Detecting liver cirrhosis in computed tomography scans using clinically-inspired and radiomic features

被引:10
|
作者
Kotowski, Krzysztof [1 ]
Kucharski, Damian [1 ]
Machura, Bartosz [1 ]
Adamski, Szymon [1 ]
Becker, Benjamin Gutierrez [2 ]
Krason, Agata [3 ]
Zarudzki, Lukasz [4 ]
Tessier, Jean [3 ]
Nalepa, Jakub [1 ,5 ]
机构
[1] Graylight Imaging, Gliwice, Poland
[2] Roche Innovat Ctr Basel, Roche Pharm Res & Early Dev Informat, Basel, Switzerland
[3] Roche Innovat Ctr Basel, Roche Pharmaceut Res & Early Dev Ctr, Early Clin Dev Oncol, Basel, Switzerland
[4] Maria Sklodowska Curie Natl Res Inst Oncol, Dept Radiol & Diagnost Imaging, Gliwice Branch, Gliwice, Poland
[5] Silesian Tech Univ, Dept Algorithm & Software, Gliwice, Poland
关键词
Liver cirrhosis; Machine learning; Feature extraction; Computed tomography; FIBROSIS; DIAGNOSIS; MANAGEMENT; ACCURACY; CT;
D O I
10.1016/j.compbiomed.2022.106378
中图分类号
Q [生物科学];
学科分类号
07 ; 0710 ; 09 ;
摘要
Hepatic cirrhosis is an increasing cause of mortality in developed countries-it is the pathological sequela of chronic liver diseases, and the final liver fibrosis stage. Since cirrhosis evolves from the asymptomatic phase, it is of paramount importance to detect it as quickly as possible, because entering the symptomatic phase commonly leads to hospitalization and can be fatal. Understanding the state of the liver based on the abdominal computed tomography (CT) scans is tedious, user-dependent and lacks reproducibility. We tackle these issues and propose an end-to-end and reproducible approach for detecting cirrhosis from CT. It benefits from the introduced clinically-inspired features that reflect the patient's characteristics which are often investigated by experienced radiologists during the screening process. Such features are coupled with the radiomic ones extracted from the liver, and from the suggested region of interest which captures the liver's boundary. The rigorous experiments, performed over two heterogeneous clinical datasets (two cohorts of 241 and 32 patients) revealed that extracting radiomic features from the liver's rectified contour is pivotal to enhance the classification abilities of the supervised learners. Also, capturing clinically-inspired image features significantly improved the performance of such models, and the proposed features were consistently selected as the important ones. Finally, we showed that selecting the most discriminative features leads to the Pareto-optimal models with enhanced feature-level interpretability, as the number of features was dramatically reduced (280x) from thousands to tens.
引用
收藏
页数:15
相关论文
共 50 条
  • [1] Using Artificial Intelligence to Predict Cirrhosis From Computed Tomography Scans
    Mazumder, Nikhilesh R.
    Enchakalody, Binu
    Zhang, Peng
    Su, Grace L.
    CLINICAL AND TRANSLATIONAL GASTROENTEROLOGY, 2023, 14 (10)
  • [2] Radiomics of Computed Tomography (CT) Liver Scans: A Novel Method to Evaluate Liver Cirrhosis and Steatosis
    Marya, Neil B.
    Shrestha, Suman
    Vedantham, Srinivasan
    Patel, Krunal
    Rupawala, Abbas H.
    Aldrugh, Summer
    Karam, Adib R.
    Zacharias, Isabel
    Karellas, Andrew
    GASTROENTEROLOGY, 2016, 150 (04) : S1166 - S1167
  • [3] Positron Emission Tomography/Computed Tomographic Scans Compared to Computed Tomographic Scans for Detecting Colorectal Liver Metastases A Systematic Review
    Patel, Seema
    McCall, Michael
    Ohinmaa, Arto
    Bigam, David
    Dryden, Donna M.
    ANNALS OF SURGERY, 2011, 253 (04) : 666 - 671
  • [4] Effect of tube current on computed tomography radiomic features
    Mackin, Dennis
    Ger, Rachel
    Dodge, Cristina
    Fave, Xenia
    Chi, Pai-Chun
    Zhang, Lifei
    Yang, Jinzhong
    Bache, Steve
    Dodge, Charles
    Jones, A. Kyle
    Court, Laurence
    SCIENTIFIC REPORTS, 2018, 8
  • [5] Effect of tube current on computed tomography radiomic features
    Dennis Mackin
    Rachel Ger
    Cristina Dodge
    Xenia Fave
    Pai-Chun Chi
    Lifei Zhang
    Jinzhong Yang
    Steve Bache
    Charles Dodge
    A. Kyle Jones
    Laurence Court
    Scientific Reports, 8
  • [6] Distinguishing granulomas from adenocarcinomas by integrating stable and discriminating radiomic features on non-contrast computed tomography scans
    Khorrami, Mohammadhadi
    Bera, Kaustav
    Thawani, Rajat
    Rajiah, Prabhakar
    Gupta, Amit
    Fu, Pingfu
    Linden, Philip
    Pennell, Nathan
    Jacono, Frank
    Gilkeson, Robert C.
    Velcheti, Vamsidhar
    Madabhushi, Anant
    EUROPEAN JOURNAL OF CANCER, 2021, 148 : 146 - 158
  • [7] Repeatability of computed tomography liver radiomic features in a nonhuman primate model of diet-induced steatosis
    Wang, Hui
    Solomon, Jeffrey
    Reza, Syed M. S.
    Yang, Hee-Jeong
    Chu, Winston T.
    Crozier, Ian
    Sayre, Philip J.
    Lee, Byeong Y.
    Mani, Venkatesh
    Friedrich, Thomas C.
    O'Connor, David H.
    Worwa, Gabriella
    Kuhn, Jens H.
    Calcagno, Claudia
    Castro, Marcelo A.
    JOURNAL OF MEDICAL IMAGING, 2023, 10 (06)
  • [8] Computed tomography findings in liver fibrosis and cirrhosis
    Adrian, Huber
    Lukas, Ebner
    Matteo, Montani
    Nasser, Semmo
    Kingshuk, Roy Choudhury
    Johannes, Heverhagen
    Andreas, Christe
    SWISS MEDICAL WEEKLY, 2014, 144
  • [9] Automated body composition analysis of clinically acquired computed tomography scans using neural networks
    Paris, Michael T.
    Tandon, Puneeta
    Heyland, Daren K.
    Furberg, Helena
    Premji, Tahira
    Low, Gavin
    Mourtzakis, Marina
    CLINICAL NUTRITION, 2020, 39 (10) : 3049 - 3055
  • [10] Improving the Reproducibility of Computed Tomography Radiomic Features Using an Enhanced Hierarchical Feature Synthesis Network
    Jeong, Dawun
    Hong, Youngtaek
    Lee, Jina
    Lee, Seul Bi
    Cho, Yeon Jin
    Shim, Hackjoon
    Chang, Hyuk-Jae
    IEEE ACCESS, 2024, 12 : 27648 - 27660