Probabilistic Heart Rate Prediction for Cardiac CT Scans
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current CT imaging systems face challenges in accurately predicting heart rates for cardiac CT scans, leading to potential motion artifacts due to the heart's movement, especially when the predicted heart rate is inaccurate.
Innovation Solution
A method for predicting upper and lower threshold values of a heart rate based on HR time series data collected during a pre-scan period, using a machine learning model that takes into account a specified confidence level, allowing for more precise configuration of CT system parameters for cardiac scans.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the gantry rotation time is reduced to minimize motion artifacts, then image quality is improved, but the system requires more precise timing and control
Solution Approach 1:
The system performs preliminary HR prediction and confidence level calculation before the actual cardiac scan to determine the optimal scanning window. This advance preparation allows the system to pre-configuring scan parameters and timing, reducing the complexity of real-time control during the actual imaging process.
Solution Approach 2:
The system dynamically adjusts scan timing based on predicted heart rate and confidence levels. By making the timing adaptive rather than fixed, the system can optimize for each patient's specific cardiac characteristics while maintaining manageable control complexity through automated adjustment.
2Productivity
If a single target HR prediction is used to configure the cardiac scan, then the scan can be performed quickly, but the scan may fail due to HR variance
Solution Approach 1:
Instead of using a single target HR value, the system changes the parameter approach by predicting an upper threshold HR and a lower threshold HR, creating a HR range. This parameter transformation from point estimate to interval estimate maintains scan efficiency while improving reliability by accounting for HR variance.
Solution Approach 2:
The system calculates a confidence level that serves as a cushion against HR prediction uncertainty. By incorporating this confidence buffer into the scan configuration, the system prepares for potential HR variations in advance, ensuring the scan remains successful even if the actual HR deviates from the predicted value.
3Manufacturing precision
If the scan duration is extended to capture the heart during minimal movement, then image quality is improved, but the patient is exposed to more motion and the heart may move during the scan
Solution Approach 1:
The system performs preliminary prediction of the upper and lower threshold HR values and calculates the confidence level before initiating the scan. This advance calculation allows the system to precisely define the scanning window, ensuring the scan is performed during the optimal cardiac phase while minimizing both duration and patient motion exposure.
Solution Approach 2:
The system uses the confidence level parameter to optimize the balance between scan duration and motion minimization. By adjusting the confidence level, the system can fine-tune the scanning parameters to achieve the shortest possible scan duration that still guarantees image quality, thereby reducing patient motion during the procedure.
Data Source
AI summary
Methods and systems are provided to predict an upper threshold and a lower threshold of a heart rate (HR) of a patient of a computed tomography (CT) imaging system using a machine learning (ML) model, based on HR time series data collected over a duration prior to a cardiac scan. The ML model takes as an additional input a desired probabilistic certainty (e.g., statistical confidence level) that the predicted upper and lower HR thresholds will be accurate, provided by an operator of the CT imaging system. Before the start of the imaging scan, the predicted upper and lower HR thresholds are used to set the start exposure and end exposure times, to ensure that the requested cardiac phases are acquired.