Assessment of neural networks training strategies for histomorphometric analysis of synchrotron radiation medical images

被引:7
作者
de Moura Meneses, Anderson Alvarenga [1 ,2 ]
Gomes Pinheiro, Christiano Jorge [3 ]
Rancoita, Paola [2 ,4 ]
Schaul, Tom [2 ]
Gambardella, Luca Maria [2 ]
Schirru, Roberto [1 ]
Barroso, Regina Cely [3 ]
de Oliveira, Luis Fernando [3 ]
机构
[1] Univ Fed Rio de Janeiro, COPPE, Nucl Engn Program, BR-21941972 Rio De Janeiro, Brazil
[2] Univ Lugano, IDSIA Dalle Molle Inst Artificial Intelligence, Lugano, Switzerland
[3] Univ Estado Rio De Janeiro, Rio De Janeiro, Brazil
[4] Univ Milan, Dept Math, I-20122 Milan, Italy
基金
瑞士国家科学基金会;
关键词
Synchrotron radiation; Micro-computed tomography; Histomorphometry; Artificial neural networks; Artificial intelligence; BONE HISTOMORPHOMETRY; ARCHITECTURE;
D O I
10.1016/j.nima.2010.05.022
中图分类号
TH7 [仪器、仪表];
学科分类号
0804 ; 080401 ; 081102 ;
摘要
Micro-computed tomography (mu CT) obtained by synchrotron radiation (SR) enables magnified images with a high space resolution that might be used as a non-invasive and non-destructive technique for the quantitative analysis of medical images, in particular the histomorphometry (HMM) of bony mass. In the preprocessing of such images, conventional operations such as binarization and morphological filtering are used before calculating the stereological parameters related, for example, to the trabecular bone microarchitecture. However, there is no standardization of methods for HMM based on mu CT images, especially the ones obtained with SR X-ray. Notwithstanding the several uses of artificial neural networks (ANNs) in medical imaging, their application to the HMM of SR-mu CT medical images is still incipient, despite the potential of both techniques. The contribution of this paper is the assessment and comparison of well-known training algorithms as well as the proposal of training strategies (combinations of training algorithms, sub-image kernel and symmetry information) for feed-forward ANNs in the task of bone pixels recognition in SR-mu CT medical images. For a quantitative comparison, the results of a cross validation and a statistical analysis of the results for 36 training strategies are presented. The ANNs demonstrated both very low mean square errors in the validation, and good quality segmentation of the image of interest for application to HMM in SR-mu CT medical images. (C) 2010 Elsevier B.V. All rights reserved.
引用
收藏
页码:662 / 669
页数:8
相关论文
共 29 条
  • [1] Mammography with synchrotron radiation: Phase-detection techniques
    Arfelli, F
    Bonvicini, V
    Bravin, A
    Cantatore, G
    Castelli, E
    Dalla Palma, L
    Di Michiel, M
    Fabrizioli, M
    Longo, R
    Menk, RH
    Olivo, A
    Pani, S
    Pontoni, D
    Poropat, P
    Prest, M
    Rashevsky, A
    Ratti, M
    Rigon, L
    Tromba, G
    Vacchi, A
    Vallazza, E
    Zanconati, F
    [J]. RADIOLOGY, 2000, 215 (01) : 286 - 293
  • [2] X-ray diffraction microtomography using synchrotron radiation
    Barroso, RC
    Lopes, RT
    de Jesus, EFO
    Oliveira, LF
    [J]. NUCLEAR INSTRUMENTS & METHODS IN PHYSICS RESEARCH SECTION A-ACCELERATORS SPECTROMETERS DETECTORS AND ASSOCIATED EQUIPMENT, 2001, 471 (1-2) : 75 - 79
  • [3] Three-dimensional microimaging (MRμI and μCT), finite element modeling, and rapid prototyping provide unique insights into bone architecture in osteoporosis
    Borah, B
    Gross, GJ
    Dufresne, TE
    Smith, TS
    Cockman, MD
    Chmielewski, PA
    Lundy, MW
    Hartke, JR
    Sod, EW
    [J]. ANATOMICAL RECORD, 2001, 265 (02): : 101 - 110
  • [4] Bone histomorphometry in three females with Rett syndrome
    Budden, SS
    Gunness, ME
    [J]. BRAIN & DEVELOPMENT, 2001, 23 : S133 - S137
  • [5] Bone microarchitecture evaluated by histomorphometry
    Carbonare, LD
    Valenti, MT
    Bertoldo, F
    Zanatta, M
    Zenari, S
    Realdi, G
    Lo Cascio, V
    Giannini, S
    [J]. MICRON, 2005, 36 (7-8) : 609 - 616
  • [6] Chappard D., 2008, Morphologie, V92, P162, DOI 10.1016/j.morpho.2008.10.003
  • [7] 3D histomorphometric quantification from 3D computed tomography
    de Oliveira, LF
    Lopes, RT
    [J]. NUCLEAR INSTRUMENTS & METHODS IN PHYSICS RESEARCH SECTION A-ACCELERATORS SPECTROMETERS DETECTORS AND ASSOCIATED EQUIPMENT, 2004, 525 (1-2) : 406 - 411
  • [8] Demuth H, 2004, MATH WORKS USERS GUI
  • [9] Periodontal histomorphometry and status of aged sheep subjected to ovariectomy, malnutrition and glucocorticoid application
    Dvorak, G.
    Reich, K.
    Tangl, S.
    Lill, C. A.
    Gottschalk-Baron, M.
    Watzek, G.
    Gruber, R.
    Haas, R.
    [J]. ARCHIVES OF ORAL BIOLOGY, 2009, 54 (09) : 857 - 863
  • [10] Image processing with neural networks - a review
    Egmont-Petersen, M
    de Ridder, D
    Handels, H
    [J]. PATTERN RECOGNITION, 2002, 35 (10) : 2279 - 2301