Advertisement
Original paper| Volume 60, P91-99, April 2019

Download started.

Ok

Analysis of a CT patient dose database with an unsupervised clustering approach

Published:March 29, 2019DOI:https://doi.org/10.1016/j.ejmp.2019.03.015

      Highlights

      • Cluster analysis (CA) can be used to summarize data of CT patient dose registries.
      • Different CA methods were tested, obtaining the best results with the Ward method.
      • CA highlights the combinations of exposure parameters most used in practice.
      • Outliers analysis associated with CA allows to investigate anomalous high dose values.

      Abstract

      Purpose

      This study investigated the benefits of implementing a cluster analysis technique to extract relevant information from a computed tomography (CT) dose registry archive.

      Methods

      A CT patient dose database consisting of about 12,000 examinations and 29,000 single scans collected from three CT systems was interrogated. The database was divided into six subsets according to the equipment and the reference phantoms in the definition of the dose indicators. Hierarchical (single, average, and complete linkage, Ward) and not hierarchical (K-means) clustering methods were implemented using R software. The suitable number of clusters for each CT system was determined by analysing the dendrogram, the within clusters sum of squares, and the cluster content. Summary statistics were produced for each cluster, and the outliers of the dose indicator distribution were investigated.

      Results

      Ward clustering identified the most common combinations of scanning parameters for each group. The optimal number of clusters for each CT equipment system ranged from 5 to 15. The main diagnostic applications were then extracted from each cluster. Outlier analysis of the dose indicator distribution of each cluster revealed potential improper settings that resulted in increased patient dose.

      Conclusions

      Clustering methods applied to CT patient dose archives provide a quick and effective overview of the main combinations of currently used exposure parameters and the consequences for dose indicator distributions, also when protocol labels and/or study descriptions are not homogeneous.

      Graphical abstract

      Keywords

      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Physica Medica: European Journal of Medical Physics
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Rehani M.M.
        Tracking of examination and dose: overview.
        Radiat Prot Dosim. 2015; 165: 50-52
        • Talati R.K.
        • Dunkin J.
        • Parikh S.
        • Moore W.H.
        Current methods of monitoring radiation exposure from CT.
        J Am Coll Radiol. 2013; 10: 702-707
        • Szczykutowicz T.P.
        • Bour R.K.
        • Rubert N.
        • Wendt G.
        • Pozniak M.
        • Ranallo F.N.
        CT protocol management: simplifying the process by using a master protocol concept.
        J Appl Clin Med Phys. 2015; 16: 228-243
        • Varian H.R.
        Big data: new tricks for econometrics.
        J Econ Perspect. 2014; 28: 3-28
        • McCollough C.H.
        Automated data mining of exposure information for dose management and patient safety initiatives in medical imaging.
        Radiology. 2012; 264: 322-324
        • Ikuta I.
        • Sodickson A.
        • Wasser E.J.
        • Warden G.I.
        • Gerbaudo V.H.
        • Khorasani R.
        Exposing exposure: enhancing patient safety through automated data mining of nuclear medicine reports for quality assurance and organ dose monitoring.
        Radiology. 2012; 264: 406-413
        • Sodickson A.
        • Warden G.I.
        • Farkas C.E.
        • Ikuta I.
        • Prevedello L.M.
        • Andriole K.P.
        • et al.
        Exposing exposure: automated anatomy-specific CT radiation exposure extraction for quality assurance and radiation monitoring.
        Radiology. 2012; 264: 397-405
        • Wang S.
        • Pavlicek W.
        • Roberts C.C.
        • Langer S.G.
        • Zhang M.
        • Hu M.
        • et al.
        An automated DICOM database capable of arbitrary data mining (including radiation dose indicators) for quality monitoring.
        J Digit Imaging. 2011; 24: 223-233
        • Han J.
        • Kamber M.
        • Pei J.
        Data mining, concepts and techniques.
        Elsevier, 2012
        • Nicol R.M.
        • Wayte S.C.
        • Bridges A.J.
        • Koller C.J.
        Experiences of using a commercial dose management system (GE Dose Watch) for CT examinations.
        Br J Radiol. 2016; 89: 20150617
        • Chormunge S.
        • Jena S.
        Efficiency and effectiveness of clustering algorithms for high dimensional data.
        Int J Comput Appl. 2015; 125: 35-40
        • Odilia Y.
        • Ramdeen K.T.
        The quantitative methods for psychology hierarchical cluster analysis: comparison of three linkage measures and application to psychological data.
        Quant Methods Psychol. 2015; 11: 8-21
        • Ward J.H.
        Hierarchical grouping to optimize an objective function.
        J Am Stat Assoc. 1963; 58: 236-244
        • Zambelli A.E.
        A data-driven approach to estimating the number of clusters in hierarchical clustering.
        F1000Res. 2016; (pii: ISCB Comm J-2809): 5
        • Origgi D.
        • Vigorito S.
        • Villa G.
        • Bellomi M.
        • Tosi G.
        Survey of computed tomography techniques and absorbed dose in Italian hospitals: a comparison between two methods to estimate the dose-length product and the effective dose and to verify fulfilment of the diagnostic reference levels.
        Eur Radiol. 2006; 16: 227-237
        • Kanal K.M.
        • Stewart B.K.
        • Kolokythas O.
        • Shuman W.P.
        Impact of operator-selected image noise index and reconstruction slice thickness on patient radiation dose in 64-MDCT.
        Am J Roentgenol. 2007; 189: 219-225
        • Tang H.
        • Yu N.
        • Jia Y.
        • Yu Y.
        • Duan H.
        • Han D.
        • et al.
        Assessment of noise reduction potential and image quality improvement of a new generation adaptive statistical iterative reconstruction (ASIR-V) in chest CT.
        Br J Radiol. 2018; 91: 20170521
        • Matsubara K.
        • Koshida K.
        • Ichikawa K.
        • Suzuki M.
        • Takata T.
        • Yamamoto T.
        • et al.
        Misoperation of CT automatic tube current modulation systems with inappropriate patient centering: phantom studies.
        Am J Roentgenol. 2009; 192: 862-865
        • Serna A.
        • Ramos D.
        • Angosto E.G.
        • Garcia-Sanchez A.J.
        • Chans M.A.
        • Benedicto-Orovitg J.M.
        • et al.
        Optimization of CT protocols using cause and effect analysis of outliers.
        Physica Med. 2018; 55: 1-7
        • Hummel M.
        • Edelmann D.
        • Kopp-Schneider A.
        Clustering of samples and variables with mixed-type data.
        PLoS One. 2017; 12e0188274
        • Dunn J.
        A fuzzy relative of the ISODATA process and its use in detecting compact well separated cluster.
        J Cybern. 1974; 3: 32-57