Application of artificial intelligence in brain molecular imaging

Area of Science:

• Artificial Intelligence
• Machine Learning
• Medical Imaging
• Molecular Imaging

Background:

• Artificial Intelligence (AI) and Machine Learning (ML) have roots in the mid-twentieth century.
• Significant advancements in computational resources, research, and investment are accelerating AI/ML applications in medical imaging.
• AI/ML demonstrates potential to enhance imaging operations, decision-making, and uncover complex relationships in multi-modal clinical data.

Purpose of the Study:

• To review the expanding applications of AI and ML in brain molecular imaging.
• To highlight the role of AI algorithms like Convolutional Neural Networks (CNNs).
• To discuss the potential impact of AI on patient care and the development of regulatory frameworks.

Main Methods:

• Review of current AI/ML algorithms and their applications in brain molecular imaging.
• Focus on image-based AI, including Convolutional Neural Networks (CNNs).
• Examination of specific applications such as image denoising, attenuation correction, segmentation, and disease detection.

Main Results:

• AI/ML applications in brain molecular imaging are rapidly increasing.
• Specific uses include image denoising, PET/MRI attenuation correction, segmentation, lesion detection, and parametric image formation.
• AI shows promise in detecting and diagnosing conditions like Alzheimer's disease.

Conclusions:

• AI/ML can significantly improve the quality of patient care in medical imaging.
• AI is viewed as a tool to augment, not replace, the role of radiologists.
• A regulatory framework is under development to support the integration of AI in medical imaging.