onset.

The past decade has witnessed rapid advances at the intersection of machine learning and medicine. Owing to the tremendous amount of digitized hospital data, machine learning is poised to bring innovation to the traditional healthcare workflow. Though machine learning models have strong predictive power, it is challenging to translate a research project into a clinical tool partly due to the lack of a rigorous validation framework. In this dissertation, I presented a range of machine learning models that were trained to classify Alzheimer’s disease - a condition with an insidious onset - using routinely collected clinical data. In addition to reporting the model performance, I discussed several considerations, including feature selection, data harmonization, effect of confounding variables, diagnostic scope, model interpretability and validation, which are critical to the design, development, and validation of machine learning models. From the methodological standpoint, I presented a multidisciplinary collaboration in which medical domain knowledge which was obtained from experts and tissue examinations was tightly integrated with the interpretable outcomes derived from our machine learning frameworks. I demonstrated that the model, which generalized well on multiple independent cohorts, achieved diagnostic performance on par with a group of medical professionals. The interpretable analysis of our model showed that its underlying decision logic corresponds with expert ratings and neuropathological findings. Taken together, this work presented a machine learning system for classification of Alzheimer’s disease, marking an important milestone towards a translatable clinical application in the future.