Texture Analysis Improves Accuracy of Computer-assisted Differentiation between Small Hepatic Cysts and Hepatocellular Carcinoma on Noncontrast CT
Bibliographic record
Abstract
Background Small (< 1.5cm in diameter) liver lesions are commonly detected incidentally on non-contrast CT exams. Characterization of these lesions as cystic or solid is crucial to patient management but is frequently not possible without other imaging modalities. Evaluation This study explored whether texture analysis could increase diagnostic confidence for the differentiation of simple Hepatic Cysts (SC) and Hepatocellular Carcinoma (HCC) from normal liver parenchyma (NP) in non-contrast CT exams. Patients (42) were retrospectively selected, 21 with ultrasound confirmed SC and 21 with positive pathology for HCC. Images were resampled to 0.75mm per pixel. Circular ROIs were drawn to fit the smallest lesion and copied nearby to overlie NP. A total of 84 ROIs were defined (21 SC + 21 HCC + 42 NP). Intensity metrics (IM) were measured, and texture features extracted, using either an S-Transform (ST) or a Wavelet Transform (WT) for each ROI. A Support Vector Machine Classifier was used to solve the classification problem. A leave-one-out cross-validation was used to estimate classification accuracy. Using IM only the accuracy was 89%. The statistical power to detect a 'biomarker of tissue type' was found to be equal to 0.98 at a 95% confidence level. Using IM and either ST or WT increased accuracy to 94%. Using IM and both ST and WT increased the accuracy to 95%. SCs were identified with 100% sensitivity (SN) and 100% specificity (SP). Differentiating HCC from NP had SN=77% (SP=94%) using IM only. Using IM and ST, the SN was 83% (SP=98%). Using IM and WT, the SN was 90% (SP=95%). Using IM and both WT and ST, the SN was 87% (SP=98%). Discussion The accuracy of distinguishing SC from NP or HCC tissue is improved by adding transform based texture features to pixel intensity analysis. Due to the small number of cases not correctly classified on the basis of intensity, this texture-enabled additional accuracy did not reach statistical significance. CONCLUSION Image intensity combined with texture analysis provides a robust tool that could aid in the discrimination between SC, HCC and NP for small liver lesions on non-contrast CT, avoiding the expense and inconvenience of additional imaging.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.009 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".