Deep-learning-assisted analysis of echocardiographic videos improves predictions of all-cause mortality

Cerna, Alvaro E. Ulloa; Jing, Linyuan; Good, Christopher W.; vanMaanen, David P.; Raghunath, Sushravya; Suever, Jonathan D.; Nevius, Christopher D.; Wehner, Gregory J; Hartzel, Dustin N.; Leader, Joseph B.; Alsaid, Amro; Patel, Aalpen A.; Kirchner, H. Lester; Pfeifer, John M.; Carry, Brendan; Pattichis, Marios S.; Haggerty, Christopher M.; Fornwalt, Brandon K.

doi:10.1038/s41551-020-00667-9

Cited by 41 publications

(13 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, AI should overcome resistances to its wide clinical acceptance through the demonstration of its usefulness for the trained and untrained professionals, improving the integration of imaging information into the wider clinical arena 1–6,11–14 . Notably, some Authors have recently demonstrated how AI can be useful both to provide new insights, and to teach novice trainees 51,52 . To date, this field represents an area of great focus.…”

Section: Challenges Pitfalls and Limitations Which Slow Down The Deve...mentioning

confidence: 99%

Artificial intelligence applied to cardiovascular imaging, a critical focus on echocardiography: The point‐of‐view from “the other side of the coin”

Dell'Angela

Nicolosi²

2022

J of Clinical Ultrasound

View full text Add to dashboard Cite

Cardiovascular imaging has achieved a crucial role in the management of cardiovascular diseases. In this field, echocardiography advantages include wide availability, portability, and affordability, at a relatively low cost. However, echocardiographic assessment requires highly trained operators, and implies high observer variability, as compared with the other cardiac imaging modalities. Hence, artificial intelligence might be extremely helpful. From the point‐of‐view of the peripheral “Spoke” Hospital potential user (“the other side of the coin”), artificial intelligence development appears very slow in the clinical arena. Many limitations are still present, and require full involvement, cooperation, and coordination of professional operators into Hub‐and‐Spoke network.

show abstract

Section: Challenges Pitfalls and Limitations Which Slow Down The Deve...mentioning

confidence: 99%

Artificial intelligence applied to cardiovascular imaging, a critical focus on echocardiography: The point‐of‐view from “the other side of the coin”

Dell'Angela

Nicolosi²

2022

J of Clinical Ultrasound

View full text Add to dashboard Cite

show abstract

“…The model's predictions were superior to the commonly utilized pooled cohort equations, the Seattle Heart Failure Score (measured in an independent dataset of over 2400 individuals with heart failure who had over 3380 echocardiograms performed on them). Predictions from the model also outperformed a machine learning model which involved 58 human‐derived variables from echocardiographic studies and 100 clinical variables taken from patient records that were stored electronically 50 . ML was also used by Samad et al.…”

Section: Mortality Prediction From Echocardiographic Imagesmentioning

confidence: 99%

Artificial intelligence in echocardiography: Review and limitations including epistemological concerns

2022

View full text Add to dashboard Cite

Background and Purpose:In this review we describe the use of artificial intelligence in the field of echocardiography. Various aspects and terminologies used in artificial intelligence are explained in an easy-to-understand manner and supplemented with illustrations related to echocardiography. Limitations of artificial intelligence, including epistemologic concerns from a philosophical standpoint, are also discussed. Methods:A narrative review of relevant papers was conducted. Conclusion:We provide an overview of the usefulness of artificial intelligence in echocardiography and focus on how it can supplement current day-to-day clinical practice in the assessment of various cardiovascular disease entities. On the other hand, there are significant limitations, including epistemological concerns, which need to be kept in perspective.

show abstract

“…The data characteristics are summarized in Table 1. Disease prevalence was based on either custom phenotypes [35]- [37] or models developed by the Electronic Medical Records and Genomics (eMERGE) Network [36], [38].…”

Section: Datasetsmentioning

confidence: 99%

Generalizability and quality control of deep learning-based 2D echocardiography segmentation models in a large clinical dataset

Zhang

Cerna

Stough

et al. 2022

Int J Cardiovasc Imaging

Self Cite

View full text Add to dashboard Cite

Use of machine learning for automated annotation of heart structures from echocardiographic videos is an active research area, but understanding of comparative, generalizable performance among models is lacking. This study aimed to 1) assess the generalizability of ve state-of-the-art machine learning-based echocardiography segmentation models within a large clinical dataset, and 2) test the hypothesis that a quality control (QC) method based on segmentation uncertainty can further improve segmentation results. Five models were applied to 47,431 echocardiography studies that were independent from any training samples. Chamber volume and mass from model segmentations were compared to clinicallyreported values. The median absolute errors (MAE) in left ventricular (LV) volumes and ejection fraction exhibited by all ve models were comparable to reported inter-observer errors (IOE). MAE for left atrial volume and LV mass were similarly favorable to respective IOE for models trained for those tasks. A single model consistently exhibited the lowest MAE in all ve clinically-reported measures. We leveraged the 10-fold cross-validation training scheme of this best-performing model to quantify segmentation uncertainty for potential application as QC. We observed that ltering segmentations with high uncertainty improved segmentation results, leading to decreased volume/mass estimation errors. The addition of contour-convexity lters further improved QC e ciency. In conclusion, ve previously published echocardiography segmentation models generalized to a large, independent clinical datasetsegmenting one or multiple cardiac structures with overall accuracy comparable to manual analyseswith variable performance. Convexity-reinforced uncertainty QC e ciently improved segmentation performance and may further facilitate the translation of such models.

show abstract

Deep-learning-assisted analysis of echocardiographic videos improves predictions of all-cause mortality

Cited by 41 publications

References 53 publications

Artificial intelligence applied to cardiovascular imaging, a critical focus on echocardiography: The point‐of‐view from “the other side of the coin”

Artificial intelligence applied to cardiovascular imaging, a critical focus on echocardiography: The point‐of‐view from “the other side of the coin”

Artificial intelligence in echocardiography: Review and limitations including epistemological concerns

Generalizability and quality control of deep learning-based 2D echocardiography segmentation models in a large clinical dataset

Contact Info

Product

Resources

About