Machine-Learned Model for Intra-Operative Lung Mass Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Minimally invasive surgery, such as laparoscopic surgery, faces challenges in detecting lung masses and nodules intra-operatively due to lung deflation and deformation, making it difficult for surgeons to accurately locate deep lung objects during procedures.

Innovation Solution

A machine-learned model, such as a convolutional neural network or U-Net, is used to predict the intra-operative location of lung objects by inputting pre-operative images and real-time images, along with clinical data, to guide surgical systems like robotic arms or bronchoscopes to the correct location during surgery.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If pre-operative images are used for surgical guidance, then surgical planning is improved, but the accuracy of object location deteriorates due to lung deflation and deformation during surgery

Engineering Contradiction:

Improvesurgical planningVSAvoidobject location accuracy

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The system dynamically adapts the pre-operative images by applying a deformation model that accounts for lung deflation and deformation during surgery. The machine-learned model transforms the static pre-operative images into dynamic representations that reflect the actual intra-operative lung state, thereby maintaining location accuracy despite anatomical changes

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the parameters of the pre-operative images by applying a deformation model that modifies the spatial coordinates and anatomical structures based on predicted lung deflation. This parameter transformation aligns the pre-operative image data with the actual intra-operative anatomy, resolving the location accuracy issue

Inventive Principle:

Principle #35Parameter changes

2Ease of operation

If lung deflation is performed during minimally invasive surgery, then surgical access is improved, but detection of lung masses deteriorates due to tissue deformation

Engineering Contradiction:

Improvesurgical accessVSAvoidlung mass detection

Core Design Contradiction:

Ease of operationVSDifficulty of detecting and measuring

Solution Approach 1:

The system introduces an intermediary deformation model that mediates between the deflated lung state (necessary for surgical access) and the pre-operative images (needed for mass detection). This model acts as a bridge, transforming the pre-operative images to match the deflated lung anatomy, thereby enabling mass detection despite the necessary lung deflation

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system replaces the mechanical visual detection method (direct visualization of lung masses) with a computational image transformation approach. Instead of relying on the surgeon's ability to visually detect masses in the deflated lung, the system uses a machine-learned deformation model to predict and display mass locations, substituting mechanical detection with computational analysis

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20240164856A1Detection in a surgical system

Publication Date: 2024.05.23 AURIS HEALTH INC

AI summary

For intraoperative guidance to an object, a machine-learned model is used to predict the intra-operative location of an object identified in pre-operative planning. The predicted location is used by the surgeon or controller during the operation, such as during a bronchoscopy.