Development of image-based decision support systems utilizing information extracted from radiological free-text report databases with text-based transformers

Article Radiology, Nuclear Medicine & Medical Imaging

Transformer-based structuring of free-text radiology report databases

S. Nowak et al.

Summary: This study investigates different labeling and pre-training strategies for on-site development of transformer-based structuring of free-text report databases. The results suggest that custom pre-training of transformers and fine-tuning on manual annotations can efficiently unlock report databases for data-driven medicine.

EUROPEAN RADIOLOGY (2023)

Add to Collection

Article Multidisciplinary Sciences

Deep learning supports the differentiation of alcoholic and other-than-alcoholic cirrhosis based on MRI

Julian A. Luetkens et al.

Summary: This proof-of-concept study explores the potential of deep learning (DL) to support imaging-based differentiation of the etiology of liver cirrhosis. The study shows that deep-learning classifiers have the potential to aid in discriminating liver cirrhosis etiology based on standard MRI.

SCIENTIFIC REPORTS (2022)

Add to Collection

Article Radiology, Nuclear Medicine & Medical Imaging

Detection of liver cirrhosis in standard T2-weighted MRI using deep transfer learning

Sebastian Nowak et al.

Summary: The study explored the use of deep transfer learning (DTL) in detecting liver cirrhosis from clinical MRI, showing promising diagnostic performance. The DTL method demonstrated higher accuracy in liver cirrhosis detection compared to radiologists, highlighting its potential as an effective tool in medical imaging diagnosis.

EUROPEAN RADIOLOGY (2021)

Add to Collection

Article Radiology, Nuclear Medicine & Medical Imaging

Deep-Learning-Based Diagnosis of Bedside Chest X-ray in Intensive Care and Emergency Medicine

Stefan M. Niehues et al.

Summary: This study developed and evaluated deep learning models for identifying clinically relevant abnormalities in bedside chest radiographs. The models showed similar performance to expert radiologists, suggesting its potential use in detecting findings during after-hours and assisting emergency and intensive care physicians in patient care.

INVESTIGATIVE RADIOLOGY (2021)

Add to Collection

Article Radiology, Nuclear Medicine & Medical Imaging