Machine learning approach using 18F-FDG-PET-radiomic features and the visibility of right ventricle 18F-FDG uptake for predicting clinical events in patients with cardiac sarcoidosis

OBJECTIVES

To investigate the usefulness of machine learning (ML) models using pretreatment 18F-FDG-PET-based radiomic features for predicting adverse clinical events (ACEs) in patients with cardiac sarcoidosis (CS).

MATERIALS AND METHODS

This retrospective study included 47 patients with CS who underwent 18F-FDG-PET/CT scan before treatment. The lesions were assigned to the training (n = 38) and testing (n = 9) cohorts. In total, 49 18F-FDG-PET-based radiomic features and the visibility of right ventricle 18F-FDG uptake were used to predict ACEs using seven different ML algorithms (namely, decision tree, random forest [RF], neural network, k-nearest neighbors, Naïve Bayes, logistic regression, and support vector machine [SVM]) with tenfold cross-validation and the synthetic minority over-sampling technique. The ML models were constructed using the top four features ranked by the decrease in Gini impurity. The AUCs and accuracies were used to compare predictive performances.

RESULTS

Patients who developed ACEs presented with a significantly higher surface area and gray level run length matrix run length non-uniformity (GLRLM_RLNU), and lower neighborhood gray-tone difference matrix_coarseness and sphericity than those without ACEs (each, p < 0.05). In the training cohort, all seven ML algorithms had a good classification performance with AUC values of > 0.80 (range: 0.841-0.944). In the testing cohort, the RF algorithm had the highest AUC and accuracy (88.9% [8/9]) with a similar classification performance between training and testing cohorts (AUC: 0.945 vs 0.889). GLRLM_RLNU was the most important feature of the modeling process of this RF algorithm.

CONCLUSION

ML analyses using 18F-FDG-PET-based radiomic features may be useful for predicting ACEs in patients with CS.

Incidence and adverse clinical events of primary sclerosing cholangitis with ulcerative colitis

PURPOSE

The aim of this study was to conduct a nationwide population-based study to estimate the incidence of primary sclerosing cholangitis in patients with ulcerative colitis (UC-PSC) and investigate healthcare use, medication use, surgery, cancer, and death as adverse clinical events of UC-PSC.

METHODS

We identified incident cases of UC with (UC-PSC) or without PSC (UC-alone) between 2008 and 2018 using health insurance claims data in Korea. Univariate (crude hazard ratio (HR)) and multivariate analyses were performed to compare the risk of adverse clinical events between groups.

RESULTS

A total of 14,406 patients with UC using population-based claims data were detected in the cohort. Overall, 3.38% (487/14,406) of patients developed UC-PSC. During a mean follow-up duration of approximately 5.92 years, the incidence of PSC in patients with UC was 185 per 100,000 person-years. The UC-PSC group showed statistically more frequent healthcare use (hospitalization and emergency department visits: HRs, 5.986 and 9.302, respectively; P < .001), higher immunomodulator and biologic use (azathioprine, infliximab, and adalimumab: HRs, 2.061, 3.457, and 3.170, respectively; P < .001), and higher surgery rate (operation for intestinal obstruction, and colectomy: HRs, 9.728 and 2.940, respectively; P < .001) than did the UC-alone group. The UC-PSC group also showed significantly higher colorectal cancer and biliary tract cancer (HRs, 2.799 and 36.343, respectively; P < .001) and mortality (HR, 4.257) rates than did the UC-alone group.

CONCLUSION

Patients with UC-PSC have higher risks of colorectal cancer, biliary tract cancer, and death than do patients with UC-alone. Although considered a rare disease, managing this complex and costly disease requires recognition of the impact of increased burden on healthcare services.

Worsening renal function, adverse clinical events and major determinants for changes of renal function in patients with atrial fibrillation: a Japanese multicenter registry substudy

Objective: To explore factors related to changing renal function and the prognostic effect of worsening renal function in patients with atrial fibrillation (AF). Methods: The present substudy was based on the SAKURA AF Registry, a Japanese multicenter prospective observational registry that includes 3267 AF patients from 63 institutions in the Tokyo area. Worsening renal function was defined as an estimated glomerular filtration rate (eGFR) decrease equaling more than 20% of the patient's baseline eGFR. Results: During a median 39.3 month follow-up period, patients' eGFR decreased annually by a mean value of 1.07 mL/min/1.73 m². Multivariable analysis showed that age ≥75 years, body weight ≤50 kg, a history of heart failure and initially preserved renal function (creatinine clearance [CrCl] ≥ 60 mL/min) were significantly associated with a decrease in eGFR, whereas a history of AF ablation was associated with a maintain in eGFR. The 194 patients with worsening renal function were at significantly increased risk of death, stroke and major bleeding (adjusted hazard ratios [HRs]: 2.06, 1.97 and 2.23, respectively). Conclusion: Age ≥75 years, body weight ≤50 kg, a history of heart failure and initially preserved renal function appear to promote renal dysfunction in patients with AF, but a history of AF ablation seems to have a favorable effect. Worsening renal function seems to increase AF patients' risk of adverse clinical events. Renal function can decline quickly; thus, early intervention including AF ablation is warranted.