000285004 001__ 285004
000285004 005__ 20260220105948.0
000285004 0247_ $$2doi$$a10.1007/s00259-025-07478-7
000285004 0247_ $$2pmid$$apmid:40932612
000285004 0247_ $$2ISSN$$a1619-7070
000285004 0247_ $$2ISSN$$a1619-7089
000285004 037__ $$aDZNE-2026-00138
000285004 041__ $$aEnglish
000285004 082__ $$a610
000285004 1001_ $$aLi, Chenyang$$b0
000285004 245__ $$aEnhancing interpretability of AI with radiomics-based deep neural network: proof of concept in the classification of Parkinsonian syndromes with 18F-FDG PET imaging.
000285004 260__ $$aHeidelberg [u.a.]$$bSpringer-Verl.$$c2026
000285004 3367_ $$2DRIVER$$aarticle
000285004 3367_ $$2DataCite$$aOutput Types/Journal article
000285004 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article$$bjournal$$mjournal$$s1771495678_26619
000285004 3367_ $$2BibTeX$$aARTICLE
000285004 3367_ $$2ORCID$$aJOURNAL_ARTICLE
000285004 3367_ $$00$$2EndNote$$aJournal Article
000285004 520__ $$aInterpretability and reproducibility remain major challenges in applying deep neural network (DNN) to neuroimaging-based diagnosis. This study proposes a radiomics-guided dual-channel deep neural network (RDDNN) to improve feature transparency and enhance clinical understanding in the classification of Parkinsonian syndromes.In this bi-centric study, we analysed two independent cohorts comprising 1,275 patients with idiopathic Parkinson's disease (IPD), multiple system atrophy (MSA), and progressive supranuclear palsy (PSP), alongside 223 healthy controls from Huashan Hospital and 90 patients with IPD, MSA, and PSP (34IPD, 17MSA, 39PSP) from the University Hospital Munich. It is a re-analysis of well-studied Chinese and German cohorts of 18F-fluorodeoxyglucose Positron emission tomography (FDG-PET) imaging of parkinsonian patients and the FDG scans were of 10-min static acquisition at 60 min post FDG injection and normalized against whole brain activity. The RDDNN model combines local features extracted via dilated convolutional networks and global features derived from Transformer-based self-attention networks. Model performance was evaluated using classification metrics and compared to radiomics and DNN approaches. The model's outputs were also compared with nuclear medicine specialists' visual assessments to assess interpretability and time efficiency. Furthermore, SHapley Additive Explanations (SHAP), Layer-wise Class Activation Mapping (Layer-CAM), and Rollout Attention Map (RAM) were employed to evaluate which features played the most critical roles in the model's final classification decisions after supervised training, and to examine how both networks spatially corresponded to known brain connectivity regions.In the internal blind-test cohort, the RDDNN achieved high accuracy (AUC = 0.99, accuracy = 0.98). SHAP and correlation analyses jointly indicated complementary information across channels, some of which were clinically interpretable. In the external cohort, the model maintained robust performance (AUC = 0.94, accuracy = 0.81), with consistent feature patterns across populations. The model significantly reduced evaluation time compared to nuclear medicine specialists' readings (p < 0.001), and the heatmaps showed disease-specific activation in anatomically relevant regions for IPD, MSA, and PSP.The RDDNN framework provides a clinically interpretable and reproducible DNN solution for classifying Parkinsonian disorders. By integrating radiomics and attention-based modeling, it enhances lesion localization, supports clinical decision-making, and offers performance comparable to human specialists-while substantially improving diagnostic efficiency.
000285004 536__ $$0G:(DE-HGF)POF4-353$$a353 - Clinical and Health Care Research (POF4-353)$$cPOF4-353$$fPOF IV$$x0
000285004 536__ $$0G:(DE-HGF)POF4-352$$a352 - Disease Mechanisms (POF4-352)$$cPOF4-352$$fPOF IV$$x1
000285004 588__ $$aDataset connected to CrossRef, PubMed, , Journals: pub.dzne.de
000285004 650_7 $$2Other$$a18F-fluorodeoxyglucosePET imaging
000285004 650_7 $$2Other$$aDual-channel neural network
000285004 650_7 $$2Other$$aModel interpretability
000285004 650_7 $$2Other$$aParkinsonian syndromes
000285004 650_7 $$2Other$$aRadiomics
000285004 650_7 $$2Other$$aSHAP
000285004 650_7 $$00Z5B2CJX4D$$2NLM Chemicals$$aFluorodeoxyglucose F18
000285004 650_2 $$2MeSH$$aHumans
000285004 650_2 $$2MeSH$$aFluorodeoxyglucose F18
000285004 650_2 $$2MeSH$$aParkinsonian Disorders: diagnostic imaging
000285004 650_2 $$2MeSH$$aParkinsonian Disorders: classification
000285004 650_2 $$2MeSH$$aPositron-Emission Tomography
000285004 650_2 $$2MeSH$$aMale
000285004 650_2 $$2MeSH$$aFemale
000285004 650_2 $$2MeSH$$aNeural Networks, Computer
000285004 650_2 $$2MeSH$$aMiddle Aged
000285004 650_2 $$2MeSH$$aAged
000285004 650_2 $$2MeSH$$aProof of Concept Study
000285004 650_2 $$2MeSH$$aDeep Learning
000285004 650_2 $$2MeSH$$aImage Processing, Computer-Assisted: methods
000285004 650_2 $$2MeSH$$aRadiomics
000285004 7001_ $$aJiao, Fangyang$$b1
000285004 7001_ $$aWu, Shaoyou$$b2
000285004 7001_ $$aWang, Chenhan$$b3
000285004 7001_ $$aWei, Min$$b4
000285004 7001_ $$aZhang, Shuoyan$$b5
000285004 7001_ $$aWang, Luyao$$b6
000285004 7001_ $$aHuang, Yu$$b7
000285004 7001_ $$aYin, Yafu$$b8
000285004 7001_ $$aTian, Rong$$b9
000285004 7001_ $$0P:(DE-2719)9002620$$aBernhardt, Alexander$$b10$$udzne
000285004 7001_ $$0P:(DE-2719)9001160$$aKatzdobler, Sabrina$$b11$$udzne
000285004 7001_ $$0P:(DE-2719)2811659$$aLevin, Johannes$$b12$$udzne
000285004 7001_ $$0P:(DE-2719)2811373$$aHöglinger, Günter U$$b13$$udzne
000285004 7001_ $$0P:(DE-2719)9001539$$aBrendel, Matthias$$b14$$udzne
000285004 7001_ $$aRominger, Axel$$b15
000285004 7001_ $$aShi, Kuangyu$$b16
000285004 7001_ $$aZuo, Chuantao$$b17
000285004 7001_ $$00000-0003-4948-3683$$aJiang, Jiehui$$b18
000285004 773__ $$0PERI:(DE-600)2098375-X$$a10.1007/s00259-025-07478-7$$gVol. 53, no. 3, p. 1962 - 1979$$n3$$p1962 - 1979$$tEuropean journal of nuclear medicine and molecular imaging$$v53$$x1619-7070$$y2026
000285004 8564_ $$uhttps://pub.dzne.de/record/285004/files/DZNE-2026-00138%20SUP.docx
000285004 8564_ $$uhttps://pub.dzne.de/record/285004/files/DZNE-2026-00138.pdf$$yOpenAccess
000285004 8564_ $$uhttps://pub.dzne.de/record/285004/files/DZNE-2026-00138.pdf?subformat=pdfa$$xpdfa$$yOpenAccess
000285004 8564_ $$uhttps://pub.dzne.de/record/285004/files/DZNE-2026-00138%20SUP.doc
000285004 8564_ $$uhttps://pub.dzne.de/record/285004/files/DZNE-2026-00138%20SUP.odt
000285004 8564_ $$uhttps://pub.dzne.de/record/285004/files/DZNE-2026-00138%20SUP.pdf
000285004 909CO $$ooai:pub.dzne.de:285004$$popenaire$$popen_access$$pVDB$$pdriver$$pdnbdelivery
000285004 9101_ $$0I:(DE-588)1065079516$$6P:(DE-2719)9002620$$aDeutsches Zentrum für Neurodegenerative Erkrankungen$$b10$$kDZNE
000285004 9101_ $$0I:(DE-588)1065079516$$6P:(DE-2719)9001160$$aDeutsches Zentrum für Neurodegenerative Erkrankungen$$b11$$kDZNE
000285004 9101_ $$0I:(DE-588)1065079516$$6P:(DE-2719)2811659$$aDeutsches Zentrum für Neurodegenerative Erkrankungen$$b12$$kDZNE
000285004 9101_ $$0I:(DE-588)1065079516$$6P:(DE-2719)2811373$$aDeutsches Zentrum für Neurodegenerative Erkrankungen$$b13$$kDZNE
000285004 9101_ $$0I:(DE-588)1065079516$$6P:(DE-2719)9001539$$aDeutsches Zentrum für Neurodegenerative Erkrankungen$$b14$$kDZNE
000285004 9131_ $$0G:(DE-HGF)POF4-353$$1G:(DE-HGF)POF4-350$$2G:(DE-HGF)POF4-300$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$aDE-HGF$$bGesundheit$$lNeurodegenerative Diseases$$vClinical and Health Care Research$$x0
000285004 9131_ $$0G:(DE-HGF)POF4-352$$1G:(DE-HGF)POF4-350$$2G:(DE-HGF)POF4-300$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$aDE-HGF$$bGesundheit$$lNeurodegenerative Diseases$$vDisease Mechanisms$$x1
000285004 9141_ $$y2026
000285004 915__ $$0StatID:(DE-HGF)0200$$2StatID$$aDBCoverage$$bSCOPUS$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)0300$$2StatID$$aDBCoverage$$bMedline$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)1030$$2StatID$$aDBCoverage$$bCurrent Contents - Life Sciences$$d2025-11-07
000285004 915__ $$0LIC:(DE-HGF)CCBY4$$2HGFVOC$$aCreative Commons Attribution CC BY 4.0
000285004 915__ $$0StatID:(DE-HGF)0600$$2StatID$$aDBCoverage$$bEbsco Academic Search$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)0100$$2StatID$$aJCR$$bEUR J NUCL MED MOL I : 2022$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)0113$$2StatID$$aWoS$$bScience Citation Index Expanded$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)3002$$2StatID$$aDEAL Springer$$d2025-11-07$$wger
000285004 915__ $$0StatID:(DE-HGF)0150$$2StatID$$aDBCoverage$$bWeb of Science Core Collection$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
000285004 915__ $$0StatID:(DE-HGF)0030$$2StatID$$aPeer Review$$bASC$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)9905$$2StatID$$aIF >= 5$$bEUR J NUCL MED MOL I : 2022$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)0160$$2StatID$$aDBCoverage$$bEssential Science Indicators$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)1110$$2StatID$$aDBCoverage$$bCurrent Contents - Clinical Medicine$$d2025-11-07
000285004 915__ $$0StatID:(DE-HGF)0199$$2StatID$$aDBCoverage$$bClarivate Analytics Master Journal List$$d2025-11-07
000285004 9201_ $$0I:(DE-2719)1111015$$kClinical Research (Munich)$$lClinical Research (Munich)$$x0
000285004 9201_ $$0I:(DE-2719)1111016$$kAG Levin$$lClinical Neurodegeneration$$x1
000285004 9201_ $$0I:(DE-2719)1110007$$kAG Haass$$lMolecular Neurodegeneration$$x2
000285004 980__ $$ajournal
000285004 980__ $$aVDB
000285004 980__ $$aUNRESTRICTED
000285004 980__ $$aI:(DE-2719)1111015
000285004 980__ $$aI:(DE-2719)1111016
000285004 980__ $$aI:(DE-2719)1110007
000285004 9801_ $$aFullTexts