Automatic microscopic detection of mycobacteria in sputum: a proof-of-concept

Area of Science:

• Microbiology
• Medical Diagnostics
• Optical Engineering

Background:

• Standard Ziehl-Neelsen staining for diagnosing lung mycobacterioses is operator-dependent and its sensitivity varies.
• Microscopic examination of sputum smears is crucial for identifying acid-fast bacilli, but is limited by subjectivity and observation time.

Purpose of the Study:

• To develop and evaluate an operator-independent automated microscopic examination for detecting and enumerating acid-fast bacilli in sputum smears.
• To assess the sensitivity, specificity, and accuracy of the automated system compared to standard microscopic examination.

Main Methods:

• Development and in-house parameterization of an automated microscope (ZEISS Axio Scan.Z1) and counting software.
• Evaluation of the limit of detection using seeded sputum samples with known concentrations of Mycobacterium bovis BCG or Mycobacterium tuberculosis H37Rv.
• Testing the automated system on 93 clinical sputum specimens (34 smear-positive, 59 smear-negative).

Main Results:

• The automated system achieved a limit of detection of 10^2 bacilli/mL with 100% positivity rate.
• Sensitivity was 97.06% and specificity was 86.44% when compared to standard microscopy.
• The system can process up to 100 slides and export results to laboratory information systems.

Conclusions:

• Automated microscopy provides a reliable and operator-independent method for detecting acid-fast bacilli in sputum.
• This technology has the potential to enhance the efficiency and accuracy of diagnosing lung mycobacterioses in routine laboratory workflows.
• The protocol is being integrated into routine workflow, with positive results confirmed by standard microscopy.