Surgical Video Obstruction Removal via Historical Frame Replacement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
During surgical procedures, especially endoscopic interventions, the visualization of tissue surfaces is obstructed by interfering objects such as instruments, blood pools, and other foreground objects, making it difficult for practitioners to assess the treated area accurately.
Innovation Solution
A device and method that utilize an identification block to distinguish tissue surface areas from obscured areas using object recognition algorithms or neural networks. Foreground objects are identified and replaced with sections from older images in the image sequence, ensuring real-time reconstruction of hidden tissue areas without interpolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If image processing algorithms are used to remove interfering objects from the field of view, then visibility of the tissue surface is improved, but the complexity of the device increases
Solution Approach 1:
The system performs preliminary identification of blind areas caused by interfering objects using object recognition algorithms or neural networks. By pre-processing the image sequence to detect and mark obscured regions before final display, the system improves tissue visibility while managing computational complexity through staged processing.
Solution Approach 2:
The system creates replacement images by copying visible tissue areas from previous time points to fill in blind areas where tissue is obscured. This copying approach reconstructs hidden tissue regions using historical data without requiring complex real-time interpolation, thereby improving visibility while controlling device complexity.
2Measurement precision
If real-time image processing is performed to remove foreground objects, then surgical navigation accuracy is improved, but processing time increases
Solution Approach 1:
The system performs preliminary identification of blind areas and selects appropriate replacement regions from previous frames before final image composition. This pre-processing step enables real-time performance by avoiding complex computations during the critical display phase, thus maintaining both surgical navigation accuracy and real-time processing requirements.
Solution Approach 2:
The system discards current frame data in blind areas and recovers tissue information from previous time points when tissue position and shape are similar. This approach maintains navigation accuracy by using reliable historical data while minimizing processing time by avoiding complex real-time reconstruction in obscured regions.
3Stability of the object's composition
If interpolation methods are used to fill blind areas, then continuity of tissue surface is improved, but risk of displaying non-existent tissue structures increases
Solution Approach 1:
Instead of interpolating tissue data, the system copies actual tissue images from previous time points to fill blind areas. This copying method maintains tissue surface continuity while ensuring reliability by displaying real, observed tissue structures rather than synthesized or interpolated data that may not exist.
Solution Approach 2:
The system preliminarily identifies suitable replacement regions from previous frames that match the spatial and temporal characteristics of blind areas. By pre-selecting appropriate historical tissue images for replacement, the system ensures both continuity of the tissue surface and accuracy of representation without relying on interpolation methods.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
A method and device according to the invention for observing a tissue surface of a surgical site provide for interfering objects within the surgeon's field of vision, such as instruments, blood pools, or the like, to be removed from the main image and rendered completely invisible or replaced by optical, non-interfering graphic representations. To this end, the interfering objects in the foreground are identified, and the tissue surface obscured by them is replaced using sections from older images of the image sequence or video stream. The method is fast, low computational intensity, and allows the surgeon improved control of the surgical procedure.