Interactive Steatosis Mapper: liver CT attenuation reading with estimated MRI-PDFF and histological grade.
Created by Dr. Sharad Maheshwari MD - imagingsimplified@gmail.com
Scientific audit & evidence calibration applied (Sept 2026 update).
Founder: BeResponsibleAI
Deterministic CT Steatosis & MASH Triage
This tool uses a two-step progressive triage architecture. In Step 1, quantify hepatic parenchymal attenuation from any unenhanced 120 kV CT to obtain an approximate imaging-based steatosis category and CT-derived MRI-PDFF equivalent (based on the Pickhardt et al. linear regression: PDFF % = -0.58 × HU + 38.2). In Step 2 (optional), explore surrogate research risk signals for active steatohepatitis (MASH) using transaminases and lipid subfractions.
Clinical & Technical Boundaries: This tool outputs imaging-based categorical approximations, not direct histologic predictions or measured MRI-PDFF ground truth. CT attenuation is influenced by biological confounders (hemochromatosis/iron overload elevates HU; acute hepatitis/edema lowers HU; amiodarone elevates HU) and acquisition parameters (contrast phase, tube voltage, iterative reconstruction). CT has marked loss of precision at low fat fractions (<5% PDFF / >57 HU).
Step 1: Unenhanced CT Liver Attenuation
Non-contrast 120 kV exam (opportunistic chest/abdomen or dedicated abdominal CT)
CT-Derived Physical Equivalent Estimates
Approximate imaging categories; not interchangeable with MRI-PDFF or biopsyStep 2: Add Clinical Biomarkers
Optional Research SignalExplore surrogate biomarkers of hepatocellular damage (ALT) and atherogenic insulin resistance (HDL-C) within the MASLD spectrum
Metabolic & Transaminase Profile
Leave blank for opportunistic pure-imaging triage, or input values if available
Zhang et al. (2026) Model-Derived Relative Log-Odds
Source & Validation Limitations: Derived from single-center bariatric surgery candidates ($n=285$, mean BMI 38.1, pretest MASH 51.6%, age 13–64). Lacks external validation in general screening cohorts. Intercept and calibration slopes were not reported. This represents a relative research signal, not an individual clinical probability or diagnostic decree.
Elevated Biomarker Risk Pattern (Exploratory Heuristic Signal)
Liver attenuation ≤ 40 HU combined with abnormal transaminases/lipids matches the exploratory composite pattern described in bariatric cohorts. Reminder: MASH cannot be diagnosed non-invasively by CT and labs alone.
⚙ Educational Clinical Synthesis (Step 1 + Step 2)
Adjust the sliders to generate a clinical synthesis.
Clinical Responsibility & Governance Notice:
The draft text below separates Objective Measurement, Interpretation Certainty, and Suggested Clinical Correlation. Radiologists must independently confirm true non-contrast 120 kV acquisition, consider non-steatotic attenuation confounders (iron overload, acute edema/hepatitis, amiodarone), and modify drafts to suit the clinical indication.
📄 Structured Dictation Template (3-Part Architecture)
Live OutputModerate Steatosis (40 HU)
Clinical Rationale & Diagnostic Nuance
Context will appear here based on selected parameters.
Evidence & Scientific Rationale
This dashboard bridges physics, population radiology cohorts, and clinical biomarkers. Below is the evidentiary breakdown and explicit boundaries of each metric.
Unenhanced 120 kV CT vs. MRI-PDFF Regression Curve
PDFF = -0.58 × HU + 38.2Derived from Pickhardt et al. (Radiology/RadioGraphics cohorts; $n=221$, 120-kV unenhanced $n=72$ within 1 month: $R^2 = 0.828$)
Demonstrates the linear relationship between unenhanced 120 kV CT radiodensity and confounder-corrected MRI-PDFF. Note that performance significantly degrades outside this narrow domain: $R^2 = 0.554$ for non-120 kV acquisitions and $R^2 = 0.565$ when imaging exams are separated by $>1$ month. Furthermore, CT precision collapses at low fat fractions (<5% PDFF / >57 HU).
The "Three Rulers" Paradox: Why Histology, PDFF, and CT Cannot Be 1:1 Translated
A central hazard in non-invasive steatosis triage is treating categorical grades (Mild, Moderate, Severe) as fixed biological entities. In reality, clinical medicine uses three fundamentally different measuring instruments that assess related but distinct physical phenomena:
Measures the fraction of hepatocytes containing macrovesicular lipid droplets (<5%, 5–33%, 34–66%, >66%). It is a two-dimensional cellular count, not a physical volume or mass fraction.
Measures the percentage of mobile protons attributable to triglyceride fatty acid chains across a 3D liver volume. Because lipid droplets occupy space differently than cellular counts, median PDFF for severe steatosis (S3) is only ~25%, not >66%!
Measures linear photon attenuation. An indirect surrogate correlated to PDFF via linear regression ($\text{PDFF} = -0.58 \times \text{HU} + 38.2$), which is in turn correlated to histology. Sensitive to iron, edema, beam energy, and reconstruction kernel.
| Study Cohort / Calibration | S0 → S1 (Mild) | S1 → S2 (Moderate) | S2 → S3 (Severe) | Reported / Derivation Nuances |
|---|---|---|---|---|
| Tang et al. (Radiology 2013) | 6.4% | 17.4% | 22.1% | Derivation cohort diagnostic cutoffs |
| Biopsy Correlation Cohort (High Specificity) | 5.75% | 15.5% | 21.35% | Median PDFF: S0 2.3% | S1 7.8% | S2 19.4% | S3 25.4% |
| Dioguardi Burgio et al. (2019) | 6.5% | 16.5% | 22.0% | Validation cohort cutoffs |
| Pickhardt et al. Regression ($\text{PDFF} = -0.58\text{HU} + 38.2$) | ≈ 57.2 HU (5.0%) | ≈ 40.0 – 42.0 HU (14–15%) | ≤ 28.0 – 30.0 HU (≥21–22%) | $R^2 = 0.828$ at 120 kV; $\le 48\text{ HU}$ has 100% spec for $\ge 30\%$ fat |
Takeaway for Dictation: Describing findings in terms of the underlying physical measurement (e.g., "Liver attenuation of 38 HU, consistent with moderate steatosis category [expected MRI-PDFF ~16% based on Pickhardt regression]") is scientifically more defensible than asserting an unequivocal biopsy grade.
Methodological Scoping: The Zhang et al. (2026) MASH Model
To maintain academic transparency, the following technical details and validation boundaries of the composite MASH model are disclosed:
The "Three Rulers" Paradox: Why Histology, PDFF, and CT Cannot Be 1:1 Translated
A central hazard in non-invasive steatosis triage is treating categorical grades (Mild, Moderate, Severe) as fixed biological entities. In reality, clinical medicine uses three fundamentally different measuring instruments that assess related but distinct physical phenomena:
Measures the fraction of hepatocytes containing macrovesicular lipid droplets (<5%, 5–33%, 34–66%, >66%). It is a two-dimensional cellular count, not a physical volume or mass fraction.
Measures the percentage of mobile protons attributable to triglyceride fatty acid chains across a 3D liver volume. Because lipid droplets occupy space differently than cellular counts, median PDFF for severe steatosis (S3) is only ~25%, not >66%!
Measures linear photon attenuation. An indirect surrogate correlated to PDFF via regression (~40 HU ≈ 15–17% PDFF), which is in turn correlated to histology. Sensitive to iron, edema, and beam energy.
| Study Cohort | S0 → S1 (Mild) | S1 → S2 (Moderate) | S2 → S3 (Severe) | Reported / Observed Values |
|---|---|---|---|---|
| Tang et al. (Radiology 2013) | 6.4% | 17.4% | 22.1% | Derivation cohort diagnostic cutoffs |
| Biopsy Correlation Cohort (High Specificity) | 5.75% | 15.5% | 21.35% | Median PDFF: S0 2.3% | S1 7.8% | S2 19.4% | S3 25.4% |
| Dioguardi Burgio et al. (2019) | 6.5% | 16.5% | 22.0% | Validation cohort cutoffs |
| CT Attenuation Equivalent (120 kV) | ≈ 48.6 – 57.2 HU | ≈ 40.0 – 48.6 HU | < 30.0 – 40.0 HU | Boyce 2010 mean: 58.8 ± 10.8 HU |
Takeaway for Dictation: When a report cites "S2 Moderate Steatosis" from a CT scan, it is applying a continuous regression model to an imperfect categorical boundary. Describing findings in terms of the underlying physical measurement (e.g., "Liver attenuation of 38 HU, consistent with moderate steatosis [expected MRI-PDFF 15–24%]") is scientifically more accurate than asserting an unequivocal biopsy grade.
1 Triage Architecture: The CT "Gatekeeper"
The tool separates macroscopic structural fat detection from inflammatory biomarker evaluation:
A. The Gatekeeper: CT Attenuation (≤ 40 HU)
Unenhanced liver attenuation ≤ 40 HU is a verified population threshold for moderate-to-severe macrovesicular steatosis (≥30% histology fat; Boyce 2010; Pickhardt 2024; Haghshomar 2024). In our logic, lab biomarkers only trigger a composite MASH pattern alert if structural fat is confirmed by this gatekeeper. Without CT fat, elevated transaminases warrant investigation for viral, autoimmune, or toxic etiologies.
B. The Composite Biomarker Heuristic (ALT & HDL-C)
This feature is an author-derived clinical heuristic inspired by Zhang et al. (BMC Gastroenterology 2026). In a retrospective study of 321 bariatric candidates with obesity (mean BMI 38.1), a multivariable logistic model incorporating continuous CT attenuation, ALT, and HDL-C yielded an AUC of 0.799 for identifying histological MASH. Important caveat: Our tool uses a dichotomized heuristic rather than the published continuous nomogram; it has not been externally validated in unselected low-prevalence screening populations.
2 Physical Principle & Biological Distribution
- ✔ Normal liver parenchyma is roughly 55 to 65 HU. Macroscopic intracellular triglycerides have negative attenuation (approx. -50 to -100 HU). As lipid droplets accumulate within hepatocytes, average voxel density decreases linearly.
- ✔ Gaussian Biological Variation: In the landmark Boyce et al. (AJR 2010) cohort of 3,357 asymptomatic adults, mean liver attenuation was 58.8 ± 10.8 HU (range -14 to 91 HU). An SD of ~11 HU implies that normal individuals naturally fall below 55 HU without disease. Therefore, attenuation between 50 and 55 HU is best categorized as a biological transition zone rather than a definitive pathologic diagnosis.
3 Biomarker Selection: Mechanisms & Evidence
Why ALT and HDL-C?
ALT (Alanine Transaminase)
The premier enzyme marker of hepatocellular membrane damage and ballooning in steatohepatitis. Notably, contemporary hepatology guidelines (AASLD / EASL 2024; Ann Hepatol 2024) recommend lower thresholds (~30 U/L for men, ~19 U/L for women) because traditional limits (>40 U/L) were calibrated on cohorts with unrecognized subclinical steatosis.
HDL-C (High-Density Lipoprotein Cholesterol)
Surrogate for hepatic insulin resistance and lipotoxicity. In progressive MASLD, hepatic insulin resistance upregulates Cholesteryl Ester Transfer Protein (CETP) activity and downregulates ApoA-I synthesis, accelerating HDL clearance and driving dyslipidemia (Borggreve et al., J Lipid Res 2003; Rashid et al., Circulation 2003). Standard sex-specific dyslipidemia thresholds are <1.0 mmol/L (40 mg/dL) for men and <1.3 mmol/L (50 mg/dL) for women.
📖 Primary Peer-Reviewed References
1. Quantitative CT to MRI-PDFF Regression Model ($n=221$)
Pickhardt PJ, et al. Assessment of Hepatic Steatosis: Comparison of Unenhanced CT and Dual-Echo Chemical-Shift MRI / RadioGraphics 2024. Derives the validated linear conversion $\text{MRI-PDFF (\%)} = -0.58 \times \text{HU} + 38.2$ ($R^2 = 0.828$ for unenhanced 120-kV CT within 1 month). Shows marked degradation ($R^2 = 0.554$) with alternate kVp and contrast.
View Article (RadioGraphics 2024) →2. Unenhanced CT Liver Attenuation Distribution (n=3,357)
Boyce CJ, Pickhardt PJ, Kim DH, et al. Hepatic Steatosis (Fatty Liver Disease) in Asymptomatic Adults Identified by Unenhanced Low-Dose CT. AJR Am J Roentgenol. 2010;194(3):623-628. Demonstrates mean attenuation 58.8 ± 10.8 HU and solidifies the ≤40 HU threshold for ≥30% steatosis.
View Article (DOI: 10.2214/AJR.09.2590) →3. CT Steatosis Diagnostic Meta-Analysis (2024)
Haghshomar M, Antonacci D, Smith AD, et al. Diagnostic Accuracy of CT for the Detection of Hepatic Steatosis: A Systematic Review and Meta-Analysis. Radiology. 2024;313(2):e241171. Highlights diagnostic accuracy of unenhanced CT and contemporary migration toward >17% MRI-PDFF for moderate steatosis.
View Article (DOI: 10.1148/radiol.241171) →4. Living Donor Transplant CT-to-Histology Assessment
Park SH, Kim PN, Kim KW, et al. Macrovesicular Hepatic Steatosis in Living Liver Donors: Use of CT for Quantitative and Qualitative Assessment. Radiology. 2006;239(1):105-112. (PMID: 16484355). Shows unenhanced CT cutoff of 48 HU has 100% specificity and 54% sensitivity for ≥30% histologic steatosis.
View Article (PMID: 16484355) →5. CT-Biomarker Composite MASH Model (Published May 2026)
Zhang H, Jiang Z, Zheng H, Wang H, Xia L, Chen J, Zhang B. Liver CT-based composite biomarkers can identify MASH and steatosis grade in people with obesity prior to bariatric surgery: a retrospective study. BMC Gastroenterology. 2026;26:419. (PMID: 42135651, DOI: 10.1186/s12876-026-04914-2).
• CT_Liver: OR 0.944 (95% CI: 0.924–0.964, p < 0.001) → β = -0.05763 / HU
• ALT: OR 1.011 (95% CI: 1.002–1.019, p = 0.013) → β = +0.01094 / (U/L)
• HDL-C: OR 0.209 (95% CI: 0.060–0.722, p = 0.013) → β = -1.56543 / (mmol/L)
Demographics: Age 13–64 (mean 31.3 ± 9.3 yrs; includes adolescents); pre-test MASH prevalence 51.6%.
6. Lipid Mechanism: CETP & HDL in Hepatic Insulin Resistance (Verified Primary Sources)
Borggreve SE, De Vries R, Dullaart RPF. Alterations in high-density lipoprotein metabolism and reverse cholesterol transport in insulin resistance and type 2 diabetes mellitus: role of lipolytic enzymes, lecithin:cholesterol acyltransferase and lipid transfer proteins. Eur J Clin Invest. 2003;33(12):1051-1069. (DOI: 10.1111/j.1365-2362.2003.01263.x).
Jiang ZG, Robson SC, Yao Z. Lipoprotein metabolism in nonalcoholic fatty liver disease. J Biomed Res. 2013;27(1):1-13. (DOI: 10.7555/JBR.27.20120077); Rashid S, et al. J Diabetes Complications. 2002;16(1):24-28.
Comprehensive Liver Disease Knowledge Base
A clinical guide to understanding the spectrum of Steatotic Liver Disease, contemporary nomenclature, and non-invasive measurement nuances.
1. The Nomenclature Shift: NAFLD to MASLD
In 2023, international hepatology societies (AASLD, EASL, ALEH) transitioned from NAFLD to MASLD (Metabolic Dysfunction-Associated Steatotic Liver Disease).
Clinical Rationale:
- Eliminates exclusionary and stigmatizing language ("non-alcoholic", "fatty").
- Establishes an affirmative diagnostic framework: steatosis plus at least one cardiometabolic risk criterion (BMI ≥ 25, type 2 diabetes / pre-diabetes, hypertension, hypertriglyceridemia, or low HDL-C).
2. Diagnostic Metrics & Threshold Nuances
Liver fat is quantified across three primary modalities: Radiodensity (CT), Proton Fraction (MRI), and Histology (Biopsy). Each operates on different physical units:
CT Attenuation (Hounsfield Units / HU)
Measures linear X-ray attenuation. Water is 0 HU; pure adipose is −50 to −100 HU. Normal parenchyma spans a broad Gaussian distribution (58.8 ± 10.8 HU in Boyce 2010). Granular CT cohorts establish: Normal ≥ 57.2 HU, Mild steatosis 48.6 to <57.2 HU, Mild-to-moderate 40 to <48.6 HU, and Moderate-to-severe ≤ 40 HU.
MRI-PDFF (Proton Density Fat Fraction)
The non-invasive reference standard. Accurately quantifies the percentage of mobile protons belonging to triglyceride molecules. Crucial distinction: Because PDFF measures proton volume/mass across the whole organ, its numerical values are much lower than histology cell counts. Severe steatosis (S3) has a median PDFF of only ~25.4% (Tang 2013), NOT >66%.
Histological Steatosis Grade (NASH-CRN / Kleiner S0 - S3)
Biopsy assessment of the fraction of liver cells containing visible lipid vacuoles: S0 (<5%), S1 (5–33%), S2 (34–66%), and S3 (>66%). This is a 2D cellular count, explaining the numerical gap with 3D volumetric PDFF.
Three Rulers Calibration Matrix: Note that histologic cell-count percentages (>33%, >66%) do not equal MRI-PDFF percentages. Categorical cutoffs are approximate central tendencies derived from cross-modality regression models.
| Biopsy Grade (Kleiner) | Hepatocytes w/ Fat (Cell Count) | Est. MRI-PDFF (Proton Vol. %) | CT Attenuation (120 kV) | Clinical Triage Category |
|---|---|---|---|---|
| S0 (Normal) | < 5% cells | < 5% – 6.4% (Median ~2.3%) | ≥ 55 – 57 HU (Mean ~59) | Normal Range |
| S1 (Mild) | 5% – 33% cells | 6.4% – 16% (Median ~7.8%) | 41 – 50 HU | Mild Steatosis |
| S2 (Moderate) | 34% – 66% cells | 16% – 22% (>17% consensus; Median ~19.4%) | 30 – 40 HU | Clinically Important |
| S3 (Severe) | > 66% cells | ≥ 22% – 25% (Median ~25.4%) | < 30 HU | Severe Steatosis |
3. Disease Progression & Fibrosis Screening
While this tool screens for steatosis and inflammatory risk (MASH), liver-related mortality is primarily governed by hepatic fibrosis. Every patient flagged with moderate-to-severe steatosis should be risk-stratified for fibrosis using non-invasive scores:
FIB-4 Index (Primary Non-Invasive Triage)
Calculated as: (Age × AST) / (Platelets [×10&sup9;/L] × √ALT).
- < 1.30 (< 2.0 if age ≥ 65): High negative predictive value (>90%) to rule out advanced fibrosis (F3–F4). Primary care follow-up.
- 1.30 – 2.67: Indeterminate risk. Second-line testing recommended (Transient Elastography / VCTE or ELF test).
- > 2.67: High probability of advanced fibrosis / cirrhosis. Hepatology referral recommended.
4. Biomarker Mechanics & Clinical Significance: ALT & HDL-C
How ALT and HDL-C interact with non-contrast CT liver attenuation, why they were chosen, and how outputs vary.
4.1 The Core Architectural Principle: The "Gatekeeper" Model
In this dashboard, liver attenuation in Hounsfield Units (HU) and biochemical laboratories (ALT / HDL-C) serve two fundamentally different diagnostic roles:
| Metric | Role in Dashboard | Physical / Biological Target |
|---|---|---|
| CT Attenuation (HU) | The Gatekeeper (Structure) | Quantifies macroscopic physical triglyceride accumulation across liver parenchyma (S0–S3, estimated MRI-PDFF). |
| ALT & HDL-C | The Modifier (Activity & Risk) | Gauges active cytolytic inflammation (ALT) and systemic metabolic insulin resistance (HDL-C). |
Key Rule: Labs Do NOT Alter Physical Steatosis Grade
- Moving the ALT or HDL-C sliders never changes the estimated Biopsy Grade (S0–S3) or the estimated MRI-PDFF range.
- Fat volume on CT is an objective physical measurement of X-ray beam attenuation.
- Labs determine whether that structural fat is quiescent (MASL) or associated with an elevated risk of active steatohepatitis (MASH).
4.2 The Two Threshold Gates
[ Non-Contrast 120 kV CT ]
│
Is Liver HU ≤ 40 HU?
│
┌───────────────┴───────────────┐
YES NO (> 40 HU)
│ │
[ GATE OPEN: Steatosis ≥ S2 ] [ GATE CLOSED: Normal / Mild ]
│ │
Are Labs Provided? Are Labs Provided?
┌──────┴──────┐ ┌──────┴──────┐
YES NO YES NO
│ │ │ │
Any Abnormal? CT-Only Any Abnormal? CT-Only
┌────┴────┐ Report ┌─────┴─────┐ Report
YES NO YES NO
│ │ │ │
COMPOSITE STRUCTURAL INVESTIGATE NORMAL
MASH ALERT FAT ONLY NON-STEATOTIC ROUTINE
(High Risk) (Labs Normal) ETIOLOGY
≤ 40 HU is the established population threshold for moderate-to-severe steatosis (≥30% macrovesicular fat on biopsy; Boyce 2010, Pickhardt 2024). If attenuation is > 40 HU, the structural fat "gate" is closed. Even if ALT is 120 U/L, the tool will not trigger a composite MASH alert because the imaging prerequisite is absent. Instead, it prompts evaluation for non-steatotic liver injury (viral hepatitis, medications, alcohol, autoimmunity).
If HU ≤ 40 and either laboratory value is abnormal:
- ALT > ULN: Female > 19 U/L, Male > 30 U/L, Unisex > 40 U/L
- HDL-C < Cutoff: Female < 1.29 mmol/L (50 mg/dL), Male < 1.03 mmol/L (40 mg/dL), Unisex < 1.10 mmol/L (42.5 mg/dL)
Transitions the dashboard to the Elevated Composite Risk Pattern.
4.3 Biological & Clinical Significance of Each Marker
A. ALT (Alanine Transaminase): The Hepatocellular Injury Marker
- Biological Mechanism: ALT is concentrated inside hepatocyte cytoplasm. In simple steatosis (MASL), lipid droplets sit passively within cells. When lipotoxicity triggers cellular ballooning, endoplasmic reticulum stress, and membrane necrosis, ALT leaks into the bloodstream.
- Clinical Role: Differentiates benign macroscopic fat accumulation from active, cytolytic steatohepatitis (MASH).
- Threshold Nuance: Historical reference limits (40 U/L) were calibrated on reference populations that inadvertently included individuals with subclinical fatty liver. Contemporary guidelines (AASLD / EASL 2024; Ann Hepatol 2024) set the true normal limit to ~19 U/L for women and ~30 U/L for men. Using sex-stratified thresholds prevents under-triaging women with early MASH.
B. HDL-C (High-Density Lipoprotein Cholesterol): The Metabolic Axis
- Biological Mechanism: Progressive liver disease is intimately tied to severe hepatic insulin resistance. Insulin resistance upregulates Cholesteryl Ester Transfer Protein (CETP) activity and blunts hepatic synthesis of Apolipoprotein A-I (ApoA-I). CETP transfers triglycerides into HDL in exchange for cholesteryl esters, creating small, dense HDL particles that undergo accelerated renal clearance (Borggreve et al., J Lipid Res 2003; Rashid et al., Circulation 2003).
- Clinical Role: Low HDL-C acts as a direct biochemical surrogate for the atherogenic dyslipidemia and hepatic insulin resistance driving fibrosis progression.
- Standalone Warning: In Zhang et al. (2026), HDL-C alone had an AUC of only 0.613 (poor standalone discriminator). It is not a liver-specific enzyme; it achieves diagnostic utility only when combined multivariably with hepatic attenuation and transaminases.
4.4 Input Permutations: Can You Use One, Both, or Neither?
The dashboard is explicitly programmed to handle all four clinical data scenarios gracefully:
| Scenario | What Happens in the Dashboard |
|---|---|
| 1. CT Only (No Labs) | Pure imaging triage; prompts for metabolic baseline labs. |
| 2. CT + ALT Only | Triage based on structural fat + cytolytic hepatocellular injury. |
| 3. CT + HDL-C Only | Triage based on structural fat + metabolic atherogenic resistance. |
| 4. CT + Both (ALT & HDL-C) | Complete composite evaluation across both activity and metabolic axes. |
- How: Leave both lab boxes blank (or click "Clear Labs").
- Behavior: Displays pure CT structural classification (e.g., 35 HU → S2 Moderate Steatosis, Est. PDFF 16–22%). Risk card reads "Clinically Important". Composite badge: "No Labs Provided (CT Only)".
- Report: Advises primary care to order ALT, lipid panel, and FIB-4.
- Utility: Opportunistic screening on unenhanced chest/trauma CTs before labs are available.
- Behavior: If HU ≤ 40 and ALT is elevated, triggers the Elevated Composite Pattern based on documented hepatocellular injury. If normal, displays "Labs Normal".
- Report: Cites ALT level and notes that lipid subfractions remain uncharacterized.
- Utility: Outpatient setting where hepatic transaminases were checked but fasting lipids were deferred.
- Behavior: If HU ≤ 40 and HDL-C is below cutoff, triggers Elevated Composite Pattern driven by atherogenic dyslipidemia.
- Report: Details macrovesicular fat alongside dyslipidemia; recommends transaminases to assess hepatocellular necrosis.
- Utility: Correlating opportunistic CT findings against a routine annual wellness lipid panel before liver enzymes have been ordered.
- Behavior: If either or both are abnormal (ALT > ULN OR HDL-C < cutoff), triggers full Elevated Composite Pattern. If both normal, badge confirms "Labs Normal".
- Report: Synthesizes both cytolytic injury and dyslipidemia into a coherent MASH risk statement.
- Utility: Comprehensive evaluation for patients with full metabolic and hepatic panels.
4.5 Summary Matrix: How the Live Output Varies
Assuming a scan demonstrating moderate hepatic steatosis (38 HU):
| Labs Entered | Values | Risk Status | Dynamic Report Output Summary |
|---|---|---|---|
| None | Blank / Blank | Clinically Important | "CT confirms moderate steatosis (≤ 40 HU). Clinical correlation with transaminases, lipid panel, and FIB-4 is advised." |
| ALT Only | 54 U/L / Blank | Elevated Pattern | "CT confirms moderate steatosis (≤ 40 HU) with elevated ALT (54 U/L), satisfying composite heuristic criteria for active MASH risk." |
| ALT Only | 18 U/L / Blank | Clinically Important | "CT confirms moderate steatosis (≤ 40 HU). Normal ALT (18 U/L) argues against severe hepatocellular necrosis, but metabolic monitoring is advised." |
| HDL-C Only | Blank / 0.8 mmol/L | Elevated Pattern | "CT confirms moderate steatosis (≤ 40 HU) with severe dyslipidemia (0.8 mmol/L), reflecting high cardiometabolic risk." |
| Both | ALT 48, HDL 0.9 | Elevated Pattern | "CT confirms moderate steatosis (≤ 40 HU). Co-occurring transaminase elevation and depressed HDL-C trigger the composite MASH heuristic alert." |
| Both | ALT 22, HDL 1.4 | Clinically Important | "CT confirms moderate steatosis (≤ 40 HU). Reported labs are currently within normal limits. Fibrosis risk stratification (FIB-4) suggested." |
💡 Key Takeaway for Clinicians
CT alone tells you: "Does the patient have fat, and how much?" (Physical organ structure & density)
Adding ALT tells you: "Is the liver actively inflamed or injured?" (Cytolytic membrane leakage & ballooning)
Adding HDL-C tells you: "Is there systemic metabolic insulin resistance driving the disease?" (Atherogenic dyslipidemia & CETP activity)
You do not need both labs to get clinical value from the dashboard; each lab independently enriches the structural CT finding with functional, actionable risk information.
5. What about LAI (Liver Attenuation Index)?
Historically, radiologists frequently evaluated steatosis using the Liver-Spleen difference or ratio, collectively known as the Liver Attenuation Index (LAI).
The Logic: A normal liver is typically 5 to 10 HU brighter than the spleen. A Liver-Spleen difference of ≤ -10 HU strongly correlates with ≥ S2 steatosis and MRI-PDFF.
Why use absolute HU (≤ 40) in this tool instead of LAI?
- Field of View: In opportunistic screening (particularly chest CTs), the spleen is frequently not fully visible.
- Splenic Disease: Anemia, portal hypertension, or splenic congestion can alter spleen attenuation, introducing errors.
- Global Standardization: Massive cohort studies prove an absolute liver attenuation of ≤ 40 HU is robust enough to diagnose moderate-to-severe fat without an internal splenic control.
Comments
Post a Comment