Overview

Schema Version

https://atlas.rsna.org/schemas/2025-11/dataset.json

Name

Progressive Supranuclear Palsy (PSP) Dataset

Link

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11140505/

Indexing

Keywords: MR Imaging, Brain/Brain Stem, Segmentation, Quantification, Diagnosis, Convolutional Neural Network, MR parkinsonism index, MRPI 2.0, Midbrain to pons area ratio, Planimetry

Content: MR, NR

RadLex: RID35976, RID10312, RID6677

SNOMED: 49049000, 192976002

Author(s)

Salvatore Nigro

Marco Filardi

Benedetta Tafuri

Martina Nicolardi

Roberto De Blasi

Alessia Giugno

Valentina Gnoni

Giammarco Milella

Daniele Urso

Stefano Zoccolella

Giancarlo Logroscino

Organization(s)

Center for Neurodegenerative Diseases and the Aging Brain, University of Bari Aldo Moro, Pia Fondazione Cardinale G. Panico, Tricase, Italy

Department of Translational Biomedicine and Neuroscience (DiBraiN), University of Bari Aldo Moro, Bari, Italy

Department of Radiology, Pia Fondazione Cardinale G. Panico, Tricase, Italy

Department of Neurosciences, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, England

Operative Unit of Neurology, San Paolo Hospital, ASL Bari, Bari, Italy

Funding

Supported by the Regione Puglia and CNR for Tecnopolo per la Medicina di Precisione, DGR number 2117 of November 21, 2018 (CUPB84I18000540002), and the Research Center of Excellence for Neurodegenerative Diseases and Brain Aging (CIREMIC), University of Bari, Aldo Moro.

Ethical review

All individuals provided informed consent, and the protocol was approved by the institutional review board at all sites.

Comments

Retrospective study developing and validating DL-based planimetric segmentation and measurement of brainstem and ventricular structures on T1-weighted MRI, with internal, external, and clinical test datasets.

Date

Published: 2024-03-20

References

[1] Nigro S, Filardi M, Tafuri B, Nicolardi M, De Blasi R, Giugno A, Gnoni V, Milella G, Urso D, Zoccolella S, Logroscino G. "Deep Learning–based Approach for Brainstem and Ventricular MR Planimetry: Application in Patients with Progressive Supranuclear Palsy". Radiology: Artificial Intelligence. 2024-05-01. doi:10.1148/ryai.230151. PMID: 38506619. PMCID: PMC11140505.

Dataset

Motivation

To develop a fast and fully automated DL-based method for planimetric segmentation and measurement of brainstem and ventricular structures to assist diagnosis of PSP.

Sampling

Retrospective selection of T1-weighted MRI from multiple cohorts and centers; controls for training and testing; age- and sex-matched PD and PSP for clinical testing.

Partitioning scheme

Training on 84 controls; Internal test 35 controls (1 excluded from 36 due to artifacts); External tests 35 FTLDNI controls and 35 ADNI controls; Clinical test 200 patients (71 PSP, 129 PD).

Relationships between instances

Clinical test set includes two diagnostic groups (PSP and PD) used for comparative analysis and classification.

Noise

One internal test image excluded due to signal intensity artifacts; automated pipeline failures (3%) due to incorrect midsagittal section identification.

External data

Data were obtained from PPMI, 4RTNI, FTLDNI, ADNI, and the Center for Neurodegenerative Diseases and the Aging Brain (University of Bari).

Confidentiality

Retrospective analysis of health data with informed consent and IRB approval across contributing sites.

Sensitive data

Neuroimaging health data involving neurodegenerative disease diagnoses.