Deep learning-based algorithm for automatic quantification of nigrosome-1 and Parkinsonism classification using susceptibility map-weighted MRI

Pae Sun Suh; Hwan Heo; Chong Hyun Suh; Myeong Oh Lee; Soohwa Song; Dong Hoon Shin; Sung Yang Jo; Sun Ju Chung; Hwon Heo; Woo Hyun Shim; Ho Sung Kim; Sang Joon Kim; Eung Yeop Kim

doi:10.3174/ajnr.A8585

Deep learning-based algorithm for automatic quantification of nigrosome-1 and Parkinsonism classification using susceptibility map-weighted MRI

AJNR Am J Neuroradiol. 2024 Nov 15:ajnr.A8585. doi: 10.3174/ajnr.A8585. Online ahead of print.

Authors

Affiliation

¹ Department of Radiology and Research Institute of Radiology (P.S.S., C.H.S., H.H., W.H.S., H.S.K., S.J.K.), Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; Department of Radiology and Research Institute of Radiological Science and Center for Clinical Imaging Data Science(P.S.S.), Yonsei University College of Medicine, Seoul, South Korea; Heuron Co.Ltd. (H.H., M.O.L., S.H.S., D.H.S.) Seoul, Republic of Korea; Department of Neurology (S.J.C., S.Y.J.), Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; Department of Radiology (E.Y.K), Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea.

PMID: 39547802
DOI: 10.3174/ajnr.A8585

Abstract

Background and purpose: To develop and validate a deep learning-based automatic quantification for nigral hyperintensity and a classification algorithm for neurodegenerative parkinsonism using susceptibility map-weighted imaging (SMwI).

Materials and methods: We retrospectively collected 450 participants (210 with idiopathic Parkinson's disease [IPD] and 240 individuals in the control group) for training data between November 2022 and May 2023, and 237 participants (168 with IPD, 58 with essential tremor, and 11 with drug-induced Parkinsonism) for validation data between July 2021 and January 2022. SMwI data were reconstructed from multi-echo GRE. Diagnostic performance for diagnosing IPD was assessed using deep learning-based automatic quantification (Heuron NI) and classification (Heuron IPD) model. Reference standard for IPD was based on ¹⁸F-FP-CIT PET finding. Additionally, the correlation between the H&Y stage and volume of nigral hyperintensity in patients with IPD was assessed.

Results: Quantification of nigral hyperintensity using Heuron NI showed AUC of 0.915 (95% CI, 0.872-0.947) and 0.928 (95% CI, 0.887-0.957) on the left and right, respectively. Classification of nigral hyperintensity abnormality using Heuron IPD showed AUC of 0.967 (95% CI, 0.881-0.991) and 0.976 (95% CI, 0.948-0.992) on the left and right, respectively. H&Y score ≥ 3 showed significant smaller nigral hyperintensity volume (1.43 ± 1.19 mm³) compared to H&Y score 1-2.5 (1.98 ± 1.63 mm³; p = 0.008).

Conclusions: Our deep learning-based model proves rapid, accurate automatic quantification of nigral hyperintensity, facilitating IPD diagnosis, symptom severity prediction, and patient stratification for personalized therapy. Further study is warranted to validate the findings across various clinical settings.

Abbreviations: IPD = Idiopathic Parkinson's disease; SN = substantia nigra; SMwI = susceptibility map weighted imaging; QSM = quantitative susceptibility mapping; CNN = convolutional neural network; ICV = intracranial volume.