Benefit of Dual Energy CT for Lesion Localization and Classification with Convolutional Neural Networks

被引:1
作者
Shapira, Nadav [1 ]
Fokuhl, Julia [2 ,3 ]
Schultheiss, Manuel [4 ]
Beck, Stefanie [4 ]
Kopp, Felix K. [4 ]
Pfeiffer, Daniela [4 ]
Dangelmaier, Julia [4 ]
Pahn, Gregor [5 ]
Sauter, Andreas P. [4 ]
Renger, Bernhard [4 ]
Fingerle, Alexander A. [4 ]
Maidment, Andrew D. A. [1 ]
Rummeny, Ernst J. [4 ]
Albarqouni, Shadi [3 ]
Navab, Nassir [3 ]
Noel, Peter B. [1 ,4 ]
机构
[1] Univ Penn, Dept Radiol, Perelman Sch Med, Philadelphia, PA 19104 USA
[2] Tech Univ Munich, Dept Diagnost & Intervent Radiol, Munich, Germany
[3] Tech Univ Munich, Comp Aided Med Procedures CAMP, Garching, Germany
[4] Tech Univ Munich, Dept Diagnost & Intervent Radiol, Munich, Germany
[5] Philips Germany GmbH, Clin Sci CT, Hamburg, Germany
来源
MEDICAL IMAGING 2020: PHYSICS OF MEDICAL IMAGING | 2020年 / 11312卷
关键词
Dual Energy CT (DECT); Convolutional Neural Networks (CNN); Hepatic imaging; Liver lesion localization; Liver lesion classification; HEPATIC METASTASES; SPECTRAL CT; PRINCIPLES;
D O I
10.1117/12.2549291
中图分类号
R318 [生物医学工程];
学科分类号
0831 ;
摘要
Dual Energy CT is a modern imaging technique that is utilized in clinical practice to acquire spectral information for various diagnostic purposes including the identification, classification, and characterization of different liver lesions. It provides additional information that, when compared to the information available from conventional CT datasets, has the potential to benefit existing computer vision techniques by improving their accuracy and reliability. In order to evaluate the additional value of spectral versus conventional datasets when being used as input for machine learning algorithms, we implemented a weakly-supervised Convolutional Neural Network (CNN) that learns liver lesion localization and classification without pixel-level ground truth annotations. We evaluated the lesion classification (healthy, cyst, hypodense metastasis) and localization performance of the network for various conventional and spectral input datasets obtained from the same CT scan. The best results for lesion localization were found for the spectral datasets with distances of 8.22 +/- 10.72 mm, 8.78 +/- 15.21 mm and 8.29 +/- 12.97 mm for iodine maps, 40 keV and 70 keV virtual mono-energetic images, respectively, while lesion localization distances of 10.58 +/- 17.65 mm were measured for the conventional dataset. In addition, the 40 keV virtual mono-energetic datasets achieved the highest overall lesion classification accuracy of 0.899 compared to 0.854 measured for the conventional datasets. The enhanced localization and classification results that we observed for spectral CT data demonstrates that combining machine-learning technology with spectral CT information may improve the clinical workflow as well as the diagnostic accuracy.
引用
收藏
页数:7
相关论文
共 33 条
  • [1] ACS, 2018, KEY STAT CANC
  • [2] Oncologic Applications of Dual-Energy CT in the Abdomen
    Agrawal, Mukta D.
    Pinho, Daniella F.
    Kulkarni, Naveen M.
    Hahn, Peter F.
    Guimaraes, Alexander R.
    Sahani, Dushyant V.
    [J]. RADIOGRAPHICS, 2014, 34 (03) : 589 - 612
  • [3] [Anonymous], 2001, TUMORS LIVER INTRAHE
  • [4] HEPATIC METASTASES - BASIC PRINCIPLES AND IMPLICATIONS FOR RADIOLOGISTS
    BAKER, ME
    PELLEY, R
    [J]. RADIOLOGY, 1995, 197 (02) : 329 - 337
  • [5] Diagnosis and management of solid benign liver lesions
    Belghiti, Jacques
    Cauchy, Francois
    Paradis, Valerie
    Vilgrain, Valerie
    [J]. NATURE REVIEWS GASTROENTEROLOGY & HEPATOLOGY, 2014, 11 (12) : 737 - 749
  • [6] Carmi R., 2004, REC NUCL SCI S, V4, P1876
  • [7] MR Imaging of Benign Focal Liver Lesions
    Cogley, Jonathan R.
    Miller, Frank H.
    [J]. RADIOLOGIC CLINICS OF NORTH AMERICA, 2014, 52 (04) : 657 - +
  • [8] De Cecco C.N., 2015, Dual Energy CT Oncol, P1
  • [9] Assessment of quantification accuracy and image quality of a full-body dual-layer spectral CT system
    Ehn, Sebastian
    Sellerer, Thorsten
    Muenzel, Daniela
    Fingerle, Alexander A.
    Kopp, Felix
    Duda, Manuela
    Mei, Kai
    Renger, Bernhard
    Herzen, Julia
    Dangelmaier, Julia
    Schwaiger, Benedikt J.
    Sauter, Andreas
    Riederer, Isabelle
    Renz, Martin
    Braren, Rickmer
    Rummeny, Ernst J.
    Pfeiffer, Franz
    Noeel, Peter B.
    [J]. JOURNAL OF APPLIED CLINICAL MEDICAL PHYSICS, 2018, 19 (01): : 204 - 217
  • [10] Feldman L. B. Mark, 2010, PATHOPHYSIOLOGY DIAG, V9th