RSS-Feed abonnieren
DOI: 10.1055/s-0040-1721755
Response: Comment and Update on “Using Artificial Intelligence to Manage Thrombosis Research, Diagnosis, and Clinical Management”
![](https://www.thieme-connect.de/media/sth/202101/lookinside/thumbnails/10-1055-s-0040-1721755_02825-1.jpg)
The Comment and Update on: “Using Artificial Intelligence to Manage Thrombosis Research, Diagnosis, and Clinical Management” by Martins et al highlights the potential of artificial neural networks (ANNs) in predictions of recurrent venous thromboembolisms (VTEs).[1] [2] Their letter provides an update on recent studies that have applied artificial intelligence (AI) techniques for developing predictive models on VTE risk factors.[2]
The self-learning capabilities of ANN models of AI make this a most interesting tool with numerous applications. ANN models can take multiple parameters into consideration, as hidden layers between input and output layers take in a set of weighted inputs and produce an output based on a predefined set of learning rules and an internal weighting system. These hidden layers are those that make ANN models superior to any other machine learning algorithms, but this comes with a downside of complexity and time consumption.[1] Martins et al reported the ability of principal component analysis (PCA) to reduce the number of predictors in their ANN model from 39 to 18 factors.[2] This not only eases the collection of clinical parameters but also lowers the training time of the model without any alterations to its accuracy. They found that factors as simple as red blood cell count, white blood cell count, hematocrit, red cell distribution width, glucose, lipids, natural anticoagulants, creatinine, age, as well as first deep vein thrombosis data (distal/proximal, D-dimer, and time of anticoagulation) could be used to develop simple clinical decision support systems to predict the rates of recurrent VTE. Furthermore, the authors compared their observations from various techniques such as weight analysis, PCA, and designs of experiments to observe that there are few factors that were repeated in all of the techniques and could be the most important ones in the prediction of recurrent VTE.[2] [3] Overall, the usage of PCA-ANN models in development of newer clinical decision support systems could not only advance the robustness of the model but might also improve its prediction quality.
Publikationsverlauf
Artikel online veröffentlicht:
01. Februar 2021
© 2021. Thieme. All rights reserved.
Thieme Medical Publishers, Inc.
333 Seventh Avenue, 18th Floor, New York, NY 10001, USA
-
References
- 1 Mishra A, Ashraf MZ. Using artificial intelligence to manage thrombosis research, diagnosis, and clinical management. Semin Thromb Hemost 2020; 46 (04) 410-418
- 2 Martins TD, Annichino-Bizzacchi JM, Romano AVC, Filho RM. Principal component analysis on recurrent venous thromboembolism. Clin Appl Thromb Hemost 2019; 25: 1076029619895323
- 3 Martins TD, Annichino-Bizzacchi JM, Romano AVC, Maciel Filho R. Artificial neural networks for prediction of recurrent venous thromboembolism. Int J Med Inform 2020; 141: 104221