In situ tissue classification during laser ablation using acoustic signals.

Area of Science:

• Biomedical Engineering
• Acoustic Signal Processing
• Medical Device Technology

Background:

• Accurate tissue identification is crucial for effective and safe vascular interventions.
• Pulsed ultraviolet laser ablation generates acoustic signals that vary with tissue type.
• Distinguishing between different tissues in real-time is a significant clinical challenge.

Purpose of the Study:

• To develop and validate a novel method for classifying ablated tissue types using acoustic signatures.
• To assess the feasibility of this method in complex and noisy environments mimicking clinical settings.
• To improve the safety and efficacy of laser-based medical procedures through real-time tissue identification.

Main Methods:

• Utilized acoustic sound waves recorded during pulsed ultraviolet laser ablation for tissue classification.
• Employed Mel-frequency cepstral coefficients (MFCCs) for feature extraction and Support Vector Machine (SVM) and fully connected deep neural network (FC-DNN) algorithms for classification.
• Validated the method through experiments classifying liquids, ex vivo porcine aorta, and bovine tendon tissues in saline and through chicken breast medium.

Main Results:

• Achieved high classification accuracy (>98%) for ex vivo tissues in both saline and through chicken breast environments.
• Demonstrated successful preliminary classification of three different liquids.
• Validated the method's robustness in noisy conditions simulating practical working environments.

Conclusions:

• The proposed acoustic signature classification method demonstrates high accuracy and robustness for tissue identification during laser ablation.
• This technique holds significant potential for real-time tissue classification in various medical procedures, improving outcomes.
• The findings pave the way for enhanced safety and efficacy in laser-assisted medical interventions.