This book, written by authors with more than a decade of experience in the design and development of artificial intelligence (AI) systems in medical imaging, will guide readers in the understanding of one of the most exciting fields today.
After an introductory description of classical machine learning techniques, the fundamentals of deep learning are explained in a simple yet comprehensive manner. The book then proceeds with a historical perspective of how medical AI developed in time, detailing which applications triumphed and which failed, from the era of computer aided detection systems on to the current cutting-edge applications in deep learning today, which are starting to exhibit on-par performance with clinical experts.
In the last section, the book offers a view on the complexity of the validation of artificial intelligence applications for commercial use, describing the recently introduced concept of software as a medical device, as well as good practices and relevant considerations for training and testing machine learning systems for medical use. Open problematics on the validation for public use of systems which by nature continuously evolve through new data is also explored.
The book will be of interest to graduate students in medical physics, biomedical engineering and computer science, in addition to researchers and medical professionals operating in the medical imaging domain, who wish to better understand these technologies and the future of the field.
- An accessible yet detailed overview of the field
- Explores a hot and growing topic
- Provides an interdisciplinary perspective
Table of Contents
Chapter 1. Fundamentals of Machine Learning
Chapter 2. Introduction to Deep Learning
Chapter 3. Applying AI in Medical Imaging
Chapter 4. Designing AI Systems for the Clinical Practice
Chapter 5. Future Perspectives
Lia Morra holds a Ph.D. degree in computer engineering from Politecnico di Torino, Italy. In 2006, she joined im3D to develop computer-aided diagnosis solutions for early cancer detection and screening. From 2014 to 2017, she served as im3D Chief Scientific Officer, overseeing technology development up to clinical testing and regulatory approval. In this book, she shares her decade-long experience in medical imaging research development. Currently, her research focuses on artificial vision and machine learning, as well as their applications from industry to healthcare. She has authored numerous papers and patents in the engineering, computer science and radiology domains.
Silvia Delsanto begins her experience in biomedical image analysis during her Ph.D. at the Turin Polytechnical Institute, where she develops systems for automated measurements on vascular US images. In the following years, she develops computer vision and machine learning algorithms integrated in commercial Computer Aided Detection systems and cooperates in the design and management of large-scale clinical validation studies. She leads the im3D Research from 2017 to 2018. Currently, she has expanded her research interests to other industrial applications. She has authored numerous papers in the engineering, radiology, and clinical trial domains.
Loredana Correale received her PhD in physics from the University of Roma "La Sapienza". She has studied models for disordered systems, spin glasses, statistical and combinatorial inference. Since 2006, she has been a researcher at im3D. Her research interests include medical imaging computer-aided diagnosis (CAD), and data-mining in various modalities such as breast and colon/CTC imaging. Within im3D, she actively contributes to the design and the analysis of large scale imaging trials to get new insight in the advantages and challenges of the application of CAD systems in clinical population. She has authored numerous papers in oncology, radiology, and clinical trial domains.