Detection of Aortic Dissection and Intramural Hematoma in Non-Contrast Chest Computed Tomography Using a You Only
Yu-Seop Kim1, Jae Guk Kim2,3, Hyun Young Choi2,3
1Department of Convergence Software, Hallym University, Chuncheon 24252, Republic of Korea.
Insights
A deep learning model accurately differentiates aortic dissection (AD) and aortic intramural hematoma (IMH) from normal aorta (NA) using non-contrast CT scans. This AI tool aids diagnosis when contrast agents are contraindicated.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Radiology
- Cardiovascular Diseases
Background:
- Aortic dissection (AD) and aortic intramural hematoma (IMH) are life-threatening conditions with overlapping clinical presentations.
- Contrast-enhanced computed tomography (CT) is typically essential for diagnosing AD and IMH.
- Developing alternative diagnostic methods is crucial, especially when contrast administration is not feasible.
Purpose of the Study:
- To develop and evaluate a deep learning algorithm for differentiating AD and IMH from normal aorta (NA) using non-contrast CT images.
- To assess the efficacy of the YOLOv4 model in identifying these aortic pathologies without contrast agents.
Main Methods:
- A retrospective analysis of 8881 non-contrast chest CT scans from 121 patients was performed.
- A deep learning model, YOLO (You Only Look Once) v4, was trained and validated on CT images categorized as NA, AD, or IMH.
- The dataset was divided into training, validation, and testing sets in an 8:1:1 ratio.
Main Results:
- The YOLOv4 deep learning model achieved over 92% accuracy in simultaneously distinguishing between AD, IMH, and NA.
- The model demonstrated robust performance in identifying these aortic conditions using non-contrast CT imaging alone.
Conclusions:
- The developed deep learning model offers a promising non-contrast imaging approach for diagnosing AD and IMH.
- This AI-powered tool can assist clinicians in identifying critical aortic pathologies when contrast-enhanced CT is challenging or contraindicated.
Abstract:
Background/Objectives: Aortic dissection (AD) and aortic intramural hematoma (IMH) are fatal diseases with similar clinical characteristics. Immediate computed tomography (CT) with a contrast medium is required to confirm the presence of AD or IMH. This retrospective study aimed to use CT images to differentiate AD and IMH from normal aorta (NA) using a deep learning algorithm. Methods: A 6-year retrospective study of non-contrast chest CT images was conducted at a university hospital in Seoul, Republic of Korea, from January 2016 to July 2021. The position of the aorta was analyzed in each CT image and categorized as NA, AD, or IMH. The images were divided into training, validation, and test sets in an 8:1:1 ratio. A deep learning model that can differentiate between AD and IMH from NA using non-contrast CT images alone, called YOLO (You Only Look Once) v4, was developed. The YOLOv4 model was used to analyze 8881 non-contrast CT images from 121 patients. Results: The YOLOv4 model can distinguish AD, IMH, and NA from each other simultaneously with a probability of over 92% using non-contrast CT images. Conclusions: This model can help distinguish AD and IMH from NA when applying a contrast agent is challenging.
