Performance comparison of instrument automation pipelines using different programming languages.

Area of Science:

• Medical Imaging
• Computer Science
• Engineering

Background:

• Automated 2D ultrasound computerized tomography (USCT) systems require efficient programming for complex tasks.
• Selecting the optimal programming language is crucial for system performance and resource optimization.

Purpose of the Study:

• To conduct a performance analysis of a fully automated, in-house developed 2D USCT system across four programming languages: LabVIEW, MATLAB, C, and Python.
• To evaluate launch performance, eight performance indices, and runtime performance for each language.
• To provide a methodology for selecting the best programming language for instrument automation.

Main Methods:

• Developed a fully automated 2D USCT system with integrated codes for sensors, instrument interfacing, real-time control, data acquisition, processing, and analysis.
• Implemented the system in LabVIEW, MATLAB, C, and Python.
• Assessed performance using launch metrics, eight indices, and runtime analysis, with and without data processing.

Main Results:

• C demonstrated the lowest processing power consumption and fewest I/O operations.
• For runtime analysis (data acquisition and real-time control), LabVIEW was fastest (365.69 s), followed by MATLAB (623.83 s), C (1252.03 s), and Python (1505.54 s).
• With data processing, MATLAB showed the best performance (640.33 s), outperforming LabVIEW (731.91 s), Python (1520.01 s), and C (1930.15 s). Python excelled in interfacing speed and RAM usage.

Conclusions:

• The choice of programming language significantly impacts the performance of automated 2D USCT systems.
• MATLAB is optimal for tasks involving data processing, while LabVIEW excels in real-time control and data acquisition.
• C offers efficiency in processing power and I/O, and Python provides advantages in interfacing and memory management, informing resource optimization strategies.