Search

EP-4343686-B1 - METHOD AND SYSTEM FOR VASCULAR CATHETER TIP DETECTION IN MEDICAL IMAGES

EP4343686B1EP 4343686 B1EP4343686 B1EP 4343686B1EP-4343686-B1

Inventors

  • LIAO, Rui
  • NARASIMHA MURTHY, VENKATESH
  • ZHANG, YUE
  • AMADOU, ABDOUL AZIZ

Dates

Publication Date
20260506
Application Date
20220923

Claims (15)

  1. Computer-implemented method (100) for detecting a tip (210T) of a vascular catheter (210) in a sequence (200) of vessel image frames (201) obtained using x-rays, comprising the steps of: determining (110), within the vessel image frames, a proximal point (210P), the proximal point corresponding to a vessel ostium; cropping (120) the sequence (200) of vessel image frames (201) to an image area surrounding the proximal point to generate a cropped sequence (200C) of cropped vessel image frames (201C); detecting (130), within the cropped vessel image frames (201C) not indicative of a contrast medium, the tip (210T) of the vascular catheter (210); and tracking (140), within the sequence (200) of vessel image frames (201), the detected tip (210T) of the vascular catheter (210).
  2. The method of claim 1, wherein the determining (110) the proximal point (210P) includes detecting (111) the proximal point (210P) within vessel image frames (201) indicative of the contrast medium.
  3. The method of claim 2, wherein the detecting (111) the proximal point (210P) includes detecting (111a) the contrast medium within the vessel image frames (201) to identify the vessel image frames indicative of the contrast medium.
  4. The method of any one of claims 2 and 3, wherein the detecting the proximal point (210P) includes (112) detecting a pattern of the contrast medium exiting the tip (210T) of the vascular catheter (210).
  5. The method of any one of claims 2 to 4, wherein the detecting (111) the proximal point (210P) is performed using K-means clustering.
  6. The method of claim 1, wherein the determining (110) the proximal point (210P) is based on receiving information identifying the proximal point (210P) in terms of a pixel coordinate system of the vessel image frames.
  7. The method of any one of the preceding claims, wherein the detecting (130) the tip (210T) of the vascular catheter (210) within the cropped vessel image frames (201C) not indicative of the contrast medium is based on classifying, a position within each cropped vessel image frame (201C) as the tip (210T).
  8. The method of any one of the preceding claims, wherein: the detecting (130) the tip (210T) of the vascular catheter (210) outputs a detected position of the tip (210T) of the vascular catheter (210) per cropped vessel image frame (201C) not indicative of a contrast medium; and the detecting (130) the tip (210T) of the vascular catheter (210) includes Kalman filtering (131) the detected positions to generate a filtered detected position.
  9. The method of any one of the preceding claims, wherein the tracking (140) the detected tip (210T) of the vascular catheter (210) is based on a correlation between temporally adjacent vessel image frames (201).
  10. The method of any one of the preceding claims, wherein: the tracking (140) of the detected catheter tip (210T) outputs a tracked position of the tip (210T) of the vascular catheter (210) per vessel image frame (201); and the tracking (140) the detected tip (210T) of the vascular catheter (210) includes Kalman filtering the tracked positions to generate a filtered tracked position.
  11. The method of any one of the preceding claims, wherein the detecting (130) the tip (210T) of the vascular catheter (210) and the tracking (140) the detected tip (210T) of the vascular catheter (210) within the sequence (200) of vessel image frames (201) are performed on a same frame number of the sequence (200) of vessel image frames (201) and the cropped sequence (200C) of cropped vessel image frames (201C).
  12. The method of claim 11, further comprising: comparing (150) a detected position of the tip (210T) of the vascular catheter (210), the detected position being output by the detecting the tip (210T) of the vascular catheter (210), with a tracked position of the tip (210T) of the vascular catheter (210), the tracked position being output by the tracking (140) of the detected tip (210T) of the vascular catheter (210); if the detected position and the tracked position differ by less than a position difference threshold, treating (151) the detected position as a valid position of the tip (210T) of the vascular catheter (210); and if the detected position and the tracked position differ by more than the position difference threshold, treating (152) the tracked position as the valid position of the tip (210T) of the vascular catheter (210).
  13. The method of claim 12, wherein the comparing (150) the detected position of the tip (210T) with the tracked position of the tip (210T) further comprises: if the detected position and the tracked position differ by more than the position difference threshold, resetting (153) the detecting (130) of the of the tip (210T) of the vascular catheter (210) based on the tracked position.
  14. A computer-readable medium (1350, 1360) comprising instructions configured to be executed by a computer (1300) including at least one processor (1310), the instructions causing the processor (1310) to perform the method (100) according to any one of claim 1 to 13.
  15. A medical imaging device (1200) including x-ray imaging means (1231, 1232) configured to record a sequence (200) of vessel image frames (201) and at least one processor (1310), the processor (1310) being configured to perform the method (100) according to any one of claim 1 to 13.

Description

TECHNICAL FIELD The invention generally relates to object detection in medical images and more precisely to detecting tips of vascular catheters in angiography and fluoroscopy images. BACKGROUND Interpretation and processing of image sequences obtained during angiography and fluoroscopy typically relies on the detection of anatomical and non-anatomical landmarks. For example, US 11,058,385 B2 discloses techniques for evaluating cardiac motion and estimating cardiac cycle using angiography image data to improve detection of a radiopaque marker in an angiography image. One type of non-anatomical landmark relied upon for interpretation and processing of these image sequences is a vascular catheter and more precisely the tip thereof. For example, US 2021 / 0 346 100 A1 discloses guidance of an interventional device in a vascular structure based on a two-dimensional image of a region of interest in which the interventional device is at least partly visible. A location of a specific part of the interventional device is detected. However, the detection of the tip of a vascular catheter is rendered difficult by various factors, including, but not limited to, occlusion by the contrast medium ejected from the vascular catheter, low-signal-to-noise (SNR) in the image sequence, interference from radio-opaque anatomical landmarks and interference from non-anatomical landmarks within the image sequence, such as parts of the medical imaging device used to obtain the image sequence or other radio-opaque non-anatomical landmarks, such as other catheters, stents or pacemakers. Therefore, it is an objective of the present invention to provide a detection of the tip of vascular catheters which overcomes the above-mentioned challenges. SUMMARY OF THE INVENTION To achieve this objective, the present invention provides a computer-implemented method for detecting a tip of a vascular catheter in a sequence of vessel image frames obtained using x-rays. The method comprises the steps of determining, within the vessel image frames, a proximal point, the proximal point corresponding to a vessel ostium, cropping the sequence of vessel image frames to an image area surrounding the proximal point to generate a cropped sequence of cropped vessel image frames, detecting, within cropped vessel image frames not indicative of a contrast medium, the tip of the vascular catheter, and tracking, within the sequence of vessel image frames, the detected tip of the vascular catheter. To further achieve this objective, the present invention provides a computer-readable medium comprising instructions configured to be executed by a computer including at least one processor, the instructions causing the processor to perform the above method. To further achieve this objective, the present invention provides a medical imaging device including x-ray imaging means configured to record a sequence of vessel image frames and at least one processor, the processor being configured to perform the method according to the above method. BRIEF DESCRIPTION OF THE DRAWINGS Embodiments of the present invention will be described with reference to the following appended drawings, in which like reference signs refer to like elements. Fig. 1 provides a flowchart of a method for detecting a tip of a vascular catheter in a sequence of vessel image frames according to embodiments of the present invention.Fig. 2 provides an example of determining a proximal point in a vessel image frame according to embodiments of the present invention.Fig. 3 provides an example of detecting a tip of a vascular catheter in a cropped vessel image frame according to embodiments of the present invention.Fig. 4 provides an example of tracking the detected tip of the vascular catheter within the sequence of vessel image frames according to embodiments of the present invention.Fig. 5 provides a diagram comparing a detected position, a tracked position and a filtered position of the tip of the vascular catheter according to embodiments of the present invention.Figs. 6, 6A and 6B show a schematic diagram of a pathological vessel guidance workflow according to embodiments of the invention.Fig. 7 shows examples of segmentations which may be obtained from a vessel image sequence according to embodiments of the invention.Fig. 8 shows an example of a vessel roadmap according to embodiments of the invention.Fig. 9 shows an example of a real-time fluoroscopy image and information associated therewith according to embodiments of the invention.Fig. 10 shows examples of laying a vessel roadmap over a real-time fluoroscopy according to embodiments of the invention.Fig. 11 shows an example of overlaying a vessel roadmap and a real-time fluoroscopy image according to embodiments of the invention.Figs. 12A and 12B show an exemplary medical imaging system according to embodiments of the invention.Fig. 13 shows an exemplary computing device according to embodiments of the present invention. It should be understood that