Quantifying Statistical Significance of Neural Network-based Image Segmentation by Selective Inference

Duy, Vo Nguyen Le; Iwazaki, Shogo; Takeuchi, Ichiro

doi:10.48550/arxiv.2010.01823

“…2 were estimated using a gradient descent-based optimizer [43]. Although the association between the local ABP shapes and hypotension occurrence can be interpreted based on these coefficient values, a deep learning model without p-values cannot provide statistical significance for this relationship [44]. Therefore, an additional logistic regression analysis based on the Newton-Raphson optimizer was applied to obtain the pvalues of these coefficients.…”

Section: B Evaluation On Interpretability 1) Statistical Significance...mentioning

confidence: 99%

Intraoperative Hypotension Prediction Based on Features Automatically Generated Within an Interpretable Deep Learning Model

Hwang¹,

Park²,

Kim³

et al. 2023

Preprint

View full text Add to dashboard Cite

Monitoring arterial blood pressure (ABP) in anesthetized patients is crucial for preventing hypotension, which can lead to adverse clinical outcomes. Thus, several efforts have been made to develop an artificial intelligence-based hypotension prediction index. Nevertheless, the use of these indices is limited because they may not provide a convincing interpretation of the association between predictors and hypotension. Herein, we developed an interpretable deep learning model that forecasts hypotension occurrences 10 min before a given 90 s ABP record. Internal and external validations of model performance reported the area under the receiver operating characteristic curve (AUC) as 0.9145 and 0.9035, respectively. Furthermore, the hypotension prediction mechanism can be physiologically interpreted by using predictors representing ABP trends that are automatically generated in the proposed model. Ultimately, we demonstrate high-applicability of a deep learning model that has a high accuracy performance and provides an interpretation of the association between ABP trends and hypotension in clinical practices.

show abstract

“…2 were estimated using the gradient descent based optimizer [39]. Although the association between the local ABP shapes and hypotension occurrence can be interpreted by these coefficient values, the deep learning model without p-values cannot provide statistical significance for this relationship [40]. Therefore, additional logistic regression analysis, based on the Newton-Raphson optimizer, was conducted to estimate the pvalues of these coefficients.…”

Section: B Evaluation On Interpretability 1) Statistical Significance...mentioning

confidence: 99%

Intraoperative Hypotension Prediction Based on Features Automatically Generated Within an Interpretable Deep Learning Model

Kim¹,

Kim²

2023

Preprint

1

0

View full text Add to dashboard Cite

Monitoring arterial blood pressure (ABP) in anesthetized patients is crucial for preventing hypotension, which can lead to adverse clinical outcomes. Thus, several efforts have been made to develop an artificial intelligence-based hypotension prediction index. Nevertheless, the use of these indices is limited because they may not provide a convincing interpretation of the association between predictors and hypotension. Herein, we developed an interpretable deep learning model that forecasts hypotension occurrences 10 min before a given 90 s ABP record. Internal and external validations of model performance reported the area under the receiver operating characteristic curve (AUC) as 0.9145 and 0.9035, respectively. Furthermore, the hypotension prediction mechanism can be physiologically interpreted by using predictors representing ABP trends that are automatically generated in the proposed model. Ultimately, we demonstrate high-applicability of a deep learning model that has a high accuracy performance and provides an interpretation of the association between ABP trends and hypotension in clinical practices.

show abstract

Intraoperative Hypotension Prediction Based on Features Automatically Generated Within an Interpretable Deep Learning Model

Kim¹,

Kim²

2022

Preprint

0

View full text Add to dashboard Cite

Monitoring arterial blood pressure (ABP) in anesthetized patients is crucial for preventing hypotension, which can lead to adverse clinical outcomes. Thus, several efforts have been made to develop an artificial intelligence-based hypotension prediction index. Nevertheless, the use of these indices is limited because they may not provide a convincing interpretation of the association between predictors and hypotension. Herein, we developed an interpretable deep learning model that forecasts hypotension occurrences 10 min before a given 90 s ABP record. Internal and external validations of model performance reported the area under the receiver operating characteristic curve (AUC) as 0.9145 and 0.9035, respectively. Furthermore, the hypotension prediction mechanism can be physiologically interpreted by using predictors representing ABP trends that are automatically generated in the proposed model. Ultimately, we demonstrate high-applicability of a deep learning model that has a high accuracy performance and provides an interpretation of the association between ABP trends and hypotension in clinical practices.

show abstract

Quantifying Statistical Significance of Neural Network-based Image Segmentation by Selective Inference

Cited by 3 publications

References 26 publications

Intraoperative Hypotension Prediction Based on Features Automatically Generated Within an Interpretable Deep Learning Model

Intraoperative Hypotension Prediction Based on Features Automatically Generated Within an Interpretable Deep Learning Model

Intraoperative Hypotension Prediction Based on Features Automatically Generated Within an Interpretable Deep Learning Model

Intraoperative Hypotension Prediction Based on Features Automatically Generated Within an Interpretable Deep Learning Model

Contact Info

Product

Resources

About