diff --git a/PLAX_Orientation.html b/PLAX_Orientation.html index da644f6..5a9f8a5 100644 --- a/PLAX_Orientation.html +++ b/PLAX_Orientation.html @@ -52,7 +52,7 @@
-

Image Quality Assessment Framework - PLAX Rotation/Orientation

+

PLAX Orientation

@@ -66,80 +66,37 @@

Image Quality Assessment Framework - PLAX Rotation/Orientation

-
-
-
-
-
-

Project Description

-
-

Image rotation can significantly affect the quality and interpretability of PLAX echocardiographic images. This project aims to create a standardized, open-access dataset of PLAX images with varying degrees of rotation, along with expert annotations. -

-
-
+
+
+
+
+
+

We are training an AI to automatically detect and quantify the degree of incorrect rotation in left ventricular (LV) images within the parasternal long axis (PLAX) view. To achieve this, we are ranking images based on the degree of rotation, from the most clockwise to the most anticlockwise. In some cases, the degree of rotation between two images may be minimal, making their exact order less critical. However, it is essential to ensure that images with significant differences in rotation are ranked accurately. Proper ranking will help the AI model better understand and quantify rotation discrepancies, contributing to improved diagnostic precision. +

+
+
+
+
+ +
+ model
- -
-
- -
- model -
Network Architecture
-
- -

Project Goals

-
    -
  • Develop a Reference Dataset: Create a comprehensive dataset of PLAX echocardiographic images with diverse rotation angles, carefully annotated by experienced echocardiographers.
    • -
  • Standardize Rotation Assessment: Establish clear and objective criteria for assessing the degree of rotation in PLAX images.
    • -
  • Empower AI Development: Provide a high-quality dataset to train and validate AI algorithms for automatic rotation detection and correction.
    • -
  • Improve Image Acquisition: Develop training materials and guidelines to help sonographers optimize image acquisition and minimize rotational artifacts.
    • -
  • Enhance Diagnostic Accuracy: Ultimately improve the accuracy and reliability of PLAX echocardiographic interpretations for better patient care.
    • -
- - - - - - - -

Your Role

-
    -
  • Review PLAX Echocardiographic Videos: You will be presented with a series of PLAX echocardiographic videos.
    • -
  • Assess Rotation: Using the provided guidelines and your expertise, you will assess the degree of rotation in each video.
    • -
  • Rank Images: You will then rank the videos based on the severity of rotation, from least to most rotated.
    • -
  • Provide Feedback (Optional): Share your insights and suggestions for improving the annotation process or the assessment criteria.
    • -
- - -

How to Participate

-
    -
  1. Create an Account: Register on the UnityImaging.net labelling platform.
    2. -
  2. Access the Project: Find and join the "PLAX Rotation Assessment Project."
    3. -
  3. Review Guidelines: Familiarize yourself with the detailed annotation instructions and criteria for assessing rotation.
    4. -
  4. Start Annotating: Begin reviewing and ranking PLAX echocardiographic videos based on the degree of rotation.
    5. -
  5. Provide Feedback (Optional): Share your insights to help refine the process and improve the dataset.
    6. -
- -

Ranking Criteria

-
    -
  • Anatomical Landmarks: Assess how well the anatomical landmarks align with standard views.
    • -
  • Overall Image Quality: Rate the overall clarity and diagnostic quality of the image, considering factors like contrast and resolution.
    • -
- -

Who Should Participate?

-
    -
  • Echocardiographers: Your expertise in image interpretation is crucial for accurate assessment.
    • -
  • Cardiologists: Your clinical knowledge can help understand the impact of rotation on diagnosis.
    • -
  • Sonographers: Your experience in image acquisition can provide valuable insights into optimizing image quality.
    • -
- - -
-
- - - - +
+
+
+

This video explains how to rank the images based on the degree of PLAX rotation using unity imaging.net application.

+
+ +
+
+
+
+ + Start Labelling +
+
+
+