Liver Radioembolization: Predicting Lung Shunt Fraction with Contrast-Enhanced CT, Eliminating the Need for 99mTc-MAA
Mehadji, B.; Marx, T.; Johnson Carter, A. I.; Goldman, R.; T. Vu, C.; Roncali, E.
Show abstract
1.Accurate estimation of the Lung Shunt Fraction (LSF) is a standard of care in yttrium-90 (90Y) radioembolization treatment planning to prevent excessive lung irradiation due to arterio-venous shunting in the liver. LSF is assessed using 99mTc macroaggregated albumin (99mTc-MAA) imaging, but this approach adds risk, complexity, and expense to the treatment planning. This study investigates the potential of Contrast-Enhanced Computed Tomography (CECT) as a non-invasive alternative for LSF estimation. MethodsWe developed a novel metric to estimate the LSF from differences between CECT phases, based on the premise that LSF results from tumor angiogenesis and arterio-venous shunting that create excessive local opacity in the CECT arterial phase. Hypervascular volumes were identified by subtracting the portal phase from the arterial phase and contouring the corresponding regions using an density threshold. The perfused volume corresponded to the lobe and to the volume injected with 99mTc-MAA. We conducted a retrospective analysis involving 30 liver cancer patients who underwent 4-phase liver CECT and 99mTc-MAA imaging prior to 90Y radioembolization to study several markers based on the subtraction images. The patient cohort included several types of liver tumors (primary or metastatic), predominantly hepatocellular carcinoma. Correlations between the hypervascular-to-perfused volume ratio from CECT and LSF values from standard-of-care 99mTc-MAA planar imaging were assessed. ResultsThe strongest correlation was obtained between the hypervascular-to-perfused volume ratio from CECT and LSF values from 99mTc-MAA planar gamma imaging (R{superscript 2}=0.95). The best correlation was demonstrated when the hypervascular volume was corrected by the hypoxic volume and then normalized by the perfused volume. In contrast, the tumor volume did not show a strong correlation with LSF (R{superscript 2}=0.38). Moreover, the predictive power of the hypervascular-to-perfused volume ratio on LSF demonstrated a 1-sigma uncertainty of 3%. ConclusionCECT-based LSF estimation shows promise as a non-invasive alternative to 99mTc-MAA imaging in 90Y treatment planning. Further validation with prospective studies on larger cohorts is necessary to confirm the accuracy and safety of this approach.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Inconsistency of AI in Intracranial Aneurysm Detection with Varying Dose and Image Reconstruction 94%
- Developing a Fully Automated Imaging Biomarker for HCC Risk Assessment via MRI-Based Tumor Segmentation and EPM 93%
- Identifying relationships between imaging phenotypes and lung cancer-related mutation status: EGFR and KRAS 93%
Similar papers in this journal
- Fully Automated Explainable Abdominal CT Contrast Media Phase Classification Using Organ Segmentation and Machine Learning 94%
- PSMA-Hornet: fully-automated, multi-target segmentation of healthy organs in PSMA PET/CT images 93%
- Phase Recognition in Contrast-Enhanced CT Scans based on Deep Learning and Random Sampling 92%
Similar papers in this journal
- Radiomic-Based Approaches in the Multi-metastatic Setting: A Quantitative Review 93%
- Detection of local microvascular proliferation in IDH wild-type Glioblastoma using relative Cerebral Blood Volume 91%
- Inference of core needle biopsy whole slide images requiring definitive therapy for prostate cancer 89%
Similar papers in this journal
- Reproducible spectral CT thermometry with liver-mimicking phantoms for image-guided thermal ablation 95%
- Patient-derived PixelPrint phantoms for evaluating clinical imaging performance of a deep learning CT reconstruction algorithm 94%
- Contrast-enhanced photon-counting micro-CT of tumor xenograft models 93%
Similar papers in this journal
- Deep-Learning-Based Generation of Synthetic High-Resolution MRI from Low-Resolution MRI for Use in Head and Neck Cancer Adaptive Radiotherapy 92%
- Segmentation stability of human head and neck medical images for radiotherapy applications under de-identification conditions: benchmarking for data sharing and artificial intelligence use-cases 91%
- SPECT/CT imaging, biodistribution and radiation dosimetry of a 177Lu-DOTA-integrin αvβ6 cystine knot peptide in a pancreatic cancer xenograft model 90%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.