Intelligent Intraoperative Haptic-AR Navigation for COVID-19 Lung Biopsy Using Deep Hybrid Model.

Area of Science:

• Medical Imaging and Image-Guided Interventions
• Surgical Robotics and Navigation
• Artificial Intelligence in Medicine

Background:

• Accurate image-guided lung biopsy is crucial for diagnosing COVID-19.
• Existing navigation techniques may lack precision and user immersion.
• Integrating advanced technologies can enhance surgical performance.

Purpose of the Study:

• To develop and evaluate a novel intelligent navigation technique for accurate image-guided COVID-19 lung biopsy.
• To combine augmented reality (AR), haptic feedback, and deep learning (DL) for customized surgical navigation.
• To improve user experience and surgical precision in interventional procedures.

Main Methods:

• A novel system integrating AR, haptic-enabled surgical tools, and DL was developed.
• Clinic data from 341 COVID-19 positive and 1598 control patients were used for model training and evaluation.
• A WPD-CNN-LSTM (WCL) model was employed for patient-specific surgical modeling, and ResNet for intraoperative force classification.
• A 3-D user interface (3DUI) with real-time response was integrated with haptic-AR navigational views.

Main Results:

• The proposed WCL model achieved a root-mean-square error of 0.0128.
• Intraoperative force classification accuracy reached 97%.
• The AR with DL intelligent model demonstrated significantly higher performance compared to existing perception navigation techniques.

Conclusions:

• The novel AR and DL-based intelligent navigation technique significantly enhances accuracy and performance in COVID-19 lung biopsies.
• This framework represents a breakthrough in interventional surgical integration for COVID-19.
• The study opens new research avenues for integrating AR, haptic rendering, and DL in surgical navigation.