The Reply

Abstract

We agree with the comments in the letter to the editor from Norris and Raza. The areas where artificial intelligence (AI) can be transformative is beyond what we summarized. 1 It is important to mention that the field of AI has been especially transformative in areas where clinical imaging is well integrated into the clinical practice, areas such as radiology, pathology, cancer, and surgery.The AI transformation is mainly a result of the push from major high-tech companies and has strong roots in the gaming industry, face recognition, personalized advertising, and more recently, in traffic monitoring systems and autonomous driving among others. Machine learning, and more specifically, deep-learning models are designed for use in imaging applications and have more recently been used in other areas as well, including data augmentation to improve access to otherwise constrained, 2 or limited, 3,4 clinical data. Today's high hope to use machine learning in the medical field is driven by 1) large repositories of digitalized images; 2) easier access to connect various data modality to complement insight gained from imaging and clinical features of the patient; 3) prospect of availability of ultrafast 5G connectivity; and 4) advanced computing infrastructures (such as federated computing, fog computing, etc.), platforms, and standards. Furthermore, augmented reality, another application of AI, which also uses data from wearables sensors, could have a key role in advancing the field of surgery, medical education, mental health, and addiction, etc. However, unlike nonclinical applications of machine learningfor instance, gaming-reaching the most clinically valuable results, such as diagnostic accuracy, is more important than understanding the logic behind the algorithm's decision.As it was highlighted in the letter to the Editor, if intelligent systems can one day outperform expert pathologists or radiologists, then the future for digital histology will be different. However, we would like to think that if we do it right, technology can be a tool to assist physicians through a digital deliberation, giving them back their valuable patient-physician time and helping them take better care of their patients by focusing on the patients rather than the computer screen. And maybe when we think of our next generation of physicians, we may want to invest in training programs where the focus is on the comprehensive wellbeing of the patients. Maybe we should invest in promoting a field of medicine that confronts the transformational challenges that are in part due to the digitalization of health care, paving the way to more successful adoption of novel technologically driven solutions.