Tuberculosis diagnosis using artificial intelligence: current trends and future prospects

Area of Science:

• Medical Diagnostics
• Infectious Diseases
• Artificial Intelligence in Medicine

Background:

• Tuberculosis (TB), caused by Mycobacterium tuberculosis, remains a significant global health concern.
• China faces a high TB incidence, with challenges in microscopic detection leading to low accuracy.
• Traditional diagnostic methods like chest radiography and pathology struggle with sensitivity due to small bacilli size.

Purpose of the Study:

• To review the literature on machine learning-based models for automated tuberculosis bacilli detection.
• To highlight the advantages of integrating AI in improving TB diagnosis accuracy and sensitivity.
• To emphasize the importance of understanding TB onset and progression for effective management.

Main Methods:

• Comprehensive literature review on AI and machine learning applications in TB detection.
• Analysis of studies focusing on automated detection of Mycobacterium tuberculosis.
• Evaluation of the integration of AI tools in diagnostic workflows.

Main Results:

• Machine learning models demonstrate potential to enhance the accuracy and sensitivity of TB detection.
• AI integration can overcome limitations of traditional microscopic identification of bacilli.
• Automated detection systems offer improved diagnostic performance compared to manual methods.

Conclusions:

• AI and machine learning offer a promising approach to improve tuberculosis diagnosis.
• Enhanced detection capabilities can lead to earlier and more accurate identification of TB cases.
• Further research and integration of AI are crucial for combating the global TB epidemic.