Fully automated segmentation of oncological PET volumes using a combined multiscale and statistical model

DWG Montgomery, A Amira, H Zaidi

    Research output: Contribution to journalArticle

    87 Citations (Scopus)

    Abstract

    The widespread application of positron emission tomography (PET) in clinical oncology has driven this imaging technology into a number of new research and clinical arenas. Increasing numbers of patient scans have led to an urgent need for efficient data handling and the development of new image analysis techniques to aid clinicians in the diagnosis of disease and planning of treatment. Automatic quantitative assessment of metabolic PET data is attractive and will certainly revolutionize the practice of functional imaging since it can lower variability across institutions and may enhance the consistency of image interpretation independent of reader experience. In this paper, a novel automated system for the segmentation of oncological PET data aiming at providing an accurate quantitative analysis tool is proposed. The initial step involves expectation maximization (EM)-based mixture modeling using a k-means clustering procedure, which varies voxel order for initialization. A multiscale Markov model is then used to refine this segmentation by modeling spatial correlations between neighboring image voxels. An experimental study using an anthropomorphic thorax phantom was conducted for quantitative evaluation of the performance of the proposed segmentation algorithm. The comparison of actual tumor volumes to the volumes calculated using different segmentation methodologies including standard k-means, spatial domain Markov Random Field Model (MRFM), and the new multiscale MRFM proposed in this paper showed that the latter dramatically reduces the relative error to less than 8% for small lesions (7 mm radii) and less than 3.5% for larger lesions (9 rum radii). The analysis of the resulting segmentations of clinical oncologic PET data seems to confirm that this methodology shows promise and can successfully segment patient lesions. For problematic images, this technique enables the identification of tumors situated very close to nearby high normal physiologic uptake. The use of this technique to estimate tumor volumes for assessment of response to therapy and to delineate treatment volumes for the purpose of combined PET/CT-based radiation therapy treatment planning is also discussed. (c) 2007 American Association of Physicists in Medicine.
    LanguageEnglish
    Pages722-736
    JournalMedical Physics
    Volume34
    Issue number2
    DOIs
    Publication statusPublished - Feb 2007

    Fingerprint

    Statistical Models
    Positron-Emission Tomography
    Tumor Burden
    Medical Oncology
    Therapeutics
    Cluster Analysis
    Radiotherapy
    Thorax
    Medicine
    Technology
    Research
    Neoplasms

    Keywords

    • positron emission tomography
    • medical image segmentation
    • multiscale Markov modeling
    • Gaussian mixture modeling
    • wavelet

    Cite this

    Montgomery, DWG ; Amira, A ; Zaidi, H. / Fully automated segmentation of oncological PET volumes using a combined multiscale and statistical model. In: Medical Physics. 2007 ; Vol. 34, No. 2. pp. 722-736.
    @article{57dbde1ca03c4d5fa608e75b85289df6,
    title = "Fully automated segmentation of oncological PET volumes using a combined multiscale and statistical model",
    abstract = "The widespread application of positron emission tomography (PET) in clinical oncology has driven this imaging technology into a number of new research and clinical arenas. Increasing numbers of patient scans have led to an urgent need for efficient data handling and the development of new image analysis techniques to aid clinicians in the diagnosis of disease and planning of treatment. Automatic quantitative assessment of metabolic PET data is attractive and will certainly revolutionize the practice of functional imaging since it can lower variability across institutions and may enhance the consistency of image interpretation independent of reader experience. In this paper, a novel automated system for the segmentation of oncological PET data aiming at providing an accurate quantitative analysis tool is proposed. The initial step involves expectation maximization (EM)-based mixture modeling using a k-means clustering procedure, which varies voxel order for initialization. A multiscale Markov model is then used to refine this segmentation by modeling spatial correlations between neighboring image voxels. An experimental study using an anthropomorphic thorax phantom was conducted for quantitative evaluation of the performance of the proposed segmentation algorithm. The comparison of actual tumor volumes to the volumes calculated using different segmentation methodologies including standard k-means, spatial domain Markov Random Field Model (MRFM), and the new multiscale MRFM proposed in this paper showed that the latter dramatically reduces the relative error to less than 8{\%} for small lesions (7 mm radii) and less than 3.5{\%} for larger lesions (9 rum radii). The analysis of the resulting segmentations of clinical oncologic PET data seems to confirm that this methodology shows promise and can successfully segment patient lesions. For problematic images, this technique enables the identification of tumors situated very close to nearby high normal physiologic uptake. The use of this technique to estimate tumor volumes for assessment of response to therapy and to delineate treatment volumes for the purpose of combined PET/CT-based radiation therapy treatment planning is also discussed. (c) 2007 American Association of Physicists in Medicine.",
    keywords = "positron emission tomography, medical image segmentation, multiscale Markov modeling, Gaussian mixture modeling, wavelet",
    author = "DWG Montgomery and A Amira and H Zaidi",
    year = "2007",
    month = "2",
    doi = "10.1118/1.2432404",
    language = "English",
    volume = "34",
    pages = "722--736",
    journal = "Medical Physics",
    issn = "0094-2405",
    number = "2",

    }

    Fully automated segmentation of oncological PET volumes using a combined multiscale and statistical model. / Montgomery, DWG; Amira, A; Zaidi, H.

    In: Medical Physics, Vol. 34, No. 2, 02.2007, p. 722-736.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Fully automated segmentation of oncological PET volumes using a combined multiscale and statistical model

    AU - Montgomery, DWG

    AU - Amira, A

    AU - Zaidi, H

    PY - 2007/2

    Y1 - 2007/2

    N2 - The widespread application of positron emission tomography (PET) in clinical oncology has driven this imaging technology into a number of new research and clinical arenas. Increasing numbers of patient scans have led to an urgent need for efficient data handling and the development of new image analysis techniques to aid clinicians in the diagnosis of disease and planning of treatment. Automatic quantitative assessment of metabolic PET data is attractive and will certainly revolutionize the practice of functional imaging since it can lower variability across institutions and may enhance the consistency of image interpretation independent of reader experience. In this paper, a novel automated system for the segmentation of oncological PET data aiming at providing an accurate quantitative analysis tool is proposed. The initial step involves expectation maximization (EM)-based mixture modeling using a k-means clustering procedure, which varies voxel order for initialization. A multiscale Markov model is then used to refine this segmentation by modeling spatial correlations between neighboring image voxels. An experimental study using an anthropomorphic thorax phantom was conducted for quantitative evaluation of the performance of the proposed segmentation algorithm. The comparison of actual tumor volumes to the volumes calculated using different segmentation methodologies including standard k-means, spatial domain Markov Random Field Model (MRFM), and the new multiscale MRFM proposed in this paper showed that the latter dramatically reduces the relative error to less than 8% for small lesions (7 mm radii) and less than 3.5% for larger lesions (9 rum radii). The analysis of the resulting segmentations of clinical oncologic PET data seems to confirm that this methodology shows promise and can successfully segment patient lesions. For problematic images, this technique enables the identification of tumors situated very close to nearby high normal physiologic uptake. The use of this technique to estimate tumor volumes for assessment of response to therapy and to delineate treatment volumes for the purpose of combined PET/CT-based radiation therapy treatment planning is also discussed. (c) 2007 American Association of Physicists in Medicine.

    AB - The widespread application of positron emission tomography (PET) in clinical oncology has driven this imaging technology into a number of new research and clinical arenas. Increasing numbers of patient scans have led to an urgent need for efficient data handling and the development of new image analysis techniques to aid clinicians in the diagnosis of disease and planning of treatment. Automatic quantitative assessment of metabolic PET data is attractive and will certainly revolutionize the practice of functional imaging since it can lower variability across institutions and may enhance the consistency of image interpretation independent of reader experience. In this paper, a novel automated system for the segmentation of oncological PET data aiming at providing an accurate quantitative analysis tool is proposed. The initial step involves expectation maximization (EM)-based mixture modeling using a k-means clustering procedure, which varies voxel order for initialization. A multiscale Markov model is then used to refine this segmentation by modeling spatial correlations between neighboring image voxels. An experimental study using an anthropomorphic thorax phantom was conducted for quantitative evaluation of the performance of the proposed segmentation algorithm. The comparison of actual tumor volumes to the volumes calculated using different segmentation methodologies including standard k-means, spatial domain Markov Random Field Model (MRFM), and the new multiscale MRFM proposed in this paper showed that the latter dramatically reduces the relative error to less than 8% for small lesions (7 mm radii) and less than 3.5% for larger lesions (9 rum radii). The analysis of the resulting segmentations of clinical oncologic PET data seems to confirm that this methodology shows promise and can successfully segment patient lesions. For problematic images, this technique enables the identification of tumors situated very close to nearby high normal physiologic uptake. The use of this technique to estimate tumor volumes for assessment of response to therapy and to delineate treatment volumes for the purpose of combined PET/CT-based radiation therapy treatment planning is also discussed. (c) 2007 American Association of Physicists in Medicine.

    KW - positron emission tomography

    KW - medical image segmentation

    KW - multiscale Markov modeling

    KW - Gaussian mixture modeling

    KW - wavelet

    U2 - 10.1118/1.2432404

    DO - 10.1118/1.2432404

    M3 - Article

    VL - 34

    SP - 722

    EP - 736

    JO - Medical Physics

    T2 - Medical Physics

    JF - Medical Physics

    SN - 0094-2405

    IS - 2

    ER -