AI-Enhanced Vibrational Capsule for Minimally Invasive Detection of Abnormal Bowel Tissue.

Area of Science:

• Biomedical Engineering
• Artificial Intelligence
• Gastroenterology

Background:

• Colorectal cancer (CRC) is a leading cause of cancer mortality, with early detection crucial for prognosis.
• Conventional diagnostic methods like colonoscopy and capsule endoscopy rely on visual inspection and miss subsurface abnormalities.
• There is a need for advanced diagnostic tools capable of detecting biomechanical changes indicative of early-stage CRC.

Purpose of the Study:

• To develop and evaluate an AI-enhanced, vibration-assisted capsule for in situ biomechanical sensing of bowel tissues.
• To assess the capsule's ability to differentiate between normal and abnormal colonic tissues based on mechanical properties.
• To establish a minimally invasive, AI-driven modality for non-visual CRC detection.

Main Methods:

• A swallowable capsule integrating a vibration motor, accelerometer, and wireless communication was designed.
• The capsule induces tissue oscillations, and acceleration responses are analyzed in real-time.
• A one-class support vector machine (SVM) was trained on normal tissue data to classify abnormal tissues.
• Dynamic modeling was used to correlate lesion stiffness with classification outcomes.

Main Results:

• Vibration-induced signals captured reproducible mechanical contrasts between normal and abnormal colon tissues.
• The AI-based classification accurately identified abnormal tissues based on stiffness variations.
• Experiments on ex vivo and in vitro models validated the capsule's sensing capabilities.

Conclusions:

• The AI-enhanced vibration-assisted capsule offers a novel, minimally invasive approach for non-visual CRC detection.
• This technology has the potential to complement existing imaging diagnostics and improve early cancer screening.
• The system provides real-time, patient-friendly screening by assessing tissue biomechanics.