Volume 82, Issue 5 (August 2024)
Tehran Univ Med J 2024, 82(5): 376-383 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Soleymani Y, Aghahoseini F, Sheikhzadeh P. Correlation of radiomics features extracted from nuclear medicine images with lesion metabolism in patients with colon cancer. Tehran Univ Med J 2024; 82 (5) :376-383
URL: http://tumj.tums.ac.ir/article-1-13166-en.html
URL: http://tumj.tums.ac.ir/article-1-13166-en.html
1- Department of Neuroscience and Addiction Studies, School of Advanced Technologies in Medicine, Tehran University of Medical Sciences, Tehran, Iran.
2- Department of Nuclear Medicine, School of Medicine, Imam Khomeini Hospital Complex, Tehran University of Medical Sciences, Tehran, Iran.
3- Department of Nuclear Medicine, School of Medicine, Imam Khomeini Hospital Complex, Tehran University of Medical Sciences, Tehran, Iran. ,psh82@yahoo.com
2- Department of Nuclear Medicine, School of Medicine, Imam Khomeini Hospital Complex, Tehran University of Medical Sciences, Tehran, Iran.
3- Department of Nuclear Medicine, School of Medicine, Imam Khomeini Hospital Complex, Tehran University of Medical Sciences, Tehran, Iran. ,
Abstract: (33 Views)
Background: Nuclear medicine imaging has shown high accuracy in evaluating the metabolism of colon cancer lesions. The aim of this study was to investigate the ability of radiomics features extracted from nuclear medicine images as non-invasive biomarkers of lesion metabolism in patients with colon cancer by examining the correlation of these features with SUV (standardized uptake value) max values.
Methods: The current study was a cross-sectional study that was conducted from July 2022 to July 2023 in the nuclear medicine department of Tehran University of Medical Sciences. In this study, PET/CT (positron emission tomography/computed tomography) images of 60 patients with primary colon cancer were used. Colon cancer lesions were manually delineated on PET images by an experienced physician and saved as VOIs (volumes of interest). Thirty-two textural radiomics features were extracted from each VOI, including feature groups of gray-level co-occurrence matrix (GLCM), gray-level run-length matrix (GLRLM), neighborhood grey-level difference matrix (NGLDM), and gray-level zone length matrix (GLZLM). Then, the correlation of these features with SUVmax values was investigated using the Spearman correlation coefficient statistical test. Also, the value of p<0.05 was considered as the significance level of the test.
Methods: The current study was a cross-sectional study that was conducted from July 2022 to July 2023 in the nuclear medicine department of Tehran University of Medical Sciences. In this study, PET/CT (positron emission tomography/computed tomography) images of 60 patients with primary colon cancer were used. Colon cancer lesions were manually delineated on PET images by an experienced physician and saved as VOIs (volumes of interest). Thirty-two textural radiomics features were extracted from each VOI, including feature groups of gray-level co-occurrence matrix (GLCM), gray-level run-length matrix (GLRLM), neighborhood grey-level difference matrix (NGLDM), and gray-level zone length matrix (GLZLM). Then, the correlation of these features with SUVmax values was investigated using the Spearman correlation coefficient statistical test. Also, the value of p<0.05 was considered as the significance level of the test.
|
Results: A comprehensive analysis revealed that more than 96% of the examined radiomics features specifically, 31 out of 32 exhibited a statistically significant correlation with lesion metabolism values, as indicated by p-values less than 0.05. Among these features, GLZLM_HGZE stood out with a high correlation coefficient of 0.9881, alongside a significance level of less than 0.0001. Similarly, GLZLM_SZHGE also demonstrated a strong correlation, with a coefficient of 0.9723 and a significance level below 0.0001, indicating a robust relationship with SUVmax values. In contrast, GLZLM_LZHGE was the only feature that failed to show a significant correlation with lesion metabolism values (p>0.05).
Conclusion: The radiomics method has the potential to be used as a completely non-invasive method to evaluate the metabolism of colon cancer lesions and facilitate the monitoring and treatment of patients with colon cancer. |
Type of Study: Original Article |
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
