Medical imaging has been positively affected by key developments in artificial intelligence (AI) and cognitive computing in recent years. In particular, advanced AI algorithms, such as deep learning, have found numerous applications in the field of medical image processing and analysis, ranging from image reconstruction, tumor detection, feature extraction, segmentation, and classification, to wound healing assessment and revealing cardiovascular abnormalities. Other applications include the diagnosis of neurological conditions, flagging thoracic complications and conditions, physical simulation, breast imaging, and the computation of diabetic retinopathy severity. These applications have all been significantly advanced due to the substantial strides in perception made by AI. The latest developments in deep learning have shown that those tasks distinctly linked to perception, including segmentation, detection, and recognition, are well handled by the employment of state-of-the-art architectures such as computer vision, even in the absence of prior knowledge. In addition, the inherent compatibility of deep learning methods to various classical approaches and to each other makes the end-to-end training of algorithms possible by using fusion on a network level. Specifically, the convolutional neural network (CNN) architectures that have emerged from deep learning have demonstrated high-computational efficiency and enable multi-task learning by integrating with classical algorithms. Similarly, cognitive computing/intelligence, comprising of self-learning and reasoning techniques that enable computers to mimic human intelligence, has demonstrated remarkable progress in the field of clinical image analysis. They can support and augment human decision making in cases of complexity, uncertainty, and high information volume.

However, there are key barriers limiting the practical applications of AI and cognitive computing in the field of medical imaging. The regulatory process remains challenging, particularly for machine learning-based products, there is a need for large-scale validation studies to be conducted to show the performance of AI algorithms and for the underlying computing technology to be tested in clinical environments, and there is a lack of integration of the results of AI-based analysis tools into clinical workflows.

Therefore, this Special Issue aims to focus primarily on novel AI and cognitive computing algorithms and methods proposed for medical image analysis, to highlight the current challenges for practical applications of AI and cognitive computing in clinical environments, and to propose possible solutions for overcoming the barriers stopping AI becoming the mainstream in medical imaging.

Potential topics include but are not limited to the following:

• Integration of cognitive computing and AI in medical image processing
• Challenges of adopting AI-based medical image analysis
• Artificial intelligence in medical image analysis (feature extraction, segmentation, classification, and reliable measurements)
• Medical image analysis of the breast, retina, lungs, heart, nerves, musculoskeletal system, and abdomen using AI
• Feature extraction in medical imaging using machine learning/ deep learning/ convolutional neural networks
• Machine learning/deep learning/convolutional neural networks in lesion, organ, and substructure segmentation
• Machine learning-based lesion segmentation methods from MRI/CT/ultrasound images
• Object/lesion detection and classification using machine learning /deep learning/ neural networks/ evolutionary computation
• Medical image registration using artificial intelligence techniques
• Artificial intelligence in medical image reconstruction
• Cognitive computing applications in medical image processing and analysis (feature extraction, segmentation, classification, and reliable measurements)
• Artificial intelligence/cognitive science in automatic diagnostic systems
• Artificial intelligence/cognitive science in revealing cardiovascular abnormalities and diagnosis of neurological conditions