Evaluation of the feasibility of explainable computer-aided detection of cardiomegaly on chest radiographs using deep learning

Lee, Mu Sook; Kim, Yong Soo; Kim, Minki; Usman, Muhammad; Byon, Shi Sub; Kim, Sung Hyun; Lee, Byoung Il; Lee, Byoung-Dai

doi:10.1038/s41598-021-96433-1

Cited by 22 publications

(26 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Discussionmentioning

confidence: 99%

“…Erroneous results are a problem faced by all DL models in scenarios where the contours are not clearly demarcated, as such cardiac contour has a more major impact than the thoracic contour. 22 Saiviroonporn et al 23 used the VGG-16 U-Net model to assess the impact of AI-enabled reporting of CTR. Their findings concluded that AI alone had higher variations than human readers, but it could support the radiologist by reducing observer variation and operation time.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Observer performance evaluation of the feasibility of a deep learning model to detect cardiomegaly on chest radiographs

Ajmera

Kharat

Gupte³

et al. 2022

Acta Radiologica Open

View full text Add to dashboard Cite

Background Cardiothoracic ratio (CTR) is the ratio of the diameter of the heart to the diameter of the thorax. An abnormal CTR (>0.55) is often an indicator of an underlying pathological condition. The accurate prediction of an abnormal CTR chest X-rays (CXRs) aids in the early diagnosis of clinical conditions. Purpose We propose a deep learning (DL)-based model for automatic CTR calculation to assist radiologists with rapid diagnosis of cardiomegaly and thus optimise the radiology flow. Material and Methods The study population included 1012 posteroanterior CXRs from a single institution. The Attention U-Net DL architecture was used for the automatic calculation of CTR. An observer performance test was conducted to assess the radiologist’s performance in diagnosing cardiomegaly with and without artificial intelligence assistance. Results U-Net model exhibited a sensitivity of 0.80 [95% CI: 0.75, 0.85], specificity >99%, precision of 0.99 [95% CI: 0.98, 1], and a F1 score of 0.88 [95% CI: 0.85, 0.91]. Furthermore, the sensitivity of the reviewing radiologist in identifying cardiomegaly increased from 40.50% to 88.4% when aided by the AI-generated CTR. Conclusion Our segmentation-based AI model demonstrated high specificity (>99%) and sensitivity (80%) for CTR calculation. The performance of the radiologist on the observer performance test improved significantly with provision of AI assistance. A DL-based segmentation model for rapid quantification of CTR can therefore have significant potential to be used in clinical workflows by reducing radiologists’ burden and alerting to an abnormal enlarged heart early on.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Observer performance evaluation of the feasibility of a deep learning model to detect cardiomegaly on chest radiographs

Ajmera

Kharat

Gupte³

et al. 2022

Acta Radiologica Open

View full text Add to dashboard Cite

show abstract

“…The extracted regions of interest are subsequently used to compute the CTR and a threshold of 0.50 is used in order to distinguish between normal and cardiomegaly instances. A similar approach consisting of optimizing 2 distinctive segmentation models in order to extract both cardiac and thoracic areas of interest before computing the CTR is proposed by Lee et al ( 2021 ).…”

Section: Related Workmentioning

confidence: 99%

Unsupervised domain adaptation for the detection of cardiomegaly in cross-domain chest X-ray images

Thiam¹,

Lausser²,

Kloth³

et al. 2023

Front. Artif. Intell.

View full text Add to dashboard Cite

In recent years, several deep learning approaches have been successfully applied in the field of medical image analysis. More specifically, different deep neural network architectures have been proposed and assessed for the detection of various pathologies based on chest X-ray images. While the performed assessments have shown very promising results, most of them consist in training and evaluating the performance of the proposed approaches on a single data set. However, the generalization of such models is quite limited in a cross-domain setting, since a significant performance degradation can be observed when these models are evaluated on data sets stemming from different medical centers or recorded under different protocols. The performance degradation is mostly caused by the domain shift between the training set and the evaluation set. To alleviate this problem, different unsupervised domain adaptation approaches are proposed and evaluated in the current work, for the detection of cardiomegaly based on chest X-ray images, in a cross-domain setting. The proposed approaches generate domain invariant feature representations by adapting the parameters of a model optimized on a large set of labeled samples, to a set of unlabeled images stemming from a different data set. The performed evaluation points to the effectiveness of the proposed approaches, since the adapted models outperform optimized models which are directly applied to the evaluation sets without any form of domain adaptation.

show abstract

“…Although classical image processing-based studies have reported promising results on limited private datasets, these methods lack generalization ability, making them inefficient for real-time application. On the other hand, deep learning-based methods have made vast inroads into various computer-aided medical applications [16], such as disease detection [17] and the segmentation of affected regions [18]. Consequently, deep learning has also been applied for mandibular canal segmentation to boost performance.…”

Section: Related Workmentioning

confidence: 99%

Dual-Stage Deeply Supervised Attention-Based Convolutional Neural Networks for Mandibular Canal Segmentation in CBCT Scans

Usman

Rehman

Saleem

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

Accurate segmentation of mandibular canals in lower jaws is important in dental implantology. Medical experts manually determine the implant position and dimensions from 3D CT images to avoid damaging the mandibular nerve inside the canal. In this paper, we propose a novel dual-stage deep learning-based scheme for the automatic segmentation of the mandibular canal. In particular, we first enhance the CBCT scans by employing the novel histogram-based dynamic windowing scheme, which improves the visibility of mandibular canals. After enhancement, we designed 3D deeply supervised attention UNet architecture for localizing the Volumes Of Interest (VOIs), which contain the mandibular canals (i.e., left and right canals). Finally, we employed the Multi-Scale input Residual UNet (MSiR-UNet) architecture to segment the mandibular canals using VOIs accurately. The proposed method has been rigorously evaluated on 500 and 15 CBCT scans from our dataset and from the public dataset, respectively. The results demonstrate that our technique improves the existing performance of mandibular canal segmentation to a clinically acceptable range. Moreover, it is robust against the types of CBCT scans in terms of field of view.

show abstract

Evaluation of the feasibility of explainable computer-aided detection of cardiomegaly on chest radiographs using deep learning

Cited by 22 publications

References 12 publications

Observer performance evaluation of the feasibility of a deep learning model to detect cardiomegaly on chest radiographs

Observer performance evaluation of the feasibility of a deep learning model to detect cardiomegaly on chest radiographs

Unsupervised domain adaptation for the detection of cardiomegaly in cross-domain chest X-ray images

Dual-Stage Deeply Supervised Attention-Based Convolutional Neural Networks for Mandibular Canal Segmentation in CBCT Scans

Contact Info

Product

Resources

About