Key result
Deep learning adds modest diagnostic value to clinical ultrasound parameters for identifying at-risk MASLD.
Why the study?
The rising prevalence of MASLD underscores the need for tools to stratify individual risk of disease progression, prompting evaluation of deep learning-augmented B-mode ultrasound models to identify patients with increased fibrosis risk.
Do deep learning-based predictions from B-mode ultrasound improve the identification of increased fibrosis risk in patients with MASLD when added to clinical parameters?
Observational (n=205)
Do deep learning-based predictions from B-mode ultrasound improve the identification of increased fibrosis risk in patients with MASLD when added to clinical parameters?
Deep learning predictions from B-mode ultrasound add only modest incremental value to clinical parameters like FIB-4 for identifying increased fibrosis risk in MASLD.
Adds minimal value to clinical parameters for MASLD fibrosis risk; leaves open whether deep learning enhances ultrasound in prospective validation.
OBJECTIVE: The prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD) continues to rise, underscoring the need for tools to stratify individual risk of disease progression. We evaluated whether logistic regression models augmented by deep learning-based predictions (DLPs) can improve the B-mode ultrasound-based identification of at-risk MASLD, defined as patients with increased fibrosis risk. METHODS: We retrospectively analyzed 205 patients with a total of 636 ultrasound images. We developed a model that reproduces the LSM-based dichotomous fibrosis risk classification using clinical parameters and ultrasound image-derived deep learning pipelines. Patients were classified by same-day liver stiffness measurement (LSM) (<8 kPa: low fibrosis risk; ≥8 kPa: increased fibrosis risk). We assessed the incremental value of DLPs when added to the parameters sex, age, BMI, diabetes mellitus type 2 status and the fibrosis-4 score (FIB-4) based on accuracy, AUROC, and related statistics. RESULTS: The logistic regression model combining the clinical parameters and the DLPs achieved acceptable performance with an AUROC of 0.73 and a test accuracy of 68%. The same model without DLPs showed an AUROC of 0.72 and a test accuracy of 61%. Including FIB-4 improved performance further (AUROC 0.92, accuracy 88%). Models based solely on image data demonstrated limited diagnostic performance. CONCLUSION: B-mode ultrasound provides a weak fibrosis-related signal, yielding limited diagnostic performance. Meaningful discrimination required the incorporation of clinical parameters, with FIB-4 offering the greatest improvement among the parameters assessed. Deep learning predictions added only modest incremental value. Prospective validation is needed to clarify clinical utility.
No takes yet. Share an insight, caveat, or question.
Stansch et al. (2026) conducted an observational in Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD) (n=205). Logistic regression models augmented by deep learning-based predictions (DLPs) on B-mode ultrasound vs. Logistic regression models without DLPs (clinical parameters only) was evaluated on Identification of at-risk MASLD (increased fibrosis risk defined as liver stiffness measurement ≥8 kPa). Deep learning-based predictions added to clinical parameters for B-mode ultrasound identification of at-risk MASLD yielded modest incremental diagnostic value (AUROC 0.73 vs 0.72 without DLPs).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: