Bronchus Line Structure Estimation in Medical Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical image processing techniques for 3D trachea and bronchus images struggle to precisely extract and reproduce line structures, especially during respiratory cycles, due to lung movement, leading to errors in data processing and the inability to accurately detect tiny tube portions and capillary vessels.
Innovation Solution
A medical image processing apparatus comprising obtaining, calculating, and transforming circuitry that selects a benchmark time phase (expiration or inspiration) to obtain a benchmark line structure, calculates motion amounts between time phases, and transforms the line structure to estimate its position at other time phases, effectively removing lung movement-induced errors and enhancing precision in detecting airway changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If images are taken at all time phases during respiratory cycle, then complete airway morphology can be captured, but processing time and data volume increase significantly
Solution Approach 1:
The patent applies preliminary action by pre-selecting a benchmark time phase (expiration or inspiration) where the airway structure is most stable and clearly defined. By establishing the line structure at this reference phase beforehand, the system avoids the need to process and analyze images from all time phases, thereby reducing processing time while maintaining detection precision through subsequent transformation of the benchmark line structure to other time phases.
2Productivity
If central line extraction is performed at end-systolic period, then processing efficiency is improved, but detection precision of tiny vessels decreases
Solution Approach 1:
The patent uses the benchmark line structure at the selected time phase as an intermediary representation. Instead of directly extracting and analyzing central lines at end-systolic period where tiny vessels are difficult to detect, the system transforms the benchmark line structure to other time phases using calculated motion amounts. This intermediary approach allows efficient processing while maintaining the ability to detect tiny vessels through the transformed line structure rather than direct extraction from difficult phases.
3Ease of manufacture
If direct airway extraction is performed at individual time phases, then processing is simpler, but lung movement errors are introduced
Solution Approach 1:
The patent applies copying by creating a transformed line structure that replicates the airway morphology at different time phases based on the benchmark line structure. Instead of directly extracting airways at each time phase (which introduces lung movement errors), the system copies the benchmark line structure and transforms it using calculated motion amounts, thereby maintaining processing simplicity while improving reliability by removing lung movement-induced position errors.
4Productivity
If benchmark time phase is selected from end-systolic period, then processing speed increases, but capillary vessel extraction completeness decreases
Solution Approach 1:
The patent applies preliminary action by pre-selecting an appropriate benchmark time phase that optimizes capillary vessel visibility, rather than defaulting to end-systolic period. By establishing the benchmark line structure at a phase where capillary vessels are clearly defined, and then transforming this structure to other time phases, the system maintains processing speed while preventing loss of capillary vessel information that would occur if end-systolic period were used as the benchmark.
Data Source
AI summary
A medical image processing apparatus according to one embodiment includes obtaining circuitry, calculating circuitry, and transforming circuitry. The obtaining circuitry takes an expiration time phase or an inspiration time phase as a benchmark time phase and obtains a benchmark line structure, which is a line structure of a bronchus in a lung field, from a medical image at the benchmark time phase. The calculating circuitry calculates a motion amount between a component depicted in the medical image at the benchmark time phase and a component depicted in a medical image at at least one other time phase than the benchmark time phase. The transforming circuitry transforms the benchmark line structure based on the motion amount to obtain an estimated line structure, which is estimated as a line structure at the at least one other time phase.