Artificial intelligence in single photon emission computed tomography (SPECT) imaging: a narrative review

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Nuclear Medicine

Background:

• Artificial intelligence (AI) has a long history in medical imaging, with applications in single-photon emission computed tomography (SPECT) dating back nearly 30 years.
• Recent advancements in AI technology have accelerated its integration and impact on SPECT imaging.
• AI's role in SPECT is expanding across various disease categories, including neurological disorders, kidney failure, cancer, and heart disease.

Purpose of the Study:

• To review and discuss the progress of AI technology specifically within SPECT imaging.
• To provide a comprehensive overview of AI applications in SPECT for researchers.
• To highlight the evolution of AI integration with physics and traditional methods in SPECT.

Main Methods:

• Review of AI applications in disease prediction and diagnosis.
• Analysis of AI for post-reconstruction image denoising in SPECT.
• Examination of AI in attenuation map generation and image reconstruction for SPECT.

Main Results:

• AI applications are diverse, covering disease prediction, diagnosis, image denoising, attenuation correction, and reconstruction.
• AI integration enhances SPECT imaging by addressing limitations of conventional methods.
• AI demonstrates significant success in improving SPECT diagnostic capabilities and image quality.

Conclusions:

• AI is critical for the advancement of modern SPECT technology, offering solutions to conventional imaging deficiencies.
• AI applications in SPECT have shown unparalleled success in various medical fields.
• Challenges and limitations of AI in SPECT imaging are acknowledged, with future directions and countermeasures proposed.