Arterial spin labeling-based Z-maps have high specificity and positive predictive value for neurodegenerative dementia compared to FDG-PET.

OBJECTIVE: Cerebral perfusion analysis based on arterial spin labeling (ASL) MRI has been proposed as an alternative to FDG-PET in patients with neurodegenerative disease. Z-maps show normal distribution values relating an image to a database of controls. They are routinely used for FDG-PET to demonstrate disease-specific patterns of hypometabolism at the individual level. This study aimed to compare the performance of Z-maps based on ASL to FDG-PET. METHODS: Data were combined from two separate sites, each cohort consisting of patients with Alzheimer's disease (n = 18 + 7), frontotemporal dementia (n = 12 + 8) and controls (n = 9 + 29). Subjects underwent pseudocontinuous ASL and FDG-PET. Z-maps were created for each subject and modality. Four experienced physicians visually assessed the 166 Z-maps in random order, blinded to modality and diagnosis. RESULTS: Discrimination of patients versus controls using ASL-based Z-maps yielded high specificity (84%) and positive predictive value (80%), but significantly lower sensitivity compared to FDG-PET-based Z-maps (53% vs. 96%, p < 0.001). Among true-positive cases, correct diagnoses were made in 76% (ASL) and 84% (FDG-PET) (p = 0.168). CONCLUSION: ASL-based Z-maps can be used for visual assessment of neurodegenerative dementia with high specificity and positive predictive value, but with inferior sensitivity compared to FDG-PET. KEY POINTS: • ASL-based Z-maps yielded high specificity and positive predictive value in neurodegenerative dementia. • ASL-based Z-maps had significantly lower sensitivity compared to FDG-PET-based Z-maps. • FDG-PET might be reserved for ASL-negative cases where clinical suspicion persists. • Findings were similar at two study sites.

Clinical applicability of artificial intelligence for patients with an inherited heart disease: A scoping review.

The number of inherited heart disease (IHD) studies using artificial intelligence (AI) has increased rapidly over the last years. In this scoping review, we aimed to present an overview of the current literature available on the applicability of AI within the field of IHD. The literature search resulted in eighteen articles in which an AI model was trained and tested, mostly for diagnostic and predictive purposes. The areas under the receiver operating characteristic curves ranged from 0.78-0.96, but varied between IHD types, used methods and outcome measures. Only three out of eighteen did perform validation on an external dataset and most studies did not use explainable deep learning models. To be able to integrate AI as a tool to aid physicians in their diagnoses and clinical decisions within the IHD field, generalizability has to be better evaluated and explainability of DL models has to be increased.