Image Contrast, Image Pre-Processing, and T<sub>1</sub> Mapping Affect MRI Radiomic Feature Repeatability in Patients with Colorectal Cancer Liver Metastases
Imaging biomarkers require technical, biological, and clinical validation to be translated into robust tools in research or clinical settings. This study contributes to the technical validation of radiomic features from magnetic resonance imaging (MRI) by evaluating the repeatability of features fro...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-01-01
|
| Series: | Cancers |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-6694/13/2/240 |
| _version_ | 1797413387937775616 |
|---|---|
| author | Damien J. McHugh Nuria Porta Ross A. Little Susan Cheung Yvonne Watson Geoff J. M. Parker Gordon C. Jayson James P. B. O’Connor |
| author_facet | Damien J. McHugh Nuria Porta Ross A. Little Susan Cheung Yvonne Watson Geoff J. M. Parker Gordon C. Jayson James P. B. O’Connor |
| author_sort | Damien J. McHugh |
| collection | DOAJ |
| description | Imaging biomarkers require technical, biological, and clinical validation to be translated into robust tools in research or clinical settings. This study contributes to the technical validation of radiomic features from magnetic resonance imaging (MRI) by evaluating the repeatability of features from four MR sequences: pre-contrast T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula>- and T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>2</mn></msub></semantics></math></inline-formula>-weighted images, pre-contrast quantitative T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula> maps (qT<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula>), and contrast-enhanced T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula>-weighted images. Fifty-one patients with colorectal cancer liver metastases were scanned twice, up to 7 days apart. Repeatability was quantified using the intraclass correlation coefficient (ICC) and repeatability coefficient (RC), and the impact of non-Gaussian feature distributions and image normalisation was evaluated. Most radiomic features had non-Gaussian distributions, but Box–Cox transformations enabled ICCs and RCs to be calculated appropriately for an average of 97% of features across sequences. ICCs ranged from <inline-formula><math display="inline"><semantics><mrow><mn>0.30</mn></mrow></semantics></math></inline-formula> to <inline-formula><math display="inline"><semantics><mrow><mn>0.99</mn></mrow></semantics></math></inline-formula>, with volume and other shape features tending to be most repeatable; volume ICC > 0.98 for all sequences. 19% of features from non-normalised images exhibited significantly different ICCs in pair-wise sequence comparisons. Normalisation tended to increase ICCs for pre-contrast T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula>- and T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>2</mn></msub></semantics></math></inline-formula>-weighted images, and decrease ICCs for qT<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula> maps. RCs tended to vary more between sequences than ICCs, showing that evaluations of feature performance depend on the chosen metric. This work suggests that feature-specific repeatability, from specific combinations of MR sequence and pre-processing steps, should be evaluated to select robust radiomic features as biomarkers in specific studies. In addition, as different repeatability metrics can provide different insights into a specific feature, consideration of the appropriate metric should be taken in a study-specific context. |
| first_indexed | 2024-03-09T05:17:09Z |
| format | Article |
| id | doaj.art-5e453909953a40aeb946649f29ab1043 |
| institution | Directory Open Access Journal |
| issn | 2072-6694 |
| language | English |
| last_indexed | 2024-03-09T05:17:09Z |
| publishDate | 2021-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Cancers |
| spelling | doaj.art-5e453909953a40aeb946649f29ab10432023-12-03T12:43:53ZengMDPI AGCancers2072-66942021-01-0113224010.3390/cancers13020240Image Contrast, Image Pre-Processing, and T<sub>1</sub> Mapping Affect MRI Radiomic Feature Repeatability in Patients with Colorectal Cancer Liver MetastasesDamien J. McHugh0Nuria Porta1Ross A. Little2Susan Cheung3Yvonne Watson4Geoff J. M. Parker5Gordon C. Jayson6James P. B. O’Connor7Division of Cancer Sciences, The University of Manchester, Manchester M13 9PL, UKClinical Trials and Statistics Unit, Institute of Cancer Research, London SW3 6JB, UKDivision of Cancer Sciences, The University of Manchester, Manchester M13 9PL, UKDivision of Cancer Sciences, The University of Manchester, Manchester M13 9PL, UKDivision of Cancer Sciences, The University of Manchester, Manchester M13 9PL, UKCentre for Medical Image Computing, University College London, London WC1V 6LJ, UKDivision of Cancer Sciences, The University of Manchester, Manchester M13 9PL, UKDivision of Cancer Sciences, The University of Manchester, Manchester M13 9PL, UKImaging biomarkers require technical, biological, and clinical validation to be translated into robust tools in research or clinical settings. This study contributes to the technical validation of radiomic features from magnetic resonance imaging (MRI) by evaluating the repeatability of features from four MR sequences: pre-contrast T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula>- and T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>2</mn></msub></semantics></math></inline-formula>-weighted images, pre-contrast quantitative T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula> maps (qT<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula>), and contrast-enhanced T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula>-weighted images. Fifty-one patients with colorectal cancer liver metastases were scanned twice, up to 7 days apart. Repeatability was quantified using the intraclass correlation coefficient (ICC) and repeatability coefficient (RC), and the impact of non-Gaussian feature distributions and image normalisation was evaluated. Most radiomic features had non-Gaussian distributions, but Box–Cox transformations enabled ICCs and RCs to be calculated appropriately for an average of 97% of features across sequences. ICCs ranged from <inline-formula><math display="inline"><semantics><mrow><mn>0.30</mn></mrow></semantics></math></inline-formula> to <inline-formula><math display="inline"><semantics><mrow><mn>0.99</mn></mrow></semantics></math></inline-formula>, with volume and other shape features tending to be most repeatable; volume ICC > 0.98 for all sequences. 19% of features from non-normalised images exhibited significantly different ICCs in pair-wise sequence comparisons. Normalisation tended to increase ICCs for pre-contrast T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula>- and T<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>2</mn></msub></semantics></math></inline-formula>-weighted images, and decrease ICCs for qT<inline-formula><math display="inline"><semantics><msub><mrow></mrow><mn>1</mn></msub></semantics></math></inline-formula> maps. RCs tended to vary more between sequences than ICCs, showing that evaluations of feature performance depend on the chosen metric. This work suggests that feature-specific repeatability, from specific combinations of MR sequence and pre-processing steps, should be evaluated to select robust radiomic features as biomarkers in specific studies. In addition, as different repeatability metrics can provide different insights into a specific feature, consideration of the appropriate metric should be taken in a study-specific context.https://www.mdpi.com/2072-6694/13/2/240radiomicsMRIrepeatabilityrepeatability coefficientintraclass correlation coefficientliver metastases |
| spellingShingle | Damien J. McHugh Nuria Porta Ross A. Little Susan Cheung Yvonne Watson Geoff J. M. Parker Gordon C. Jayson James P. B. O’Connor Image Contrast, Image Pre-Processing, and T<sub>1</sub> Mapping Affect MRI Radiomic Feature Repeatability in Patients with Colorectal Cancer Liver Metastases Cancers radiomics MRI repeatability repeatability coefficient intraclass correlation coefficient liver metastases |
| title | Image Contrast, Image Pre-Processing, and T<sub>1</sub> Mapping Affect MRI Radiomic Feature Repeatability in Patients with Colorectal Cancer Liver Metastases |
| title_full | Image Contrast, Image Pre-Processing, and T<sub>1</sub> Mapping Affect MRI Radiomic Feature Repeatability in Patients with Colorectal Cancer Liver Metastases |
| title_fullStr | Image Contrast, Image Pre-Processing, and T<sub>1</sub> Mapping Affect MRI Radiomic Feature Repeatability in Patients with Colorectal Cancer Liver Metastases |
| title_full_unstemmed | Image Contrast, Image Pre-Processing, and T<sub>1</sub> Mapping Affect MRI Radiomic Feature Repeatability in Patients with Colorectal Cancer Liver Metastases |
| title_short | Image Contrast, Image Pre-Processing, and T<sub>1</sub> Mapping Affect MRI Radiomic Feature Repeatability in Patients with Colorectal Cancer Liver Metastases |
| title_sort | image contrast image pre processing and t sub 1 sub mapping affect mri radiomic feature repeatability in patients with colorectal cancer liver metastases |
| topic | radiomics MRI repeatability repeatability coefficient intraclass correlation coefficient liver metastases |
| url | https://www.mdpi.com/2072-6694/13/2/240 |
| work_keys_str_mv | AT damienjmchugh imagecontrastimagepreprocessingandtsub1submappingaffectmriradiomicfeaturerepeatabilityinpatientswithcolorectalcancerlivermetastases AT nuriaporta imagecontrastimagepreprocessingandtsub1submappingaffectmriradiomicfeaturerepeatabilityinpatientswithcolorectalcancerlivermetastases AT rossalittle imagecontrastimagepreprocessingandtsub1submappingaffectmriradiomicfeaturerepeatabilityinpatientswithcolorectalcancerlivermetastases AT susancheung imagecontrastimagepreprocessingandtsub1submappingaffectmriradiomicfeaturerepeatabilityinpatientswithcolorectalcancerlivermetastases AT yvonnewatson imagecontrastimagepreprocessingandtsub1submappingaffectmriradiomicfeaturerepeatabilityinpatientswithcolorectalcancerlivermetastases AT geoffjmparker imagecontrastimagepreprocessingandtsub1submappingaffectmriradiomicfeaturerepeatabilityinpatientswithcolorectalcancerlivermetastases AT gordoncjayson imagecontrastimagepreprocessingandtsub1submappingaffectmriradiomicfeaturerepeatabilityinpatientswithcolorectalcancerlivermetastases AT jamespboconnor imagecontrastimagepreprocessingandtsub1submappingaffectmriradiomicfeaturerepeatabilityinpatientswithcolorectalcancerlivermetastases |