Initial findings creating a temperature prediction model using vibroacoustic signals originating from tissue needle...

Area of Science:

• Biomedical Engineering
• Acoustics
• Medical Device Technology

Background:

• Minimally invasive temperature monitoring is crucial for thermotherapy.
• Current methods often require extra sensors or complex simulations.
• Accurate real-time thermal feedback is needed for effective treatment.

Purpose of the Study:

• To investigate the use of vibroacoustic signals for temperature estimation during tissue-tool interaction.
• To develop a minimally invasive method for temperature monitoring using a modified needle.
• To assess the feasibility of extracting thermal information from audio signals during needle insertion.

Main Methods:

• A MEMS audio sensor was integrated with a conventional aspiration needle.
• Needle insertion experiments were conducted in water-saturated foams at varying temperatures (25-55°C).
• Vibroacoustic signals were analyzed using signal processing and a deep learning algorithm.

Main Results:

• Temperature-dependent features were identified in the acquired audio signals.
• The deep learning model predicted temperature with a root mean squared error of approximately 3°C.
• The study demonstrated a correlation between vibroacoustic signals and temperature.

Conclusions:

• Vibroacoustic signal analysis shows promise for non-invasive temperature monitoring.
• This technique could lead to improved real-time thermal management in medical procedures.
• Further research and refinement are needed for clinical application.