Medical Navigation With 3D Model Alignment for Laparoscopy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Laparoscopic surgeries lack haptic feedback and provide limited visualization of sub-surface structures, relying on 2D ultrasound images that require mental integration with 3D space, complicating intraoperative navigation.
Innovation Solution
A method using three artificial neural networks to align a virtual 3D target object model with a laparoscopic video image by detecting anatomical landmarks, allowing real-time repositioning and alignment for enhanced spatial navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If 2D ultrasound images are used to visualize sub-surface structures, then visualization capability is provided, but spatial navigation becomes complicated due to mental integration requirements
Solution Approach 1:
The patent transforms 2D ultrasound images into a 3D virtual model of the target organ, adding the third dimension to enable direct spatial navigation. The 3D model allows surgeons to visualize and navigate through anatomical structures in their actual three-dimensional space, eliminating the need to mentally integrate 2D slices.
Solution Approach 2:
The patent creates a virtual copy (3D model) of the target organ based on preoperative imaging data. This virtual model serves as an interactive replica that can be rotated, zoomed, and navigated within the surgical environment, providing intuitive spatial relationships without requiring mental reconstruction from 2D images.
2Ease of operation
If preoperative image data is transformed into a 3D model and aligned with intraoperative images, then spatial navigation is improved, but image registration complexity increases
Solution Approach 1:
The system automatically performs image registration by detecting anatomical landmarks in both preoperative and intraoperative images, then using these landmarks to align the 3D virtual model with the current surgical field. This self-aligning mechanism reduces manual intervention and simplifies the registration process.
Solution Approach 2:
The system continuously tracks anatomical landmarks during the surgical procedure and provides real-time feedback to maintain alignment between the 3D virtual model and the actual target organ. This feedback loop ensures the model remains accurately registered even as the organ moves or deforms during surgery.
3Ease of manufacture
If laparoscopic surgery is performed, then smaller incisions and shorter recovery time are achieved, but haptic feedback is lost
Solution Approach 1:
The patent introduces a virtual 3D model as an intermediary between the surgeon and the target organ. This virtual model compensates for the lost haptic feedback by providing visual information about the organ's internal structure, texture, and spatial relationships, allowing the surgeon to 'feel' the organ through visual inspection rather than physical contact.
Solution Approach 2:
The system replaces the mechanical sense of touch with visual information processing. Instead of relying on haptic feedback from direct contact, the surgeon uses the 3D visual model to understand the organ's anatomy and navigate surgical structures, substituting tactile perception with visual cognition.
Data Source
AI summary
A method for operating a medical navigation system for an image-guided surgical procedure including: training a first artificial neural network on a set of surgical image video data of a target object of a patient representing a video image, and identifying the structure of the target object; training a second artificial neural network on at least one virtual three-dimensional target object model of the target object and identifying the structure of the virtual three-dimensional target object model by the identified structure of the target object; training a third artificial neural network by the identified structure of the target object and the structure of the virtual three-dimensional target object model and aligning the identified structure of the target object, with the structure of the corresponding target object of the virtual three-dimensional target object model, which is identified by the second artificial neural network.