Abstract:Cardiovascular Diseases (CVDs) are the leading cause of death worldwide, taking 17.9 million lives annually. Abdominal Aortic Calcification (AAC) is an established marker for CVD, which can be observed in lateral view Vertebral Fracture Assessment (VFA) scans, usually done for vertebral fracture detection. Early detection of AAC may help reduce the risk of developing clinical CVDs by encouraging preventive measures. Manual analysis of VFA scans for AAC measurement is time consuming and requires trained human assessors. Recently, efforts have been made to automate the process, however, the proposed models are either low in accuracy, lack granular level score prediction, or are too heavy in terms of inference time and memory footprint. Considering all these shortcomings of existing algorithms, we propose 'AACLiteNet', a lightweight deep learning model that predicts both cumulative and granular level AAC scores with high accuracy, and also has a low memory footprint, and computation cost (Floating Point Operations (FLOPs)). The AACLiteNet achieves a significantly improved one-vs-rest average accuracy of 85.94% as compared to the previous best 81.98%, with 19.88 times less computational cost and 2.26 times less memory footprint, making it implementable on portable computing devices.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop a lightweight deep - learning model for automatically detecting and scoring fine - grained abdominal aortic calcification (AAC) from vertebral fracture assessment (VFA) scan images. Specifically, the existing methods have the following problems: 1. **Insufficient accuracy**: Some existing models perform poorly in predicting fine - grained AAC scores. 2. **Lack of fine - grained score prediction**: Existing models usually only predict the cumulative score and ignore the fine - grained score of each vertebral body. 3. **High computational cost**: Many models are very expensive in terms of inference time and memory usage, which limits their application on mobile devices. To solve these problems, the authors propose a lightweight deep - learning model named "AACLiteNet". This model aims to achieve high - accuracy AAC score prediction, including cumulative scores and fine - grained scores, with lower computational cost and less memory usage. ### Main features of the model - **Lightweight architecture**: By using techniques such as depthwise separable convolution, the number of model parameters and computational complexity are reduced. - **Efficient attention mechanism**: An efficient global attention mechanism, similar to the self - attention mechanism in the Transformer architecture, is introduced to enhance the feature representation ability. - **Multi - task learning**: The problem is framed as a combination of regression and multi - class multi - label classification tasks, thereby improving the overall performance of the model. ### Experimental results The experimental results show that AACLiteNet significantly outperforms existing methods in multiple metrics: - **Average accuracy**: It reaches 85.94%, which is a significant improvement compared to the previous best of 81.98%. - **Computational cost**: The computational cost is reduced by 19.88 times, and the memory usage is reduced by 2.26 times. - **Clinical relevance**: The AAC scores predicted by the model are highly correlated with the scores labeled by humans (Pearson correlation coefficient is 0.89), and the risk ratio in predicting major acute cardiovascular events (MACE) is consistent with the results of human assessment. In conclusion, AACLiteNet provides an efficient and accurate solution that can realize the automatic detection and scoring of abdominal aortic calcification in resource - constrained environments.

AACLiteNet: A Lightweight Model for Detection of Fine-Grained Abdominal Aortic Calcification

Machine learning for abdominal aortic calcification assessment from bone density machine-derived lateral spine images

Automatic Segmentation and Quantification of Abdominal Aortic Calcification in Lateral Lumbar Radiographs Based on Deep-Learning-Based Algorithms.

SCOL: Supervised Contrastive Ordinal Loss for Abdominal Aortic Calcification Scoring on Vertebral Fracture Assessment Scans

Automated Measurement of Vascular Calcification in Femoral Endarterectomy Patients Using Deep Learning

DLFFNet: A new dynamical local feature fusion network for automatic aortic valve calcification recognition using echocardiography

End-to-End, Pixel-Wise Vessel-Specific Coronary and Aortic Calcium Detection and Scoring Using Deep Learning

DeepAAA: clinically applicable and generalizable detection of abdominal aortic aneurysm using deep learning

Quantifying the calcification of abdominal aorta and major side branches with deep learning

Independent Associations of Aortic Calcification with Cirrhosis and Liver Related Mortality in Veterans with Chronic Liver Disease

Automated Screening for Abdominal Aortic Aneurysm in CT Scans under Clinical Conditions Using Deep Learning

Resource efficient aortic distensibility calculation by end to end spatiotemporal learning of aortic lumen from multicentre multivendor multidisease CMR images

Automated Analysis of Femoral Artery Calcification Using Machine Learning Techniques

Abdominal Aortic Calcification and Cognitive Function among Older Adults: Cross-sectional Analysis of National Health and Nutrition Examination Survey, 2013-2014.

Artificial Intelligence in Coronary Artery Calcium Scoring Detection and Quantification

Multiscale dilated convolutional neural network for Atrial Fibrillation detection

Diagnosis of atrial fibrillation based on lightweight detail-semantic network

Automated Coronary Artery Calcium Scoring Using Convolutional Neural Networks: Enhancing Cardiovascular Risk Assessment in Chest CT Scans

Artificial Intelligence-Based Opportunistic Coronary Calcium Screening in the Veterans Affairs National Healthcare System

Automated vessel specific coronary artery calcification quantification with deep learning in a large multi-center registry

Prognostic Value of Abdominal Aortic Calcification: A Systematic Review and Meta-Analysis of Observational Studies