Modeling Patient-Informed Liver Contrast Perfusion in Contrast-enhanced Computed Tomography

Development and Clinical Evaluation of a Contrast Optimizer for Contrast-Enhanced CT Imaging of the Liver

OBJECTIVE

Patient characteristics, iodine injection, and scanning parameters can impact the quality and consistency of contrast enhancement of hepatic parenchyma in CT imaging. Improving the consistency and adequacy of contrast enhancement can enhance diagnostic accuracy and reduce clinical practice variability, with added positive implications for safety and cost-effectiveness in the use of contrast medium. We developed a clinical tool that uses patient attributes (height, weight, sex, age) to predict hepatic enhancement and suggest alternative injection/scanning parameters to optimize the procedure.

METHODS

The tool was based on a previously validated neural network prediction model that suggested adjustments for patients with predicted insufficient enhancement. We conducted a prospective clinical study in which we tested this tool in 24 patients aiming for a target portal-venous parenchyma CT number of 110 HU ± 10 HU.

RESULTS

Out of the 24 patients, 15 received adjustments to their iodine contrast injection parameters, resulting in median reductions of 8.8% in volume and 9.1% in injection rate. The scan delays were reduced by an average of 42.6%. We compared the results with the patients' previous scans and found that the tool improved consistency and reduced the number of underenhanced patients. The median enhancement remained relatively unchanged, but the number of underenhanced patients was reduced by half, and all previously overenhanced patients received enhancement reductions.

CONCLUSIONS

Our study showed that the proposed patient-informed clinical framework can predict optimal contrast enhancement and suggest empiric injection/scanning parameters to achieve consistent and sufficient contrast enhancement of hepatic parenchyma. The described GUI-based tool can prospectively inform clinical decision-making predicting optimal patient's hepatic parenchyma contrast enhancement. This reduces instances of nondiagnostic/insufficient enhancement in patients.

A patient-informed approach to predict iodinated-contrast media enhancement in the liver

OBJECTIVE

To devise a patient-informed time series model that predicts liver contrast enhancement, by integrating clinical data and pharmacokinetics models, and to assess its feasibility to improve enhancement consistency in contrast-enhanced liver CT scans.

METHODS

The study included 1577 Chest/Abdomen/Pelvis CT scans, with 70-30% training/validation-testing split. A Gaussian function was used to approximate the early arterial, late arterial, and the portal venous phases of the contrast perfusion curve of each patient using their respective bolus tracking and diagnostic scan data. Machine learning models were built to predict the Gaussian parameters of each patient using the patient attributes (weight, height, age, sex, BMI). Pearson's coefficient, mean absolute error, and root mean squared error were used to assess the prediction accuracy.

RESULTS

The integration of the pharmacokinetics model with a two-layered neural network achieved the highest prediction accuracy on the test data (R2 = 0.61), significantly exceeding the performance of the pharmacokinetics model alone (R2 = 0.11). Applying the model demonstrated that adjusting the contrast administration directed by the model may reduce clinical enhancement inconsistency by up to 40 %.

CONCLUSIONS

A new model using a Gaussian function and supervised machine learning can be used to build liver parenchyma contrast enhancement prediction model. The model can have utility in clinical settings to optimize and improve consistency in contrast-enhanced liver imaging.