Parkinson's Disease Recognition Using SPECT Image and Interpretable AI: A Tutorial

被引:18
作者
Pianpanit, Theerasarn [1 ]
Lolak, Sermkiat [2 ]
Sawangjai, Phattarapong [3 ]
Sudhawiyangkul, Thapanun [3 ]
Wilaiprasitporn, Theerawit [3 ]
机构
[1] Kasetsart Univ, Fac Sci, Dept Appl Radiat & Isotopes, Bangkok 10900, Thailand
[2] Mahidol Univ, Ramathibodi Hosp, Fac Med, Dept Clin Epidemiol & Biostat, Bangkok 10400, Thailand
[3] Vidyasirimedhi Inst Sci & Technol VISTEC, Sch Informat Sci & Tech IST, Bioinspired Robot & Neural Engn BRAIN Lab, Rayong 21210, Thailand
关键词
Single photon emission computed tomography; Solid modeling; Support vector machines; Image recognition; Diseases; Tutorials; Three-dimensional displays; Parkinson's disease; SPECT image; computer-aided diagnosis (CAD); explainable AI (XAI); deep learning tutorial; CONVOLUTIONAL NEURAL-NETWORKS; AUTOMATIC CLASSIFICATION; DIAGNOSIS; PROGRESSION;
D O I
10.1109/JSEN.2021.3077949
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
In the past few years, there are several researches on Parkinson's disease (PD) recognition using single-photon emission computed tomography (SPECT) images with deep learning (DL) approach. However, the DL model's complexity usually results in difficultmodel interpretation when used in clinical. Even though there are multiple interpretation methods available for the DL model, there is no evidence of which method is suitable for PD recognition application. This tutorial aims to demonstrate the procedure to choose a suitable interpretationmethod for the PD recogni-tion model. We exhibit four DCNN architectures as an example and introduce six well-known interpretationmethods. Finally, we propose an evaluation method to measure the interpretation performance and a method to use the interpreted feedback for assisting in model selection. The evaluation demonstrates that the guided backpropagation and SHAP interpretation methods are suitable for PD recognition methods in different aspects. Guided backpropagation has the best ability to show fine-grained importance, which is proven by the highest Dice coefficient and lowest mean square error. On the other hand, SHAP can generate a better quality heatmap at the uptake depletion location, which outperforms other methods in discriminating the difference between PD and NC subjects. Shortly, the introduced interpretationmethods can contribute to not only the PD recognition application but also to sensor data processing in an AI Era (interpretable-AI) as feedback in constructing well-suited deep learning architectures for specific applications.
引用
收藏
页码:22304 / 22316
页数:13
相关论文
共 57 条
[1]  
Abadi M, 2016, PROCEEDINGS OF OSDI'16: 12TH USENIX SYMPOSIUM ON OPERATING SYSTEMS DESIGN AND IMPLEMENTATION, P265
[2]   Gait Spatiotemporal Signal Analysis for Parkinson's Disease Detection and Severity Rating [J].
Alharthi, Abdullah S. ;
Casson, Alexander J. ;
Ozanyan, Krikor B. .
IEEE SENSORS JOURNAL, 2021, 21 (02) :1838-1848
[3]   CADA-computer-aided DaTSCAN analysis [J].
Augimeri, Antonio ;
Cherubini, Andrea ;
Cascini, Giuseppe Lucio ;
Galea, Domenico ;
Caligiuri, Maria Eugenia ;
Barbagallo, Gaetano ;
Arabia, Gennarina ;
Quattrone, Aldo .
EJNMMI PHYSICS, 2016, 3 (01)
[4]   On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation [J].
Bach, Sebastian ;
Binder, Alexander ;
Montavon, Gregoire ;
Klauschen, Frederick ;
Mueller, Klaus-Robert ;
Samek, Wojciech .
PLOS ONE, 2015, 10 (07)
[5]   SPECT imaging evaluation in movement disorders: far beyond visual assessment [J].
Badiavas, Kosmas ;
Molyvda, Elisavet ;
Iakovou, Ioannis ;
Tsolaki, Magdalini ;
Psarrakos, Kyriakos ;
Karatzas, Nikolaos .
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING, 2011, 38 (04) :764-773
[6]   Deep Neural Networks Rival the Representation of Primate IT Cortex for Core Visual Object Recognition [J].
Cadieu, Charles F. ;
Hong, Ha ;
Yamins, Daniel L. K. ;
Pinto, Nicolas ;
Ardila, Diego ;
Solomon, Ethan A. ;
Majaj, Najib J. ;
DiCarlo, James J. .
PLOS COMPUTATIONAL BIOLOGY, 2014, 10 (12)
[7]   Non-motor symptoms of Parkinson's disease: dopaminergic pathophysiology and treatment [J].
Chaudhuri, K. Ray ;
Schapira, Anthony H. V. .
LANCET NEUROLOGY, 2009, 8 (05) :464-474
[8]   Refining diagnosis of Parkinson's disease with deep learning-based interpretation of dopamine transporter imaging [J].
Choi, Hongyoon ;
Ha, Seunggyun ;
Im, Hyung Jun ;
Paek, Sun Ha ;
Lee, Dong Soo .
NEUROIMAGE-CLINICAL, 2017, 16 :586-594
[9]  
Chollet F., 2015, KERAS 20 COMPUTER SO
[10]   Approximate statistical tests for comparing supervised classification learning algorithms [J].
Dietterich, TG .
NEURAL COMPUTATION, 1998, 10 (07) :1895-1923