Using Machine Learning to Identify Intravenous Contrast Phases on Computed Tomography

2021 ◽  
pp. 106603
Author(s):  
Raouf Muhamedrahimov ◽  
Amir Bar ◽  
Jonathan Laserson ◽  
Ayelet Akselrod-Ballin ◽  
Eldad Elnekave
Keyword(s):  
Machine Learning ◽  
Computed Tomography ◽  
Intravenous Contrast

Use of deep learning to predict acute kidney injury after intravenous contrast media administration (Preprint)

2021 ◽  
Author(s):  
Donghwan Yun ◽  
Semin Cho ◽  
Yong Chul Kim ◽  
Dong Ki Kim ◽  
Kook-Hwan Oh ◽  
...  
Keyword(s):  
Machine Learning ◽  
Computed Tomography ◽  
Deep Learning ◽  
Contrast Media ◽  
Learning Algorithm ◽  
Kidney Injury ◽  
Time Varying ◽  
Intravenous Contrast ◽  
Deep Learning Algorithm ◽  
Intravenous Contrast Media

BACKGROUND Precise prediction of contrast media-induced acute kidney injury (CIAKI) is an important issue because of its relationship with worse outcomes. OBJECTIVE Herein, we examined whether a deep learning algorithm could predict the risk of intravenous CIAKI better than other machine learning and logistic regression models in patients undergoing computed tomography. METHODS A total of 14,185 cases that underwent intravenous contrast media for computed tomography under the preventive and monitoring facility in Seoul National University Hospital were reviewed. CIAKI was defined as an increase in serum creatinine ≥0.3 mg/dl within 2 days and/or ≥50% within 7 days. Using both time-varying and time-invariant features, machine learning models, such as the recurrent neural network (RNN), light gradient boosting machine, extreme boosting machine, random forest, decision tree, support vector machine, κ-nearest neighboring, and logistic regression, were developed using a training set, and their performance was compared using the area under the receiver operating characteristic curve (AUROC) in a test set. RESULTS CIAKI developed in 261 cases (1.8%). The RNN model had the highest AUROC value of 0.755 (0.708–0.802) for predicting CIAKI, which was superior to those obtained from other machine learning models. Although CIAKI was defined as an increase in serum creatinine ≥0.5 mg/dl and/or ≥25% within 3 days, the highest performance was achieved in the RNN model with an AUROC of 0.716 (0.664–0.768). In the feature ranking analysis, albumin level was the most highly contributing factor to RNN performance, followed by time-varying kidney function. CONCLUSIONS Application of a deep learning algorithm improves the predictability of intravenous CIAKI after computed tomography, representing a basis for future clinical alarming and preventive systems.

scholarly journals Improved contrast and noise of megavoltage computed tomography (MVCT) through cycle‐consistent generative machine learning

2020 ◽  
Author(s):  
Luciano Vinas ◽  
Jessica Scholey ◽  
Martina Descovich ◽  
Vasant Kearney ◽  
Atchar Sudhyadhom
Keyword(s):  
Machine Learning ◽  
Computed Tomography ◽  
Megavoltage Computed Tomography

Radiology During the COVID-19 Pandemic: Mapping Radiology Literature in 2020

2021 ◽  
Vol 18 ◽  
Author(s):  
Nosaiba Al-Ryalat ◽  
Lna Malkawi ◽  
Ala'a Abu Salhiyeh ◽  
Faisal Abualteen ◽  
Ghaida Abdallah ◽  
...  
Keyword(s):  
Machine Learning ◽  
Computed Tomography ◽  
Nuclear Medicine ◽  
Medical Imaging ◽  
Core Collection ◽  
Web Of Science ◽  
Vos Viewer ◽  
Literature Mapping ◽  
Link Strength ◽  
Linkage Strength

Objectives: Our aim was to assess articles published in the field of radiology, nuclear medicine, and medical imaging in 2020, analyzing the linkage of radiology-related topics with coronavirus disease 2019 (COVID-19) through literature mapping, along with a bibliometric analysis for publications. Methods: We performed a search on Web of Science Core Collection database for articles in the field of radiology, nuclear medicine, and medical imaging published in 2020. We analyzed the included articles using VOS viewer software, where we analyzed the co-occurrence of keywords, which represents major topics discussed. Of the resulting topics, literature map created, and linkage analysis done. Results: A total of 24,748 articles were published in the field of radiology, nuclear medicine, and medical imaging in 2020. We found a total of 61,267 keywords, only 78 keywords occurred more than 250 times. COVID-19 had 449 occurrences, 29 links, with a total link strength of 271. MRI was the topic most commonly appearing in 2020 radiology publications, while “computed tomography” has the highest linkage strength with COVID-19, with a linkage strength of 149, representing 54.98% of the total COVID-19 linkage strength, followed by “radiotherapy, and “deep and machine learning”. The top cited paper had a total of 1,687 citations. Nine out of the 10 most cited articles discussed COVID-19 and included “COVID-19” or “coronavirus” in their title, including the top cited paper. Conclusion: While MRI was the topic that dominated, CT had the highest linkage strength with COVID-19 and represent the topic of top cited articles in 2020 radiology publications.

Sign in / Sign up

Export Citation Format

Share Document