Noninvasive diagnostic criteria for hepatocellular carcinoma

Angio-Computed Tomograph-Guided Immediate Lipiodol Computed Tomograph for Diagnosis of Small Hepatocellular Carcinoma Lesions during Transarterial Chemoembolization

BACKGROUND

The diagnosis and treatment of small hepatocellular carcinoma (HCC) play a vital role in the prognosis of patients with HCC. The purpose of our study was to evaluate angio-computed tomography (angio-CT)-guided immediate lipiodol CT (a CT scan performed immediately after transarterial chemoembolization [TACE]) in the diagnosis of potential HCCs ≤1 cm in diameter.

METHODS

This study retrospectively analyzed 31 patients diagnosed with HCCs after routine imaging (contrast-enhanced CT or magnetic resonance imaging) or pathologic examinations with undefined or undetermined tumor lesions (diameter ≤1 cm) from February 2016 to September 2016. After TACE guided by digital subtraction angiography of the angio-CT system, potential HCC lesions with a diameter ≤1 cm were diagnosed by immediate lipiodol CT. The number of well-demarcated lesions was recorded to calculate the true positive rate. The correlation between the number of small HCCs detected by immediate lipiodol CT and the size of HCC lesions (diameter >1 cm) diagnosed preoperatively was analyzed 1 month after TACE. A paired t-test was used to analyze differences in liver function. Pearson analysis was used to analyze correlation. Chi-square test was used to compare the rates.

RESULTS

Fifty-eight lesions were detected on preoperative routine imaging examinations in 31 patients including 15 lesions with a diameter ≤1 cm. Ninety-one lesions were detected on immediate lipiodol CT, of which 48 had a diameter ≤1 cm. After 1 month, CT showed that 45 lesions had lipiodol deposition and three lesions had lipiodol clearance. Correlation analysis showed that the number of small HCCs detected by lipiodol CT was positively correlated with the size of HCC lesions diagnosed by conventional imaging examination (R2 = 0.54, P < 0.05).

CONCLUSION

Immediate lipiodol CT may be a useful tool in the diagnosis of potential HCC lesions with a diameter of ≤1 cm.

Novel Mahalanobis-based feature selection improves one-class classification of early hepatocellular carcinoma

Detection of early hepatocellular carcinoma (HCC) is responsible for increasing survival rates in up to 40%. One-class classifiers can be used for modeling early HCC in multidetector computed tomography (MDCT), but demand the specific knowledge pertaining to the set of features that best describes the target class. Although the literature outlines several features for characterizing liver lesions, it is unclear which is most relevant for describing early HCC. In this paper, we introduce an unconstrained GA feature selection algorithm based on a multi-objective Mahalanobis fitness function to improve the classification performance for early HCC. We compared our approach to a constrained Mahalanobis function and two other unconstrained functions using Welch's t-test and Gaussian Data Descriptors. The performance of each fitness function was evaluated by cross-validating a one-class SVM. The results show that the proposed multi-objective Mahalanobis fitness function is capable of significantly reducing data dimensionality (96.4%) and improving one-class classification of early HCC (0.84 AUC). Furthermore, the results provide strong evidence that intensity features extracted at the arterial to portal and arterial to equilibrium phases are important for classifying early HCC.

Incidence and risk factors for surveillance failure in patients with regular hepatocellular carcinoma surveillance

BACKGROUND

Initial presentation of hepatocellular carcinoma (HCC) at an advanced stage in patients under a regular surveillance program is a devastating problem.

AIMS

We assessed the prevalence and factors associated with this surveillance failure.

METHODS

A total of 304 HCC patients who received regular surveillance were retrospectively reviewed. Surveillance failure was defined when the tumor was diagnosed at beyond the Milan criteria.

RESULTS

Surveillance failure rate was 5.9 %. Macronodular cirrhosis (MC), ultrasonography-only surveillance (US-S) and infiltrative tumor type were independent factors associated with surveillance failure. The surveillance failure rate was higher in patients with MC (10.3 vs. 3.2 %, p = 0.022), US-S (14.6 vs. 4.3 %, p = 0.013) and when the tumor was infiltrative type (57.1 vs. 2.1 %, p < 0.001). Based on the two baseline factors (MC and US-S), the surveillance failure rates were 35.7, 6.8, 5.9 and 2.6 % for MC(+)/US-S(+), MC(+)/US-S(-), MC(-)/US-S(+) and MC(-)/US-S(-), respectively (p < 0.001).

CONCLUSION

The HCC surveillance failure was not rare in clinical practice. These data suggest that special attention for surveillance failure might be needed for patients with MC who receive US-S.