PROJECT DESCRIPTION
Identification of unstable carotid plaques associated with symptoms using ultrasonic image analysis and plaque motion analysis
The overall objective of the AtheroRisk project is to develop an integrated intelligent software system for the identification of unstable carotid plaques associated with stroke by combining different clinical and imaging, based features on the analysis of motion and morphology as well as raw pixel intensities of the ultrasound video and plaque images in asymptomatic patients with moderate to severe stenosis.
The ultimate aim is to develop a method that can be used for stroke risk stratification.
Utrasound Video Analysis
Model differentiating between potentially stable and unstable (leading to stroke) plaque motion.
Adaptive multi-scale Amplitude Modulation-Frequency Modulation (AM-FM)
Texture analysis system classifying (predicting) plaques associated with stroke based image characteristics.
Deep Learning Models
DL-based automated carotid plaque segmentation and classification based on CNN transfer-learning models and ML models, leveraging AM-FM and classical plaque-originating imaging features.
Funding Source
AtheroRisk, a two-year project has secured funding by the Cyprus Research and Innovation Foundation (Proposal Num.:EXCELLENCE/0421/0292), receving 255,850.00 €.
Project Coordination: eHealth Lab at CUT
WORK PACKAGES
WP1-Project Management
This WP carry out all project management activities to guarantee the smooth running of the project.
WP2 - Dissemination and Exploitation Activities
The dissemination and outreach activities plan details the array and type of activities undertaken to communicate and engage relevant stakeholders, including citizens, health professionals, patient associations, policy makers, industry with respect to the outcomes of the Atherorisk project.
WP3 - User Requirements and Reference Architecture
This WP includes collection, analysis and evaluation of clinical requirements and use cases of Stroke risk assessment, and the critical technological areas related to ultrasound image and video and image analysis for the prediction of the risk of stroke. Definition of the AtheroRisk Architecture, covering the design and documentation of an interoperable, intelligent tool for the predicition of the risk of stroke.
WP4 - AtheroRisk System Development
This WP involves the development of the AtheroRisk intelligent platform, the integration of the different software components (spaces) and the technical evaluation of the platform.
WP5 - Clinical and System Performance Evaluation
Evaluation Strategy of Platform and Quality control Criteria. First and Early adopters program. Clinical Evaluation.
Deliverables
|
Deliverable
|
WP
|
Dissemination Class
|
|---|---|---|
|
Quality Assurance Plan (M3)
|
1
|
Confidential
|
|
Annual Progress Report (M12)
|
1
|
Public
|
|
2
|
Confidential
|
|
|
Progress Dissemination Report.pdf (M12)
|
2
|
Public
|
|
User needs and Requirements (M6)
|
3
|
Public
|
|
Atherorisk - Technical Specifications (M9)
|
3
|
Confidential
|
|
Clinical Evaluation Guidelines.pdf (M6)
|
5
|
Public
|
|
AtheroRisk Integrated Software (M24)
|
24
|
Confidential
|
|
Clinical Evaluation Results (M24)
|
24
|
Public
|
P a r t n e r s
ATHERORISK PROJECT FINDINGS PRESENTED IN CONFERENCES
BHI2022 and DSP2023
The project description was presented in a Special Session in the BHI2022 Conference in September 2022, in Ioannina city, Greece.
In 2023, we presented our first findings on automatic carotid plaques segmentation in Ultrasound longitudinal grayscale images, based on a convolutional neural network training and evaluation.
