Background Copeptin, the C-terminal fragment of provasopressin, has emerged as a potential prognostic biomarker in sepsis. However, its predictive accuracy for mortality in adult patients with sepsis remains uncertain. We conducted a systematic review and meta-analysis to evaluate the diagnostic performance of elevated blood copeptin levels for mortality prediction in this population. Methods We systematically searched PubMed, Embase, Web of Science, Wanfang Data, and CNKI from inception to 22 May 2025, for observational studies assessing copeptin levels at admission or within 48 h in adults with sepsis. Pooled sensitivity, specificity, likelihood ratios, diagnostic odds ratio (DOR), and area under the summary receiver operating characteristic curve (AUC) were calculated using a random-effects model. Study quality was assessed using QUADAS-2. Results Ten prospective studies involving 1,637 patients were included. Pooled sensitivity and specificity of elevated copeptin for predicting mortality were 0.77 (95% CI: 0.70–0.83; I 2 = 52%) and 0.76 (95% CI: 0.67–0.83; I 2 = 86%), respectively. The pooled positive and negative likelihood ratios were 3.16 (95% CI: 2.33–4.29) and 0.30 (95% CI: 0.23–0.40), with a DOR of 10.40 (95% CI: 6.62–16.33). The summary AUC was 0.83 (95% CI: 0.79–0.86), indicating good overall prognostic accuracy. Subgroup analysis according to the cutoffs of copeptin did not significantly affect the results. No significant publication bias was detected ( p = 0.58). Conclusion Elevated blood copeptin levels within 48 h of sepsis diagnosis show good prognostic accuracy for short-term mortality in adult patients with sepsis. These findings support the potential clinical utility of copeptin as a risk stratification tool in sepsis management. Systematic review registration https://www.crd.york.ac.uk/prospero/ , identifier CRD42024587540.

Stroke-associated pneumonia (SAP) is a frequent and severe complication following stroke. Recently, several machine learning (ML) models have been developed to predict SAP. We aimed to evaluate the predictive performance of these models in SAP prediction. We searched PubMed, Embase, Scopus, and Web of Science up to 18 June 2025, for studies developing ML, deep learning (DL), or neural network (NN) models for SAP prediction. The pooled estimates of area under the curve (AUC), accuracy (ACC), sensitivity (SEN), specificity (SPE), and diagnostic odds ratio (DOR) were calculated using the R program. A total of 27 studies were included, with a prevalence of SAP at 18.9%. Most models were ML based (77.8%), and clinical data were the most common input (77.8%). The pooled AUC was 0.84 [95% (CI): 0.80-0.87], and the pooled ACC was 0.80 (95% CI: 0.76-0.84). SEN and SPE were 0.73 (95% CI: 0.63-0.81) and 0.85 (95% CI: 0.77-0.90), respectively. The pooled DOR was 15.4 (95% CI: 10.2-23.3), and the summary receiver operating characteristic (SROC) curve showed an AUC of 0.853 with a false positive rate of 0.153 (95% CI: 0.096-0.235). No significant differences were found between ischemic and hemorrhagic subgroups. ML-based models demonstrated promising performance in predicting SAP and can help physicians through the early identification of high-risk cases. However, further external validation and integration into clinical workflows are required before widespread clinical adoption.

Diagnostic accuracy of prenatal ultrasound and MRI in predicting survival in severe isolated congenital diaphragmatic Hernia: A systematic review and meta-analysis.

Diagnostic Performance of Shear-Wave Dispersion Slope for Biopsy-Proven Hepatic Inflammation in MASLD: A Systematic Review and HSROC Meta-Analysis.

Diagnostic Performance of AI-Assisted Coronary CT Angiography: A Systematic Review and Meta-Analysis.

AI-driven ECG diagnostics: A game-changer for hypertrophic cardiomyopathy. A systematic review and diagnostic test accuracy meta-analysis.

Background: Neonatal jaundice is a common condition with potentially severe complications such as bilirubin-induced neurological dysfunction and kernicterus. While serum bilirubin (SBR) remains the standard laboratory measurement, point-of-care methods, such as transcutaneous bilirubinometry (TcB) and blood gas analysers (BGAs), offer rapid, less invasive alternatives. Direct comparisons of their diagnostic accuracy remain limited. Objective: The aim of this study was to assess and compare diagnostic accuracy and clinical utility of TcB and BGA against SBR in neonatal hyperbilirubinaemia screening. Methods: This retrospective study included neonates (n = 221) with concurrent SBR, BGA, and TcB measurements (n = 333). Assessment was via Passing–Bablok regression, Bland–Altman analysis, and Spearman correlation. Diagnostic performance was evaluated against jaundice thresholds in phototherapy charts (≥95th percentile threshold). Subgroup analyses considered phototherapy status, haemoglobin concentration, and Fitzpatrick skin type. Results: BGA showed stronger agreement with SBR (R2 = 0.88) than TcB (R2 = 0.43). BGA remained accurate regardless of phototherapy or haemoglobin levels. TcB accuracy declined post-phototherapy with reduced predictive value in darker-skinned neonates (Fitzpatrick III–VI) and increased false discovery rates. Both methods demonstrated low sensitivity (45.8%) but high specificity (>95%) and negative predictive value (~91%) for clinically significant hyperbilirubinaemia. BGA had a higher diagnostic odds ratio (47.5) than TcB (19.3). When individual patient sequential SBR and BGA measurements were compared for jaundice tracking (n = 175), there was high correlation, (r = 0.971) with no statistical differences, and 50% of measurements achieving agreement within 10 μmol/L. Conclusions: BGA is a more reliable alternative to SBR than TcB, particularly in time-critical or resource-limited settings. While TcB remains a non-invasive screening tool, limited accuracy post-phototherapy and with darker skinned neonates indicate confirmatory SBR testing. These findings support the selective and context-aware use of BGA and TcB to optimise neonatal hyperbilirubinaemia management and reduce interventions.

Diagnostic and prognostic significance of circulating HPV cfDNA in cervical cancer: A systematic review and meta-analysis.

Fecal DNA methylation of the syndecan-2(SDC2)gene is being explored as a noninvasive biomarker for colorectal cancer (CRC) detection. However, its diagnostic performance necessitates thorough evaluation. A systematic search of PubMed, Embase, and Web of Science was conducted to identify studies investigating fecal SDC2methylation (mSDC2) for CRC diagnosis. Eligible studies included adult CRC patients with histological confirmation and controls with either normal mucosa or benign colorectal lesions. Pooled sensitivity and specificity were synthesized using a Reitsma bivariate random-effects model, and summary receiver operating characteristic (SROC) curves with corresponding area under the curve (AUC) values were derived from this hierarchical model. Twenty-five studies encompassing 3,427 CRC patients, 3,267 individuals with benign lesions, and 5,372 with normal mucosa were included. For the comparison of CRC versus normal mucosa (24 studies), the pooled sensitivity and specificity were 0.86 (95% confidence interval [CI]: 0.82-0.89; I² = 88%) and 0.93 (95% CI: 0.90-0.95; I² = 95%), respectively. The pooled diagnostic odds ratio (DOR) was 81.73 (95% CI: 51.60-129.46), with an AUC of 0.95 (95% CI: 0.93-0.97). In the comparison against benign lesions (22 studies), the sensitivity was 0.85 (95% CI: 0.81-0.89; I² = 87%), specificity was 0.66 (95% CI: 0.59-0.71; I² = 91%), DOR was 11.10 (95% CI: 7.61-16.19), and AUC was 0.83 (95% CI: 0.80-0.86). Deeks' funnel plot asymmetry tests indicated no statistically significant publication bias (p= 0.48 and 0.54).In conclusion, fecal mSDC2testing demonstrates high diagnostic accuracy for CRC detection when compared to individuals with normal mucosa and moderate performance against benign colorectal lesions. These findings suggest that mSDC2may serve as a promising noninvasive biomarker to complement existing CRC screening methodologies.

Epilepsy surgery outcomes after intracranial EEG remain suboptimal necessitating the discovery of additional biomarkers to define the epileptogenic zone. Fast ripples (FRs) are a promising, new interictal epilepsy biomarker. By analyzing a multicenter data set consisting of overnight stereo-EEG (SEEG) recordings, we aimed at validating FRs as an accurate marker of the epileptogenic zone. We hypothesized that removing ≥60% of total FR events would significantly increase the odds of good postsurgical outcome (Engel class I). In addition, we compared FRs with spikes, and spikes co-occurring with FRs (spike-FRs) as surgery outcome predictors. This retrospective cohort study included consecutive patients from 4 epilepsy surgery centers in Canada, Finland, and Denmark, who underwent SEEG followed by resective surgery or a preplanned ablation procedure separate from the SEEG and had at least 1 year of follow-up. We detected FRs and spikes automatically from overnight SEEG recordings edited for artifacts. To calculate resection ratios of the detected events, we determined resected SEEG contacts by superimposing the peri-implantation and postresection images. We evaluated postsurgical seizure outcomes from medical records. Of the 73 included patients (mean age 23 ± 12 years, 41% female), 46 had good and 27 had poor (Engel classes II-IV) outcome at the latest follow-up. Patients with FR resection ratio ≥0.6 were more likely to achieve good postsurgical outcome (p < 0.001, diagnostic odds ratio [DOR] 10, 95% CI 2.7-39). Of those with ≥0.6 FR resection ratio, 26 of 29 (90%, 95% CI 74%-96%) achieved good outcome, whereas of those with <0.6 FR resection ratio, 24 of 44 (55%, 95% CI 46%-63%) had poor outcome, with overall accuracy of 68% (95% CI 57%-79%). In addition, the spike-FR resection ratio ≥0.6 was associated with good postsurgical outcome (p = 0.007, DOR 4.1, 95% CI 1.4-12, accuracy 64%, 95% CI 52%-75%), whereas the spike resection ratio ≥0.6 was not. In accordance with our hypothesis, the FR resection ratio ≥0.6 significantly increased the odds of attaining good postsurgical seizure outcome. Although the FR resection ratio ≥0.6 accurately predicted good postsurgical outcome, resecting <0.6 of FRs did not necessarily mean poor outcome. As predictors of postsurgical outcome, spikes fared poorly, whereas spike-FRs were comparable with FRs.

This systematic review and meta-analysis evaluated the diagnostic performance of artificial intelligence (AI) models that analyze preoperative prostate MRI images in conjunction with clinical parameters for predicting extraprostatic extension (EPE) in prostate cancer. A comprehensive search of PubMed, Embase, and Web of Science up to July 2025 identified 14 eligible studies involving 2,131 patients. The pooled analysis demonstrated that integrated radiomics-clinical models achieved high diagnostic performance, with a sensitivity of 0.83 (95% CI: 0.78-0.87), specificity of 0.82 (95% CI: 0.77-0.86), and an area under the curve (AUC) of 0.89 (95% CI: 0.86-0.92). The diagnostic odds ratio (DOR) was 19.82 (95% CI: 12.33-31.86), indicating robust discrimination between EPE-positive and EPE-negative cases. Subgroup analysis suggested models using deep learning algorithms had marginally higher accuracy (DOR: 24.6) than those using traditional machine learning (DOR: 17.3), though the difference was not statistically significant. Heterogeneity among studies stemmed from variations in MRI protocols, segmentation methods, and modeling approaches. No significant publication bias was detected. The results affirm that integrating radiomic features from multiparametric MRI (e.g., T2-weighted, diffusion-weighted imaging) with clinical variables (e.g., PSA, Gleason score) significantly outperforms conventional assessments for preoperative EPE prediction, demonstrating excellent diagnostic accuracy and supporting its potential clinical application in risk stratification. This supports the potential of combined models to enhance risk stratification and guide personalized surgical planning. Future research should prioritize standardized radiomics workflows, external validation, and multi-center collaborations to facilitate clinical adoption.

Blood culture is the diagnostic gold standard for bacteremia in the emergency department (ED), but its turnaround time can delay appropriate antimicrobial therapy, highlighting the need for rapid, accessible biomarkers. We retrospectively analyzed adult ED patients from July 2023 to June 2024 who underwent blood culture testing and had complete data for monocyte distribution width (MDW), white blood cell count (WBC), C-reactive protein (CRP), and neutrophil-to-lymphocyte ratio (NLR). Discrimination was assessed using area under the receiver operating characteristic curve (AUROC) and diagnostic accuracy using sensitivity, specificity, and diagnostic odds ratio (DOR); combined models were compared with net reclassification improvement (NRI) and integrated discrimination improvement (IDI). Among 19,325 patients, 2011 (10.4%) had positive blood cultures. MDW had the highest AUROC (0.760) versus CRP (0.730), NLR (0.695), and WBC (0.642); at a cut-off of 22, MDW showed 0.72 sensitivity, 0.68 specificity, and DOR 5.46. The best combined model was MDW+NLR (AUROC 0.785; DOR 6.39; NRI 0.428; IDI 0.770). MDW is a rapid and effective marker for identifying bacteremia in the ED, and performance improves when combined with NLR.

Oral potentially malignant disorders (OPMDs) can lead to oral cancer, which is one of the most common cancers worldwide. Prevention is crucial in the avoidance of malignant transformations of OPMDs. Artificial intelligence (AI) provides a new and noninvasive tool for analyzing medical data, such as patient data, radiologic images, and clinical photographs. These AI-based tools can help in the decision-making process. However, histological examination is still the gold standard for diagnosing OPMDs. This systematic review and meta-analysis aimed to investigate the diagnostic accuracy of artificial intelligence on intraoral photographs of patients with OPMDs. A systematic search was conducted on 5 major databases (MEDLINE, Embase, Cochrane Library, Scopus, and Web of Science) on November 10, 2023. Included studies compared AI methods to histology examination as the reference. A quantitative analysis was carried out to assess sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), diagnostic odds ratio (DOR), positive likelihood ratio (LR+), and negative likelihood ratio (LR-) calculated with 95% confidence intervals (CIs). Six eligible articles were included, with 898 images out of 4,046 tested using AI-based architectures. Five studies investigated at least 2 AI models. The overall sensitivity, specificity, DOR, LR+, and LR- were 0.94 (95% CI, 0.88 to 0.95), 0.95 (95% CI, 0.85 to 0.98), 212.39 (95% CI, 56.39 to 800.00), 16.89 (95% CI, 5.72 to 48.68), and 0.08 (95% CI, 0.05 to 0.13) for the best-performing AI-based architectures in terms of sensitivity, respectively. AI-based diagnostic tools have high negative predictive value that could help identify OPMD lesions using intraoral photographs.Knowledge Transfer Statement:This systematic review on AI-based methods to diagnose oral potentially malignant disorders showed that although their high negative predictive value could reduce unnecessary specialist consultations, clinical judgment remains paramount. Further prospective studies are needed to evaluate the integration of AI diagnostics into routine care and screening and policies to enhance efficiency and support early detection and prevention of oral cancer.

This review was done to evaluate diagnostic accuracy of LUS against established ARDS reference standards in adult populations. Four electronic databases were searched through June 2025 without language restrictions. Two reviewers independently screened studies, extracted data, and assessed quality using QUADAS‑2. Bivariate random‑effects models generated pooled sensitivity, specificity, likelihood ratios, diagnostic odds ratios (DORs), and hierarchical summary receiver‑operating characteristic (HSROC) curves. Subgroup analyses explored pattern‑ versus score‑based protocols and prospective study designs. Publication bias was assessed via Deeks' funnel‑plot test. Fourteen studies (531 ARDS-positive; 1354 ARDS-negative; pre‑test probability=28%) met inclusion criteria. Overall pooled sensitivity was 0.84 (95%CI: 0.69-0.92) and specificity 0.94 (95%CI: 0.83-0.98), with an AUROC of 0.95 (95%CI: 0.93-0.96). The positive likelihood ratio was 13.3 (95%CI: 5.0-35.9) and negative likelihood ratio 0.17 (95%CI: 0.09-0.34), corresponding to a DOR of 77 (95%CI: 26-227). Pattern‑based protocols achieved sensitivity 0.82 and specificity 0.96 (AUROC=0.96), while score‑based approaches yielded sensitivity 0.90 and specificity 0.83 (AUROC=0.93). Deeks' test indicated potential publication bias (p=0.004). LUS demonstrates excellent rule‑in and rule‑out performance for ARDS in critically ill adults, rivalling CT accuracy without its drawbacks. Adoption of standardized LUS protocols and integration into ARDS diagnostic pathways could enhance early detection, optimize management, and reduce reliance on ionizing imaging.

Background: Quantitative flow ratio (QFR) is a novel technology to assess the functional significance of coronary stenoses based on standard coronary angiography, which can be alternatives to invasive fractional flow reserve (FFR) assessment. However, the evidence is limited to single-center studies and small sample sizes. This study systematically determined the diagnostic performance of QFR to diagnose functionally significant stenosis with FFR as the reference standard. Methods: A systematic review and meta-analysis of studies assessing the diagnostic performance of angiography-derived QFR systems were performed. All relevant studies from six literature databases were searched and screened according to the inclusion and exclusion criteria. The pooled sensitivity, specificity, positive likelihood ratio (LR+), negative likelihood ratio (LR−), and diagnostic odds ratio (DOR), along with their 95% confidence intervals (CIs), were calculated using DerSimonian–Laird methodology. The summary receiver operating characteristic (SROC) curve and area under the curve were estimated. Meta-regression analysis was performed to identify a potential source of heterogeneity. Results: Fifty-seven studies comprising 13,215 patients and 16,125 vessels were included in the final analysis. At the vessel level, the pooled sensitivity and specificity of QFR for detecting a significant coronary stenosis were 0.826 (95% CI: 0.798–0.851) and 0.919 (95% CI: 0.902–0.933). Pooled LR+ and LR− were 10.198 (95% CI: 8.469–12.281) and 0.189 (95% CI: 0.163–0.219), with a pooled DOR of 53.968 (95% CI: 42.888–67.910). The SROC revealed an area under the curve (AUC) of 0.94 (95% CI: 0.91–0.96). The summary AUCs were 0.90 (95% CI: 0.87–0.92) for fixed-flow QFR (fQFR), 0.95 (95% CI: 0.92–0.96) for contrast-flow QFR (cQFR), 0.97 (95% CI: 0.95–0.98) for Murray law-based QFR (μQFR), and 0.91 (95% CI: 0.89–0.94) for non-specified QFR. The adjusted pooled DORs were as follows: 126.25 for μQFR, 45.49 for cQFR, 26.12 for adenosine-flow QFR (aQFR), 25.88 for fQFR, and 36.54 for non-specified QFR. Conclusions: The accuracy of angiography-derived QFR was strong to assess the functional significance of coronary stenoses with FFR as a reference. μQFR demonstrated the highest diagnostic performance among the five evaluated modes.

Tuberculosis (TB) remains a major global public health challenge, particularly in low-resource countries where access to trained radiologists is limited, making Chest X-ray (CXR) screening difficult to scale. The advancement of Artificial Intelligence (AI) and Computer-Aided Detection (CAD) technology offers a potential solution by providing automated TB detection and supporting diagnostic workflows. To assess their clinical readiness, this systematic review and meta-analysis was conducted using the PRISMA 2020 protocol and included studies from PubMed, Scopus, and Semantic Scholar that evaluated AI-CAD systems (Index Test) against microbiological or extended reference standards (Reference Standard). The Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2 and QUADAS-C) tools were applied to measure risk of bias, and a random-effects model was used to estimate pooled Diagnostic Odds Ratio (DOR). Six studies with approximately 38,940 participants were eligible for analysis. Results showed a pooled DOR of 0.133 (95% CI: 0.047–0.377), indicating a significantly lower diagnostic error rate (P=0.000). Although sensitivity was consistently high (83.3%–100%), specificity varied widely (26.8%–98.9%), resulting in notable heterogeneity and a wide prediction interval (0.003–6.411). These findings conclude that AI-CAD tools demonstrate strong potential for TB screening but should undergo local validation, threshold calibration, and operational evaluation before broad clinical implementation, especially where specificity remains below the WHO Target Product Profile.

Predictive Value of Cervical Length Measured After 24 Weeks for Spontaneous Preterm Birth: Systematic Review and Meta-Analysis.

Background: Mutations in the KRAS gene play a pivotal role in lung cancer development and progression and are becoming increasingly important in therapeutic decision-making. The detection of these mutations in circulating tumor DNA (ctDNA) has attracted attention as a minimally invasive diagnostic approach. However, the accuracy reported in different studies varies widely. Methods: We conducted a systematic review and meta-analysis in accordance with the PRISMA-DTA guidelines. Eligible studies evaluated the detection of KRAS mutations in ctDNA in plasma or serum for lung cancer diagnosis and reported sufficient data to construct 2 × 2 contingency tables. Primary pooled estimates of sensitivity, specificity and likelihood ratios were calculated using aggregated 2 × 2 contingency tables. Additionally, a bivariate random-effects model was applied in a secondary analysis to investigate between-study heterogeneity. Results: Nine diagnostic study arms comprising 691 patients met the inclusion criteria. Across all datasets, there were 255 true positives, 19 false positives, 136 false negatives, and 281 true negatives. The pooled sensitivity was 65.2%, while the pooled specificity was 93.7%. The positive likelihood ratio was 10.35, and the negative likelihood ratio was 0.37, resulting in a diagnostic odds ratio of 28.0, which indicates strong rule-in capability. Sensitivity showed moderate heterogeneity across studies. In contrast, specificity demonstrated minimal heterogeneity. Conclusions: ctDNA-based detection of KRAS mutations demonstrates high specificity but moderate sensitivity for diagnosing lung cancer. These findings suggest that a KRAS liquid biopsy could be a valuable complementary diagnostic tool, particularly when a tissue biopsy is not possible or is inadequate, and it could support more personalized decision-making as analytical technologies continue to advance.

Diabetic retinopathy (DR) persists as a predominant cause of preventable vision loss globally, with its prevalence escalating in conjunction with the diabetes epidemic. Efficient, automated screening is needed to enable earlier detection of DR at scale. Artificial intelligence (AI)-driven platforms, such as EyeArt® (Eyenuk Inc.), offer a scalable solution with potential to alleviate the burden on healthcare systems. A systematic review (SR) and meta-analysis were conducted following PRISMA and MOOSE guidelines. This review was prospectively registered in PROSPERO (CRD42024571137). Observational studies published between 2016 and 2024 assessing the diagnostic performance of the EyeArt® system for DR detection were retrieved from PubMed, Scopus, and Embase. Data on sensitivity, specificity, and diagnostic odds ratio (DOR) were extracted, and pooled estimates were calculated using a random-effects model. Study quality was assessed using QUADAS-2 and GRADE frameworks. Seventeen studies, met the inclusion criteria. The pooled log diagnostic odds ratio (LDOR) was 3.96 (95% CI 3.54-4.39), and the area under the summary receiver operating characteristic (SROC) curve was 0.932 (95% CI 0.885-0.985), indicating high overall diagnostic accuracy. No significant heterogeneity was observed in the pooled diagnostic OR, although sensitivity and specificity varied across studies. EyeArt® demonstrates high diagnostic accuracy for detecting any-grade and referable DR across diverse clinical and geographical settings. Its integration into DR screening programs could improve early detection, optimize healthcare resource allocation, and expand access to ophthalmic care, particularly in resource-limited environments. • EyeArt®demonstrated high diagnostic accuracy for detecting referable or any-grade DR across diverse settings. • Its consistent performance supports its integration into routine DR screening workflows. • Deployment of EyeArt®for DR may optimize resource allocation, streamline diagnostic pathways, and expand access, particularly in resource-limited environments.

Metabolic dysfunction-associated steatohepatitis (MASH) is an advanced form of metabolic dysfunction-associated steatotic liver disease (MASLD), characterised by hepatocellular injury, inflammation and varying degrees of fibrosis. Non-invasive, accurate diagnostic tools are critical for identifying patients "at risk" (MAS ≥4, F ≥2). This meta-analysis evaluates the diagnostic performance of imaging-based technologies for ruling in (TRI) and ruling out (TRO) "at risk" MASH. A systematic search of Medline and Embase (inception to December 20, 2024) identified studies reporting on MRI-based diagnostic techniques for "at risk" MASH. Eligible studies were independently screened, with 20 studies meeting inclusion criteria. Sensitivity, specificity, and diagnostic odds ratios (DORs) were calculated using bivariate meta-analysis, applying pre-specified TRO and TRI thresholds to each technique. Twenty studies involving 9,480 participants were included. FAST demonstrated highest TRO sensitivity (0.871) with moderate specificity (0.567) and TRI specificity (0.900) with reduced sensitivity (0.441). MEFIB achieved high TRO sensitivity (0.812) but lower specificity (0.606); TRI specificity was 0.872, sensitivity was 0.500. MAST exhibited intermediate performance, while cT1 thresholds showed variable diagnostic accuracy. A sensitivity analysis of head-to-head studies shows superior performance in FAST compared to other diagnostic methods. FAST with its accessibility and robust diagnostic performance may be well-suited for large-scale application. MRI-based techniques are effective non-invasive options for diagnosing "at risk" MASH in MASLD and may provide strong alternatives. Rather than challenging existing perspectives, this study provides a reflective overview of current evidence on imaging-based modalities for "at risk" MASH.