Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-2077
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-21T06:54:21.875062Z
status: published
lens: g_space
sector: health
headline: Super Resolution Reconstruction: A Strategy to Improve the Diagnostic Performance of Radiomics for Predicting Microvascular Invasion in Hepatocellular Carcinoma
dek: Rationale and objectives Microvascular invasion (MVI) is a key factor associated with high postoperative recurrence rate in patients with hepatocellular carcinoma (HCC). This study aims to evaluate the value of super-resolution (SR) reconstruction based on hepatobiliary phase (HBP) MR images for improving the accuracy of MVI prediction. Materials and methods A total of 544 patients with HCC from two medical centers between 2018 and 2025 were retrospectively analyzed, including 450 patients in the training set, 5…
gain_title: Super-resolution reconstruction of hepatobiliary phase MR images raised radiomics model AUC for predicting microvascular invasion in hepatocellular carcinoma, including in an external test set.
problem_title: (none)
trace_subject: (none)
gain_reading: Super-resolution reconstruction of hepatobiliary phase MR images raised radiomics model AUC for predicting microvascular invasion in hepatocellular carcinoma, including in an external test set.
gain_evidence: The corresponding AUCs of the radiomics model based on the SR images were 0.855, 0.798, and 0.827, respectively. | SR reconstruction can improve the diagnostic performance of radiomics models for predicting MVI in patients with HCC.
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers analyzed 544 patients with hepatocellular carcinoma from two medical centers, treated between 2018 and 2025. They used a lesion-aware 3D Restormer transformer to create super-resolution versions of hepatobiliary phase MR images, then extracted radiomics features from both original and enhanced scans to predict microvascular invasion.

Microvascular invasion is linked to high recurrence after surgery, so better prediction from imaging could matter for patient management. The reported gains were largest in the external test set, but that set was small and the study was retrospective, so wider confirmation is still needed.
limitation: The analysis was retrospective and the external test set contained only 43 patients, which limits how firmly the improvement can be generalized.
tag: Evidence-backed gain
key_points: The cohort had 450 training, 51 internal validation, and 43 external test patients. | Tumor segmentation used nnU-Net, then manual correction by an experienced abdominal radiologist. | 851 radiomics features were extracted from each image type, and LASSO selected 50. | Logistic Regression, Random Forest, and LightGBM were used for model development. | In the external test set, AUC was 0.827 with super-resolution images versus 0.614 with original images.
rundown: The study tested whether improving image resolution before feature extraction could strengthen radiomics prediction of microvascular invasion, a factor associated with high postoperative recurrence in hepatocellular carcinoma.

Models built on original images reached pooled AUCs of 0.710, 0.764, and 0.614 in training, validation, and test sets. Models built on super-resolution images reached 0.855, 0.798, and 0.827 in the same sets.

The pipeline combined automated nnU-Net tumor segmentation, radiologist correction, and feature selection before model building, with data drawn from two centers.
sources:
- peer_reviewed | Academic Radiology | https://doi.org/10.1016/j.acra.2026.09.076 | 2026-10-19
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

4887acb1839a13d709a5f21450939a5e06885fd41e0b5b9b2decbd2ba753bddc

previous

0000000000000000000000000000000000000000000000000000000000000000