AI-based quantification of whole-body tumour burden on somatostatin receptor PET/CT
(2023) In European journal of hybrid imaging 7(1).- Abstract
BACKGROUND: Segmenting the whole-body somatostatin receptor-expressing tumour volume (SRETVwb) on positron emission tomography/computed tomography (PET/CT) images is highly time-consuming but has shown value as an independent prognostic factor for survival. An automatic method to measure SRETVwb could improve disease status assessment and provide a tool for prognostication. This study aimed to develop an artificial intelligence (AI)-based method to detect and quantify SRETVwb and total lesion somatostatin receptor expression (TLSREwb) from [
68Ga]Ga-DOTA-TOC/TATE PET/CT images.
METHODS: A UNet3D convolutional neural network (CNN) was used to train an AI model with [
68Ga]Ga-DOTA-TOC/TATE PET/CT images, where all... (More)BACKGROUND: Segmenting the whole-body somatostatin receptor-expressing tumour volume (SRETVwb) on positron emission tomography/computed tomography (PET/CT) images is highly time-consuming but has shown value as an independent prognostic factor for survival. An automatic method to measure SRETVwb could improve disease status assessment and provide a tool for prognostication. This study aimed to develop an artificial intelligence (AI)-based method to detect and quantify SRETVwb and total lesion somatostatin receptor expression (TLSREwb) from [
68Ga]Ga-DOTA-TOC/TATE PET/CT images.
METHODS: A UNet3D convolutional neural network (CNN) was used to train an AI model with [
68Ga]Ga-DOTA-TOC/TATE PET/CT images, where all tumours were manually segmented with a semi-automatic method. The training set consisted of 148 patients, of which 108 had PET-positive tumours. The test group consisted of 30 patients, of which 25 had PET-positive tumours. Two physicians segmented tumours in the test group for comparison with the AI model.
RESULTS: There were good correlations between the segmented SRETVwb and TLSREwb by the AI model and the physicians, with Spearman rank correlation coefficients of r = 0.78 and r = 0.73, respectively, for SRETVwb and r = 0.83 and r = 0.81, respectively, for TLSREwb. The sensitivity on a lesion detection level was 80% and 79%, and the positive predictive value was 83% and 84% when comparing the AI model with the two physicians.
CONCLUSION: It was possible to develop an AI model to segment SRETVwb and TLSREwb with high performance. A fully automated method makes quantification of tumour burden achievable and has the potential to be more widely used when assessing PET/CT images.
(Less)
- author
- Gålne, Anni LU ; Enqvist, Olof LU ; Sundlöv, Anna LU ; Valind, Kristian LU ; Minarik, David LU and Trägårdh, Elin LU
- organization
- publishing date
- 2023-08-07
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- AI, Somatostatin receptor-expressing tumour volume, [68 Ga]Ga-DOTA- TATE, [ 68 Ga]Ga-DOTA-TOC, PET/CT
- in
- European journal of hybrid imaging
- volume
- 7
- issue
- 1
- article number
- 14
- pages
- 16 pages
- publisher
- Springer
- external identifiers
-
- scopus:85167517530
- pmid:37544941
- ISSN
- 2510-3636
- DOI
- 10.1186/s41824-023-00172-7
- project
- PET-CT imaging of neuroendocrine tumors - Beyond diagnostics
- language
- English
- LU publication?
- yes
- additional info
- © 2023. The Author(s).
- id
- b24fc026-5f2d-4346-9bb4-bfeee6a280be
- date added to LUP
- 2023-08-10 21:39:03
- date last changed
- 2024-07-13 03:42:25
@article{b24fc026-5f2d-4346-9bb4-bfeee6a280be, abstract = {{<p>BACKGROUND: Segmenting the whole-body somatostatin receptor-expressing tumour volume (SRETVwb) on positron emission tomography/computed tomography (PET/CT) images is highly time-consuming but has shown value as an independent prognostic factor for survival. An automatic method to measure SRETVwb could improve disease status assessment and provide a tool for prognostication. This study aimed to develop an artificial intelligence (AI)-based method to detect and quantify SRETVwb and total lesion somatostatin receptor expression (TLSREwb) from [<br> 68Ga]Ga-DOTA-TOC/TATE PET/CT images.<br> </p><p>METHODS: A UNet3D convolutional neural network (CNN) was used to train an AI model with [<br> 68Ga]Ga-DOTA-TOC/TATE PET/CT images, where all tumours were manually segmented with a semi-automatic method. The training set consisted of 148 patients, of which 108 had PET-positive tumours. The test group consisted of 30 patients, of which 25 had PET-positive tumours. Two physicians segmented tumours in the test group for comparison with the AI model.<br> </p><p>RESULTS: There were good correlations between the segmented SRETVwb and TLSREwb by the AI model and the physicians, with Spearman rank correlation coefficients of r = 0.78 and r = 0.73, respectively, for SRETVwb and r = 0.83 and r = 0.81, respectively, for TLSREwb. The sensitivity on a lesion detection level was 80% and 79%, and the positive predictive value was 83% and 84% when comparing the AI model with the two physicians.</p><p>CONCLUSION: It was possible to develop an AI model to segment SRETVwb and TLSREwb with high performance. A fully automated method makes quantification of tumour burden achievable and has the potential to be more widely used when assessing PET/CT images.</p>}}, author = {{Gålne, Anni and Enqvist, Olof and Sundlöv, Anna and Valind, Kristian and Minarik, David and Trägårdh, Elin}}, issn = {{2510-3636}}, keywords = {{AI; Somatostatin receptor-expressing tumour volume; [68 Ga]Ga-DOTA- TATE; [ 68 Ga]Ga-DOTA-TOC; PET/CT}}, language = {{eng}}, month = {{08}}, number = {{1}}, publisher = {{Springer}}, series = {{European journal of hybrid imaging}}, title = {{AI-based quantification of whole-body tumour burden on somatostatin receptor PET/CT}}, url = {{http://dx.doi.org/10.1186/s41824-023-00172-7}}, doi = {{10.1186/s41824-023-00172-7}}, volume = {{7}}, year = {{2023}}, }