Abstract:The SYNTAX score has become a widely used measure of coronary disease severity , crucial in selecting the optimal mode of revascularization. This paper introduces a new medical regression and classification problem - automatically estimating SYNTAX score from coronary angiography. Our study presents a comprehensive dataset of 1,844 patients, featuring a balanced distribution of individuals with zero and non-zero scores. This dataset includes a first-of-its-kind, complete coronary angiography samples captured through a multi-view X-ray video, allowing one to observe coronary arteries from multiple perspectives. Furthermore, we present a novel, fully automatic end-to-end method for estimating the SYNTAX. For such a difficult task, we have achieved a solid coefficient of determination R2 of 0.51 in score predictions.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop a fully - automatic method to estimate the SYNTAX score through coronary angiography. Specifically, the paper aims to: 1. **Establish a comprehensive data set**: This data set contains multi - view X - ray videos of 1,844 patients and their corresponding SYNTAX scores. These videos can observe the coronary arteries from multiple angles. 2. **Propose a brand - new end - to - end method**: Utilize 3D ResNet and recurrent neural networks (RNN) or Transformer models to automatically predict the SYNTAX score from coronary angiography. This method has achieved a relatively high coefficient of determination \(R^2 = 0.51\) in predicting scores. 3. **Solve existing problems**: Currently, the calculation of the SYNTAX score depends on semi - automatic tools, which are highly subjective, time - consuming and error - prone. Fully - automatic SYNTAX score prediction can significantly improve the work efficiency of cardiologists and reduce human errors. ### Main contributions of the paper - **Propose a brand - new end - to - end method**: For the automatic estimation of the SYNTAX score, compared with expert evaluation, this method has achieved a coefficient of determination of \(R^2 = 0.51\). - **Introduce a new large - scale coronary angiography data set**: It contains multi - view X - ray videos and their corresponding SYNTAX scores, providing a valuable resource for future research. - **Establish a new benchmark for SYNTAX score estimation**: By independently labeling 60 cases, the score differences between different doctors are estimated (\(R^2\approx0.7\), with a standard deviation of about 8.5), which is helpful for evaluating the performance of the model. ### Method overview The paper uses a method that combines 3D ResNet and RNN or Transformer models. The specific steps include: - **Feature extraction**: Use 3D ResNet to extract features from each view. - **Feature fusion**: Fuse the extracted features through a head neural network (such as LSTM or BERT). - **Classification and regression**: Use a fully - connected layer to perform the final classification and regression tasks, predict the scores of RCA and LCA respectively, and finally add the two to get the total SYNTAX score. ### Results and discussion - **Prediction performance**: This method has achieved \(R^2 = 0.51\) in SYNTAX score prediction, showing good potential, but still needs further improvement to surpass the expert level. - **Classification performance**: In distinguishing patients with zero scores from those with non - zero scores, the accuracy rate reaches 77%, and the recall rate and F1 - score also perform well. - **Challenges and future work**: Further research is needed on how to mitigate the impact of the SYNTAX score distribution in the test set on the results, identify complex cases, and explore how to improve these situations. In general, this paper proposes a promising fully - automatic SYNTAX score prediction method, providing new ideas and technical support for the diagnosis and treatment of coronary artery diseases.

End-to-end SYNTAX score prediction: benchmark and methods

Fallacies and Possible Remedies of the SYNTAX Score.

Quantitative Prediction of SYNTAX Score for Cardiovascular Artery Disease Patients via the Inverse Problem Algorithm Technique as Artificial Intelligence Assessment in Diagnostics

Editorial comment: SYNTAX score II 2020 as a tool for decision making on revascularization strategy in high‐risk patients with complex coronary artery disease

The Predictive Value of the Syntax Score in Patients With Chronic Coronary Artery Disease Undergoing Percutaneous Coronary Intervention or Coronary Artery Bypass Grafting: A Pilot Study

Angiographic score assessment improves cardiovascular risk prediction: the clinical value of SYNTAX and Gensini application

Predicting the Outcome of Patients with Aneurysmal Subarachnoid Hemorrhage: A Machine-Learning-Guided Scorecard

The impact of SYNTAX score on culprit lesion morphology of acute coronary syndrome

Data governance and Gensini score automatic calculation for coronary angiography with deep-learning-based natural language extraction

Mis-estimation of Coronary Lesions and Rectification by SYNTAX Score Feedback for Coronary Revascularization Appropriateness

A New Approach for Interpretability and Reliability in Clinical Risk Prediction: Acute Coronary Syndrome Scenario

External validation of the SYNTAX score II 2020 in patients with chronic renal insufficiency

Impact of the SYNTAX scores I and II in patients with diabetes and multivessel coronary disease: a pooled analysis of patient level data from the SYNTAX, PRECOMBAT, and BEST trials

10 Years of SYNTAX: Closing an Era of Clinical Research After Identifying New Outcome Determinants

Learning Optimized Risk Scores

Using Text Content From Coronary Catheterization Reports to Predict 5-Year Mortality Among Patients Undergoing Coronary Angiography: A Deep Learning Approach

EventScore: An Automated Real-time Early Warning Score for Clinical Events

AutoScore-Survival: Developing interpretable machine learning-based time-to-event scores with right-censored survival data

Automated stenosis estimation of coronary angiographies using end-to-end learning

Validation of a Novel Clinical Prediction Score for Severe Coronary Artery Diseases Before Elective Coronary Angiography.

Designing optimal mortality risk prediction scores that preserve clinical knowledge