Modern machine and deep learning methods require large datasets to achieve reliable and robust results. This requirement is often difficult to meet in the medical field, due to data sharing limitations imposed by privacy regulations or the presence of a small number of patients (e.g., rare diseases). To address this data scarcity and to improve the situation, novel generative models such as Generative Adversarial Networks (GANs) have been widely used to generate synthetic data that mimic real data by representing features that reflect health-related information without reference to real patients. In this paper, we consider several GAN models to generate synthetic data used for training binary (malignant/benign) classifiers, and compare their performances in terms of classification accuracy with cases where only real data are considered. We aim to investigate how synthetic data can improve classification accuracy, especially when a small amount of data is available. To this end, we have developed and implemented an evaluation framework where binary classifiers are trained on extended datasets containing both real and synthetic data. The results show improved accuracy for classifiers trained with generated data from more advanced GAN models, even when limited amounts of original data are available.

Accurately estimating the length of stay (LOS) of patients admitted to the intensive care unit (ICU) in relation to their health status helps healthcare management allocate appropriate resources and better plan for the future. This paper presents predictive models for the LOS of ICU patients from the MIMIC-IV database based on typical demographic and administrative data, as well as early vital signs and laboratory measurements collected on the first day of ICU stay. The goal of this study was to demonstrate a practical, stepwise approach to predicting patient’s LOS in the ICU using machine learning and early available typical clinical data. The results show that this approach significantly improves the performance of models for predicting actual LOS in a pragmatic framework that includes only data with short stays predetermined by a prior classification.

The utilization of artificial intelligence (AI) in healthcare is on the rise, demanding increased accessibility to (public) medical data for benchmarking. The digitization of healthcare in recent years has facilitated medical data scientists’ access to extensive hospital data, fostering AI-based research. A notable addition to this trend is the Salzburg Intensive Care database (SICdb), made publicly available in early 2023. Covering over 27 thousand intensive care admissions at the University Hospital Salzburg from 2013 to 2021, this dataset presents a valuable resource for AI-driven investigations. This article explores the SICdb and conducts a comparative analysis with the widely recognized Medical Information Mart for Intensive Care - version IV (MIMIC-IV) database. The comparison focuses on key aspects, emphasizing the availability and granularity of data provided by the SICdb, particularly vital signs and laboratory measurements. The analysis demonstrates that the SICdb offers more detailed information with higher data availability and temporal resolution for signal data, especially for vital signs, compared to the MIMIC-IV. This is advantageous for longitudinal studies of patients’ health conditions in the intensive care unit. The SICdb provides a valuable resource for medical data scientists and researchers. The database offers comprehensive and diverse healthcare data in a European country, making it well suited for benchmarking and enhancing AI-based healthcare research. The importance of ongoing efforts to expand and make public datasets available for advancing AI applications in the healthcare domain is emphasized by the findings.

Defacing of brain magnetic resonance imaging (MRI) scans by removing identifiable facial features is essential for protecting patient privacy, yet assessing defacing quality remains challenging. While deep learning methods offer solutions, they require large labeled datasets, limiting their practical applicability. This study presents DefaceQA , a machine learning (ML) approach for automated defacing quality assessment using quantitative image features. A dataset of 200 MRI scans from the Leukodystrophy Registry at the Leipzig University Medical Center was processed using four defacing algorithms: PyDeface , QuickShear , FSL-Deface , and MRI-Deface . Image features extracted from original and defaced scans were used to classify defacing efficacy. The ML classifiers achieved an AUROC of 0.84 and an accuracy of 0.85 under a lenient criterion for successful/unsuccessful defacing, with the Feature Similarity Index Measure (FSIM) emerging as a key predictor. The findings demonstrate ML’s potential for defacing evaluation while highlighting challenges related to dataset limitations and generalizability.