Journal article
Choline PET/CT features to predict survival outcome in high risk prostate cancer restaging: a preliminary machine-learning radiomics study.
- Alongi P Nuclear Medicine Unit, Fondazione Istituto G.Giglio, Cefalù, Palermo, Italy - alongi.pierpaolo@gmail.com.
- Laudicella R Nuclear Medicine Unit, Fondazione Istituto G.Giglio, Cefalù, Palermo, Italy.
- Stefano A Institute of Molecular Bioimaging and Physiology, National Research Council (IBFM-CNR), Cefalù, Palermo, Italy.
- Caobelli F Clinic of Radiology & Nuclear Medicine, University Hospital Basel, University of Basel, Basel, Switzerland.
- Comelli A Ri.MED Foundation, Palermo, Italy.
- Vento A Department of Biomedical and Dental Sciences and of Morpho-functional Imaging, Nuclear Medicine Unit, University of Messina, Messina, Italy.
- Sardina D Department of Industrial and Digital Innovation (DIID), University of Palermo, Palermo, Italy.
- Ganduscio G Department of Industrial and Digital Innovation (DIID), University of Palermo, Palermo, Italy.
- Toia P Cellular and Molecular Pathophysiology Laboratory, Department of Surgical, Oncological and Stomatological Sciences, University of Palermo, Palermo, Italy.
- Ceci F Department of Radiology, DIBIMED, University of Palermo, Palermo, Italy.
- Mapelli P Nuclear Medicine, Department of Medical Sciences, University of Turin, Turin, Italy.
- Picchio M Nuclear Medicine, Department of Medical Sciences, University of Turin, Turin, Italy.
- Midiri M Cellular and Molecular Pathophysiology Laboratory, Department of Surgical, Oncological and Stomatological Sciences, University of Palermo, Palermo, Italy.
- Baldari S Department of Biomedical and Dental Sciences and of Morpho-functional Imaging, Nuclear Medicine Unit, University of Messina, Messina, Italy.
- Lagalla R Cellular and Molecular Pathophysiology Laboratory, Department of Surgical, Oncological and Stomatological Sciences, University of Palermo, Palermo, Italy.
- Russo G Institute of Molecular Bioimaging and Physiology, National Research Council (IBFM-CNR), Cefalù, Palermo, Italy.
- 2020-06-17
Published in:
- The quarterly journal of nuclear medicine and molecular imaging : official publication of the Italian Association of Nuclear Medicine (AIMN) [and] the International Association of Radiopharmacology (IAR), [and] Section of the Society of.... - 2020
English
BACKGROUND
Radiomic features are increasingly utilized to evaluate tumor heterogeneity in PET imaging but to date its role has not been investigated for Cho-PET in prostate cancer. The potential application of radiomics features analysis using a machine-learning radiomics algorithm was evaluated to select 18F-Cho PET/CT imaging features to predict disease progression in PCa.
METHODS
We retrospectively analyzed high-risk PCa patients who underwent restaging 18F-Cho PET/CT from November 2013 to May 2018. 18F-Cho PET/CT studies and related structures containing volumetric segmentations were imported in the "CGITA" toolbox to extract imaging features from each lesion. A Machine-learning model has been adapted using NCA for feature selection, while DA was used as a method for feature classification and performance analysis.
RESULTS
106 imaging features were extracted for 46 lesions for a total of 4876 features analyzed. No significant differences between the training and validating sets in terms of age, sex, PSA values, lesion location and size (p > 0.05) were demonstrated by the machine-learning model. Thirteen features were able to discriminate FU disease status after NCA selection. Best performance in DA classification was obtained using the combination of the 13 selected features (sensitivity 74%, specificity 58% and accuracy 66%) compared to the use of all features (sensitivity 40%, specificity 52%, and accuracy 51%). Per-site performance of the 13 selected features in DA classification were as follow: T= sensitivity 63%, specificity 83%, accuracy 71%; N= sensitivity 87%, specificity 91% of and accuracy 90%; bone-M= sensitivity 33%, specificity 77% and accuracy 66%.
CONCLUSIONS
An artificial intelligence model demonstrated to be feasible and able to select a panel of 18F-Cho PET/CT features with valuable association with PCa patients' outcome.
Radiomic features are increasingly utilized to evaluate tumor heterogeneity in PET imaging but to date its role has not been investigated for Cho-PET in prostate cancer. The potential application of radiomics features analysis using a machine-learning radiomics algorithm was evaluated to select 18F-Cho PET/CT imaging features to predict disease progression in PCa.
METHODS
We retrospectively analyzed high-risk PCa patients who underwent restaging 18F-Cho PET/CT from November 2013 to May 2018. 18F-Cho PET/CT studies and related structures containing volumetric segmentations were imported in the "CGITA" toolbox to extract imaging features from each lesion. A Machine-learning model has been adapted using NCA for feature selection, while DA was used as a method for feature classification and performance analysis.
RESULTS
106 imaging features were extracted for 46 lesions for a total of 4876 features analyzed. No significant differences between the training and validating sets in terms of age, sex, PSA values, lesion location and size (p > 0.05) were demonstrated by the machine-learning model. Thirteen features were able to discriminate FU disease status after NCA selection. Best performance in DA classification was obtained using the combination of the 13 selected features (sensitivity 74%, specificity 58% and accuracy 66%) compared to the use of all features (sensitivity 40%, specificity 52%, and accuracy 51%). Per-site performance of the 13 selected features in DA classification were as follow: T= sensitivity 63%, specificity 83%, accuracy 71%; N= sensitivity 87%, specificity 91% of and accuracy 90%; bone-M= sensitivity 33%, specificity 77% and accuracy 66%.
CONCLUSIONS
An artificial intelligence model demonstrated to be feasible and able to select a panel of 18F-Cho PET/CT features with valuable association with PCa patients' outcome.
- Language
-
- English
- Identifiers
-
- DOI 10.23736/S1824-4785.20.03227-6
- PMID 32543166
- Persistent URL
- https://sonar.ch/global/documents/160225
Statistics
Document views: 27
File downloads: