What is new in computer vision and artificial intelligence in medical image analysis applications

Olveres, Jimena; González, Germán; Torres, Fabián; Moreno-Tagle, José Carlos; Carbajal-Degante, Erik; Valencia-Rodríguez, Alejandro; Méndez-Sánchez, Nahúm; Escalante-Ramı́rez, Boris

doi:10.21037/qims-20-1151

Cited by 50 publications

(24 citation statements)

References 81 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The overall accuracy of these scores has been found to be modest. Artificial intelligence (AI) is currently being integrated with conventional diagnostic methods in the hopes of performance improvement [14]. In a recent metaanalysis of 13 studies, it was shown that AI significantly improves the diagnosis of NAFLD, NASH, and liver fibrosis.…”

Section: Introductionmentioning

confidence: 99%

Detangling the interrelations between MAFLD, insulin resistance, and key hormones

2022

View full text Add to dashboard Cite

Metabolic dysfunction–associated fatty liver disease (MAFLD) has increasingly become a significant and highly prevalent cause of chronic liver disease, displaying a wide array of risk factors and pathophysiologic mechanisms of which only a few have so far been clearly elucidated. A bidirectional interaction between hormonal discrepancies and metabolic-related disorders, including obesity, type 2 diabetes mellitus (T2DM), and polycystic ovarian syndrome (PCOS) has been described. Since the change in nomenclature from non-alcoholic fatty liver disease (NAFLD) to MAFLD is based on the clear impact of metabolic elements on the disease, the reciprocal interactions of hormones such as insulin, adipokines (leptin and adiponectin), and estrogens have strongly pointed to the intrinsic links that lead to the heterogeneous epidemiology, clinical presentations, and risk factors involved in MAFLD in different populations. The objective of this work is twofold. Firstly, there is a brief discussion regarding the change in nomenclature as well as epidemiology, risk factors, and pathophysiologic mechanisms other than hormonal effects, which include nutrition and the gut microbiome, as well as genetic and epigenetic influences. Secondly, we review the basis of the most important hormonal factors involved in the development and progression of MAFLD that act both independently and in an interrelated manner.

show abstract

Section: Introductionmentioning

confidence: 99%

Detangling the interrelations between MAFLD, insulin resistance, and key hormones

2022

View full text Add to dashboard Cite

show abstract

“…Automatic CAC quantification on non-gated chest CT is particularly cumbersome because of high noise, low resolution, and motion artifacts (13,28). The solutions to tackle this challenging task should be focused on the following: (I) accuracy, i.e., maintaining good consistency with the gold standard; (II) reliability, i.e., maintaining good consistency with manual results on chest CT; (III) easy interpretation, i.e., pixels that contribute to the calcium score are well classified in a single forward pass.…”

Section: Discussionmentioning

confidence: 99%

Automatic coronary artery calcium scoring on routine chest computed tomography (CT): comparison of a deep learning algorithm and a dedicated calcium scoring CT

Xu¹,

Guo²,

Xu³

et al. 2022

Quant Imaging Med Surg

View full text Add to dashboard Cite

Background: The aim of this study was to investigate the reliability and accuracy of automatic coronary artery calcium (CAC) scoring and risk classification in non-gated, non-contrast chest computed tomography (CT) of different slice thicknesses using a deep learning algorithm.Methods: This retrospective study was performed at 2 tertiary hospitals. Paired, dedicated calcium-scoring CT scans and non-gated, non-contrast chest CT scans taken within a month from the same patients were included. Chest CT images were grouped according to the slice thickness (group A: 1 mm; group B: 3 mm).For internal scans, the CAC score manually measured on dedicated calcium scoring CT images was used as the gold standard. The deep learning algorithm for group A was trained using 150 chest CT scans and tested using 144 scans, and that for group B was trained using 170 chest CT scans and tested using 144 scans. The intraclass correlation coefficient (ICC) was used to evaluate the correlation between the algorithm and the gold standard. Agreement between the deep learning algorithm, the manual results on chest CT, and the gold standard was determined by Bland-Altman analysis. Cardiac risk categories were compared. External validation was performed on 334 paired scans from a different organization.Results: A total of 608 internal paired scans (1 mm: 294; 3 mm: 314) of 406 individuals and 334 external paired scans (1 mm: 117; 3 mm: 117) of 117 individuals were included in the analysis. The ICCs between the deep learning algorithm and the gold standard were excellent in both group A (0.90; 95% CI: 0.85-0.93) and group B (0.94; 95% CI: 0.92-0.96). The Bland-Altman plots showed good agreement in both groups. For the cardiovascular risk category, the deep learning algorithm accurately classified 71% of cases in group A and 81% of cases in group B. The Kappa values for risk classification were 0.72 in group A and 0.82 in group B.External validation yielded equally good results. Conclusions:The automatic calculation of CAC score and cardiovascular risk stratification on non-gated chest CT using a deep learning algorithm was reliable and accurate on both 1 and 3 mm scans. Chest CT with a slice thickness of 3 mm was slightly more accurate in CAC detection and risk classification.

show abstract

“…Moreover, considering that the spatial attention module only focuses on the target region and the channel attention mechanism only focuses on the local information of each channel, researchers have proposed a mechanism based on the mixed domain, such as convolutional block attention module (CBAM) (14) and residual attention learning (15). However, although CNN-based deep learning methods have made significant progress in the field of medical image segmentation (16,17), a bottleneck remains.…”

Section: Introductionmentioning

confidence: 99%

“…However, most transformer-based methods require pretraining on large datasets to achieve satisfactory performance, yet large datasets and labels with high quality images are very expensive to obtain in the field of medical image segmentation. Some researchers have made related attempts to improve the adaptability of transformer-based methods to small datasets (16,18,27). The most typical work is medical transformer (MT) which was proposed by Valanarasu et al (18).…”

Section: Introductionmentioning

confidence: 99%

Dilated transformer: residual axial attention for breast ultrasound image segmentation

Shen¹,

Wang²,

Yu³

et al. 2022

Quant Imaging Med Surg

View full text Add to dashboard Cite

Background: The segmentation of breast ultrasound (US) images has been a challenging task, mainly due to limited data and the inherent image characteristics involved, such as low contrast and speckle noise.Although convolutional neural network-based (CNN-based) methods have made significant progress over the past decade, they lack the ability to model long-range interactions. Recently, the transformer method has been successfully applied to the tasks of computer vision. It has a strong ability to capture distant interactions. However, most transformer-based methods with excellent performance rely on pre-training on large datasets, making it infeasible to directly apply them to medical images analysis, especially that of breast US images with limited high-quality labels. Therefore, it is of great significance to find a robust and efficient transformer-based method for use on small breast US image datasets. Methods:We developed a dilated transformer (DT) method which mainly uses the proposed residual axial attention layers to build encoder blocks and the introduced dilation module (DM) to further increase the receptive field. We evaluated the proposed method on 2 breast US image datasets using the 5-fold crossvalidation method. Dataset A was a public dataset with 562 images, while dataset B was a private dataset with 878 images. Ground truth (GT) was delineated by 2 radiologists with more than 5 years of experience. The evaluation was followed by related ablation experiments. Results:The DT was found to be comparable with the state-of-the-art (SOTA) CNN-based method and outperformed the related transformer-based method, medical transformer (MT), on both datasets. Especially on dataset B, the DT outperformed the MT on metrics of Jaccard index (JI) and Dice similarity coefficient (DSC) by 2.67% and 4.68%, respectively. Meanwhile, when compared with Unet, the DT improved JI and DSC by 4.89% and 4.66%, respectively. Moreover, the results of the ablation experiments showed that each add-on part of the DT is important and contributes to the segmentation accuracy. Conclusions:The proposed transformer-based method could achieve advanced segmentation performance on different small breast US image datasets without pretraining.

show abstract

What is new in computer vision and artificial intelligence in medical image analysis applications

Cited by 50 publications

References 81 publications

Detangling the interrelations between MAFLD, insulin resistance, and key hormones

Detangling the interrelations between MAFLD, insulin resistance, and key hormones

Automatic coronary artery calcium scoring on routine chest computed tomography (CT): comparison of a deep learning algorithm and a dedicated calcium scoring CT

Dilated transformer: residual axial attention for breast ultrasound image segmentation

Contact Info

Product

Resources

About