Improved identification of tumors in 18F-FDG-PET examination by normalizing the standard uptake in the liver based on blood test data

Category	Detail
Condition	Hepatic tumors and liver abnormalities
Key Mechanisms	Normalization of liver standardized uptake value (SUV) in 18F-FDG-PET imaging using blood test-derived clinical variables and machine learning (LASSO regression) to generate personalized Z-score maps
Target Population	Adult patients undergoing whole-body PET/CT screening for liver tumor detection
Care Setting	Hospital-based imaging and diagnostic radiology departments

Liver SUV in PET imaging is influenced by multiple clinical variables including blood test results, age, BMI, and blood glucose level.
A machine learning-based normalization method using LASSO regression was developed to estimate personalized liver SUV mean and standard deviation from non-image clinical variables.
Normalized Z-score maps of liver SUV improve detection and diagnosis of hepatic masses compared to conventional SUV measurements.

Use 18F-FDG-PET/CT imaging combined with liver function blood tests (AST, ALT, ALP, bilirubin, albumin) to evaluate liver abnormalities.
Consider normalization of liver SUV using patient-specific clinical variables to enhance tumor detection accuracy.

Incorporate personalized Z-score maps derived from normalized liver SUV for daily image interpretation by physicians.
Use blood test results obtained on the same day as PET/CT scans to inform SUV normalization.

Perform repeated PET/CT imaging with concurrent blood tests to monitor liver tumor response and progression using normalized SUV metrics.

Be aware that unnormalized liver SUV can be confounded by physiological and clinical variables, potentially leading to missed hepatic lesions.
Ensure quality control of PET/CT imaging and blood test accuracy to maintain reliability of normalization.

Adults undergoing whole-body PET/CT screening for liver tumor detection

Normalization of liver SUV using blood test results and clinical variables may improve diagnostic sensitivity and specificity for hepatic tumors.

Obtain comprehensive blood test panels including liver function tests on the same day as PET/CT imaging.
Use machine learning models such as LASSO regression to integrate multiple clinical variables for SUV normalization.
Replace conventional SUV maps with personalized Z-score maps for standardized and improved interpretation of hepatic FDG uptake.
Collaborate between radiologists for independent image review and consensus diagnosis to improve accuracy.

Clinical Scorecard: Enhanced Tumor Detection in 18F-FDG-PET Imaging Through Liver Uptake Normalization Using Blood Test Results