Artificial intelligence assists operators in real-time detection of focal liver lesions during ultrasound: A randomized controlled study.

PURPOSE: Detection of hepatocellular carcinoma (HCC) is crucial during surveillance by ultrasound. We previously developed an artificial intelligence (AI) system based on convolutional neural network for detection of focal liver lesions (FLLs) in ultrasound. The primary aim of this study was to evaluate whether the AI system can assist non-expert operators to detect FLLs in real-time, during ultrasound examinations. METHOD: This single-center prospective randomized controlled study evaluated the AI system in assisting non-expert and expert operators. Patients with and without FLLs were enrolled and had ultrasound performed twice, with and without AI assistance. McNemar's test was used to compare paired FLL detection rates and false positives between groups with and without AI assistance. RESULTS: 260 patients with 271 FLLs and 244 patients with 240 FLLs were enrolled into the groups of non-expert and expert operators, respectively. In non-experts, FLL detection rate in the AI assistance group was significantly higher than the no AI assistance group (36.9 % vs 21.4 %, p < 0.001). In experts, FLL detection rates were not significantly different between the groups with and without AI assistance (66.7 % vs 63.3 %, p = 0.32). False positive detection rates in the groups with and without AI assistance were not significantly different in both non-experts (14.2 % vs 9.2 %, p = 0.08) and experts (8.6 % vs 9.0 %, p = 0.85). CONCLUSIONS: The AI system resulted in significant increase in detection of FLLs during ultrasound examinations by non-experts. Our findings may support future use of the AI system in resource-limited settings where ultrasound examinations are performed by non-experts. The study protocol was registered under the Thai Clinical Trial Registry (TCTR20201230003), which is part of the WHO ICTRP Registry Network. The registry can be accessed via the following URL: https://trialsearch.who.int/Trial2.aspx?TrialID=TCTR20201230003.

Validation and prognostic value of EZ-ALBI score in patients with intermediate-stage hepatocellular carcinoma treated with trans-arterial chemoembolization.

BACKGROUND: Heterogeneity of liver function and tumor burden in intermediate-stage hepatocellular carcinoma (HCC) results in different outcomes after transarterial chemoembolization (TACE). Easy albumin-bilirubin (EZ-ALBI), a simplified albumin-bilirubin (ALBI) score, has recently been proposed as a new prognostic score for HCC. This study aimed to validate the EZ-ALBI score and evaluate the impact of dynamic changes in patients with intermediate-stage HCC undergoing TACE. METHODS: All patients with HCC treated with TACE at King Chulalongkorn Memorial Hospital, Bangkok, Thailand, between January 2015 and December 2019 were prospectively enrolled. Intermediate-stage HCC was defined as Barcelona Clinic Liver Cancer (BCLC) stage B or unresectable single HCC with size > 5 cm in BCLC stage A. EZ-ALBI and ALBI scores were calculated and stratified into three different grades. Overall survival (OS) and prognostic factors were assessed using the Kaplan-Meier curve and Cox proportional hazard model. Decision analysis curves were used to evaluate the clinical utility of the predictive scores. RESULTS: Among 672 patients with HCC treated with TACE, 166 patients with intermediate-stage HCC who met the eligibility criteria were enrolled. The median OS of all patients in the cohort was 21 months. A good correlation between the EZ-ALBI and ALBI scores was observed (correlation coefficient 1.000, p < 0.001). The baseline EZ-ALBI grades 1, 2, and 3 were 24.5%, 70%, and 5.5%, respectively. EZ-ALBI grade can stratify patients with significantly different prognoses (p = 0.002). Baseline EZ-ALBI grade 2, 3, and serum alpha-fetoprotein > 20 ng/ml were significantly associated with OS [hazard ratio (HR) 2.20 (95% confidence interval [CI] 1.24-3.88, p = 0.007), 3.26 (95% CI 1.24-8.57, p = 0.016), and 1.77 (95% CI 1.10-2.84, p = 0.018), respectively]. Following TACE, 42 (29.6%) patients had a worsening EZ-ALBI grade. However, the EZ-ALBI grade migration was not significantly correlated with OS. EZ-ALBI and ALBI score provided improved discriminatory ability (Harrell's concordance index 0.599 and 0.602, respectively) and better net benefit compared with Child-Turcotte-Pugh and Model for End-stage Liver Disease scores. CONCLUSIONS: The baseline EZ-ALBI score demonstrated good predictive performance for survival and a strong correlation with conventional ALBI scores. Both the EZ-ALBI and ALBI scores outperformed other prognostic models in patients with intermediate-stage HCC receiving TACE. However, the dynamic change in the EZ-ALBI grade after TACE was not associated with postprocedural survival.

Genetic screening to avoid adverse drug reactions from medication use and approach patients' better outcomes: A lesson learn from the report of the Queen Savang Vadhana Memorial Hospital.

Background and Aims: The polymerase chain reaction (PCR) technique is adopted for pharmacogenetic testing and adverse drug reaction (ADR) analysis. Methods: PCR was used for testing of pharmacogenetic markers for HLA and non-HLA polymorphism related to specific drugs. Results: Among 76 cases that underwent genetic screening, 7.7%, 11.1%, and 2.7% of the patients were found to be genetically positive for allopurinol, carbamazepine, and abacavir, respectively. Two cases were genetically positive for interferon, and two cases of extensive metabolizers were positive for clopidogrel. One case of a NAT2 slow acetylator for isoniazid was found. Among the 74 cases with complete outcomes, 39.2% showed improvements and 18.9% reported a deterioration. Although no serious ADR was observed, two HLA-B*5701-negative cases reported ADRs (2.7%). All patients positive for IL28B were improved. One patient receiving clopidogrel showed improvements, but another showed deterioration. Finally, the outcome of slow acetylation NAT2 was worse without ADR. Conclusion: PCR-based pharmacogenetic testing is critical for ADR monitoring in a cost-effective manner.

Development and validation of artificial intelligence to detect and diagnose liver lesions from ultrasound images.

Artificial intelligence (AI) using a convolutional neural network (CNN) has demonstrated promising performance in radiological analysis. We aimed to develop and validate a CNN for the detection and diagnosis of focal liver lesions (FLLs) from ultrasonography (USG) still images. The CNN was developed with a supervised training method using 40,397 retrospectively collected images from 3,487 patients, including 20,432 FLLs (hepatocellular carcinomas (HCCs), cysts, hemangiomas, focal fatty sparing, and focal fatty infiltration). AI performance was evaluated using an internal test set of 6,191 images with 845 FLLs, then externally validated using 18,922 images with 1,195 FLLs from two additional hospitals. The internal evaluation yielded an overall detection rate, diagnostic sensitivity and specificity of 87.0% (95%CI: 84.3-89.6), 83.9% (95%CI: 80.3-87.4), and 97.1% (95%CI: 96.5-97.7), respectively. The CNN also performed consistently well on external validation cohorts, with a detection rate, diagnostic sensitivity and specificity of 75.0% (95%CI: 71.7-78.3), 84.9% (95%CI: 81.6-88.2), and 97.1% (95%CI: 96.5-97.6), respectively. For diagnosis of HCC, the CNN yielded sensitivity, specificity, and negative predictive value (NPV) of 73.6% (95%CI: 64.3-82.8), 97.8% (95%CI: 96.7-98.9), and 96.5% (95%CI: 95.0-97.9) on the internal test set; and 81.5% (95%CI: 74.2-88.8), 94.4% (95%CI: 92.8-96.0), and 97.4% (95%CI: 96.2-98.5) on the external validation set, respectively. CNN detected and diagnosed common FLLs in USG images with excellent specificity and NPV for HCC. Further development of an AI system for real-time detection and characterization of FLLs in USG is warranted.

Surveillance for Hepatocellular Carcinoma Reduces Mortality: an Inverse Probability of Treatment Weighted Analysis

ABSTRACT Background. Evidence supporting benefit of hepatocellular carcinoma (HCC) surveillance in reducing mortality is not well-established. The effect of HCC surveillance in reducing mortality was assessed by an inverse probability of treatment weighting (IPTW)- based analysis controlled for inherent bias and confounders in observational studies. Material and methods. This retrospective cohort study was conducted on 446 patients diagnosed with HCC between 2007 and 2013 at a major referral center. Surveillance was defined as having at least 1 ultrasound test within a year before HCC diagnosis. Primary outcome was survival estimated using the Kaplan-Meier method with lead-time bias adjustment and compared using the log-rank test. Hazard ratio (HR) and 95% confidence interval (Cl) were computed using conventional Cox and weighted Cox proportional hazards analysis with IPTW adjustment. Results. Of the 446 patients, 103 (23.1%) were diagnosed with HCC through surveillance. The surveillance group had more patients with the Barcelona-Clinic Liver Cancer stage A (80.6% vs. 33.8%, P < 0.0001), more patients eligible for potentially curative treatment (73.8% vs. 44.9%, P < 0.0001), and longer median survival (49.6 vs. 15.9 months, P < 0.0001). By conventional multivariate Cox analysis, HR (95% Cl) of surveillance was 0.63 (0.45-0.87), P = 0.005. The estimated effect of surveillance remained similar in the IPTW-adjusted Cox analysis (HR: 0.57; 95% Cl: 0.43-0.76, P < 0.001). Conclusions. HCC surveillance by ultrasound is associated with a 37% reduction in mortality. Even though surveillance is recommended in all guidelines, but in practice, it is underutilized. Interventions are needed to increase surveillance rate for improving HCC outcome.

Surveillance for Hepatocellular Carcinoma Reduces Mortality: an Inverse Probability of Treatment Weighted Analysis.

BACKGROUND: Evidence supporting benefit of hepatocellular carcinoma (HCC) surveillance in reducing mortality is not well-established. The effect of HCC surveillance in reducing mortality was assessed by an inverse probability of treatment weighting (IPTW)-based analysis controlled for inherent bias and confounders in observational studies. MATERIAL AND METHODS: This retrospective cohort study was conducted on 446 patients diagnosed with HCC between 2007 and 2013 at a major referral center. Surveillance was defined as having at least 1 ultrasound test within a year before HCC diagnosis. Primary outcome was survival estimated using the Kaplan-Meier method with lead-time bias adjustment and compared using the log-rank test. Hazard ratio (HR) and 95% confidence interval (CI) were computed using conventional Cox and weighted Cox proportional hazards analysis with IPTW adjustment. RESULTS: Of the 446 patients, 103 (23.1%) were diagnosed with HCC through surveillance. The surveillance group had more patients with the Barcelona-Clinic Liver Cancer stage A (80.6% vs. 33.8%, P < 0.0001), more patients eligible for potentially curative treatment (73.8% vs. 44.9%, P < 0.0001), and longer median survival (49.6 vs. 15.9 months, P < 0.0001). By conventional multivariate Cox analysis, HR (95% CI) of surveillance was 0.63 (0.45-0.87), P = 0.005. The estimated effect of surveillance remained similar in the IPTW-adjusted Cox analysis (HR: 0.57; 95% CI: 0.43-0.76, P < 0.001). CONCLUSIONS: HCC surveillance by ultrasound is associated with a 37% reduction in mortality. Even though surveillance is recommended in all guidelines, but in practice, it is underutilized. Interventions are needed to increase surveillance rate for improving HCC outcome.