| 1. |
- Kaboteh, Reza, et al.
(author)
-
Convolutional neural network based quantification of choline uptake in PET/CT studies is associated with overall survival in patients with prostate cancer
- 2017
-
In: European Journal of Nuclear Medicine and Molecular Imaging. - 1619-7070 .- 1619-7089. ; 44:supplement 2
-
Journal article (peer-reviewed)abstract
- Aim : To develop a convolutional neural network (CNN) based automated method for quantification of 18F-choline uptake in the prostate gland in PET/CT studies and to study the association between this measure, clinical data and overall survival in patients with prostate cancer. Methods : A CNN was trained to segment the prostate gland in CT images using manual segmentations performed by a radiologist in a group of 100 patients, who had undergone 18F-FDG PET/CT. After the training process, the CNN automatically segmented the prostate gland in the CT images and SUV values in the corresponding PET images were automatically analyzed in a separate validation group consisting of 45 patients with biopsy-proven hormone-naïve prostate cancer. All patients had undergone an 18F-choline PET/CT as part of a previous research project. Voxels localized in the prostate gland and having a SUV >2.65 were defined as abnormal, as proposed by Reske S et al. (2006). Automated calculation of the following five PET measurements was performed: maximal SUV within the prostate gland - SUVmax; average SUV for voxels with SUV >2.65 - SUVmean; volume of voxels with SUV >2.65 - VOL; fraction of VOL related to the whole volume of the prostate gland - FRAC; product SUVmean x FRAC defined as Total Lesion Uptake - TLU. The association between the automated PET measurements, age, PSA, Gleason score and overall survival (OS) was evaluated using a univariate Cox proportional hazards regression model. Kaplan-Meier analysis was used to estimate the survival difference (log-rank test). Results : TLU and FRAC were significantly associated with OS in the Cox analysis while the other three PET measurements; age, PSA and Gleason score were not. Kaplan-Meier analysis showed that patients with SUVmax <5.3, SUVmean <3.5 and TLU <1 showed significantly longer survival times than patients with values higher than these thresholds. No significant differences were found when patients were stratified based on the other two PET measurements, PSA or Gleason score. Conclusion : Measurements reflecting 18F-choline PET uptake in the prostate gland obtained using a completely automated method were significantly associated with OS in patients with hormone-naïve prostate cancer. This type of objective quantification of PET/CT studies could be of value also for other PET tracers and other cancers in the future.
|
|
| 2. |
- Lind, Erica, et al.
(author)
-
Automated quantification of reference levels in liver and mediastinum (blood pool) for the Deauville therapy response classification using FDG-PET/CT in lymphoma patients
- 2017
-
In: European Journal of Nuclear Medicine and Molecular Imaging. - 1619-7070 .- 1619-7089. ; 44:supplement 2
-
Journal article (peer-reviewed)abstract
- Aim : To develop and validate a convolutional neural network (CNN) based method for automated quantification of reference levels in liver and mediastinum (blood pool) for the Deauville therapy response classification using FDG-PET/CT in lymphoma patients. Methods : CNNs were trained to segment the liver and the mediastinum, defined as the thoracic part of the aorta, in CT images from 81 consecutive lymphoma patients, who had undergone FDG-PET/CT examinations. Trained image readers segmented the liver and aorta manually in each of the CT images and these segmentations together with the CT images were used to train the CNN. After the training process, the CNN method was applied to a separate validation group consisting of six consecutive lymphoma patients (17-82 years, 3 female). First, the liver and mediastinum were automatically segmented in the CT images. Second, voxels in the corresponding FDG-PET images, which were localized in the liver and mediastinum, were selected and the median standard uptake value (SUV) was calculated. The CNN based analysis was compared to corresponding manual segmentations by two experienced radiologists. The Dice index was used to analyse the overlap between the segmentations by the CNN and the two radiologists. A Dice index of 1.00 indicates perfect matching. Results : The mean Dice indices for the comparison between CNN based liver segmentations and those of the two radiologists in the validation group were 0.95 and 0.95. A corresponding comparison between the two radiologists also resulted in a Dice index of 0.95. The mean liver volumes were 1,752ml, 1,757ml and 1,768ml for the CNN and two radiologists, respectively. The median SUV for the liver was on average 1.8 and the differences between median SUV based on CNN and manual segmentations were less or equal to 0.1. The mean Dice indices for the mediastinum were 0.80, 0.83 (CNN vs radiologists) and 0.86 (comparing the two radiologists). The mean mediastinum (aorta) volumes were 147ml, 140ml and 125ml for the CNN and two radiologists, respectively. The median SUV for the mediastinum was on average 1.4 and the differences between median SUV based on CNN and manual segmentations were less or equal to 0.14. Conclusion : A CNN based method for automated quantification of reference levels in liver and mediastinum show good agreement with results obtained by experienced radiologists, who manually segmented the CT images. This is a first and promising step towards a completely objective treatment response evaluation in patients with lymphoma based on FDG-PET/CT.
|
|
| 3. |
- Sadik, May, 1970, et al.
(author)
-
3D prostate gland uptake of 18F-choline - association with overall survival in patients with hormone-naïve prostate cancer
- 2017
-
In: Journal of Nuclear Medicine. - 0161-5505 .- 2159-662X. ; 58:supplement 1, s. 544-
-
Journal article (peer-reviewed)abstract
- Objectives : To develop a completely automated method to quantify prostate gland uptake of 18F-choline in PET/CT images and to study the relationship between this measure, clinical data and overall survival in patients with prostate cancer. Methods : An automated method for segmentation of the prostate gland in CT images was developed using a training group of 100 patients who had undergone PET/CT scanning. The algorithms were trained based on the manual segmentations of the prostate gland in the 100 CT scans performed by a single radiologist. A multi-atlas-based method was used applied for automated segmentation of the prostate gland. Each of a subset of the training images was registered separately to the test image. By applying the resulting transformations to the manual delineations a rough segmentation of the test image was obtained. This segmentation was refined using a random-forest classifier and the final segmentation was obtained with graph cuts. Voxels in the 18F-choline PET scans having a standard uptake value (SUV) >2.65 and localized in the prostate gland in the corresponding CT scan were defined as abnormal. Automated calculation of the following five PET measurements was performed: The maximal SUV within the prostate gland - SUVmax The average SUV within the abnormal part of the prostate gland - SUVmean The volume of abnormal uptake within the prostate gland - VOL The product SUVmean x VOL defined as Total Lesion Uptake - TLU The fraction of the prostate with abnormal uptake related to the whole volume of the prostate gland - FRAC The automated quantification method was retrospectively applied to a separate test group of 46 prostate cancer patients, aged 53-94 years, who had undergone 18F-choline PET/CT. These patients have previously been selected for a study aiming to compare whole-body bone scans, 18F-choline-PET/CT and 18F-NaF PET/CT with magnetic resonance imaging. The study entry criteria were biopsy-proven prostate cancer, a positive whole-body bone scan consistent with bone metastases, and no history of androgen deprivation. The association between the automated PET measurements, age, PSA, Gleason score and overall survival was evaluated using a univariate Cox proportional hazards regression model. Kaplan-Meier estimates were used to estimate the survival difference between patients with values above and below the median value for all variables analyzed. Results : The fraction of the prostate with abnormal uptake related to the whole volume of the prostate gland - FRAC and age were significantly associated with overall survival (Table) while PSA, Gleason score and other PET measurements were not. The patients with a FRAC above the median value (58.2%) had a significantly shorter median survival time than patients with a value below the median value (2.8 years vs. 5.5 years; p=0.04), see Figure. Conclusion : A completely automated method of quantifying 18F-choline PET uptake in the prostate gland yielded a measure of disease extent that was significantly associated with overall survival in patients with hormone-naïve prostate cancer. The method can also be applied to PET/CT scans with other tracers such as FDG or PSMA-targeted agents. It is our hope that these preliminary data will inspire further evaluation of this type of objective quantification of PET/CT scans.
|
|
| 4. |
- Sadik, May, 1970, et al.
(author)
-
Analytical validation of an automated method for segmentation of the prostate gland in CT images
- 2017
-
In: European Journal of Nuclear Medicine and Molecular Imaging. - : Springer Science and Business Media LLC. - 1619-7070 .- 1619-7089. ; 44:supplement issue 2
-
Journal article (peer-reviewed)abstract
- Aim : Uptake of PET tracers in the prostate gland may serve as guidance for management of patients with prostate cancer. PET studies alone do, however, not allow for accurate segmentation of the gland, instead the corresponding CT images contain the required anatomical information. Our long-term aim is to develop an objectively measured PET/CT imaging biomarker reflecting PET tracer uptake. In this study we take the first step and develop and validate a completely automated method for 3D-segmentation of the prostate gland in CT images. Methods : A convolutional neural network (CNN) was trained to segment the prostate gland in CT images using manual segmentations performed by a radiologist in a group of 100 patients, who had undergone 18F-FDG PET/CT. After the training process, the CNN automatically segmented the prostate gland in CT images in a separate validation group consisting of 45 patients with prostate cancer. All patients had undergone a 18F-choline PET/CT as part of a previous research project. The CNN segmentations were compared to manual segmentations performed independently by two radiologists. The volume of the prostate gland was calculated based on segmentations by the CNN and radiologists. The Sørensen-Dice index was used to analyse the overlap between the segmentations by the CNN and the two radiologists. Results : The prostate volumes were on average 79mL (range 9-212mL) in the 45 patients, measured as mean volumes for the two radiologists. The mean difference in prostate volumes between the two radiologists was 14mL (SD 29mL). The mean volume difference between the CNN segmentation and the mean values from the two radiologists was 22mL (SD 43mL). For the subgroup of patients with prostate volumes <100 mL (n=36), the difference between the radiologists was 9mL (SD 11mL) compared to difference CNN vs radiologists of 7mL (SD 15mL). The Sørensen-Dice index was 0.69 and 0.70 for the comparison between CNN segmentation and the two radiologists, respectively and 0.83 for the comparison between the two radiologists. The corresponding Sørensen-Dice index in the 36 patients with volumes <100 mL were 0.74, 0.75 and 0.83, respectively Conclusion : Our CNN based method for automated segmentation of the prostate gland in CT images show good agreement with the corresponding manual segmentations by two radiologists especially for prostade glands with a volume less than 100 mL.
|
|
| 5. |
- Sadik, May, 1970, et al.
(author)
-
Automated 3D segmentation of the prostate gland in CT images - a first step towards objective measurements of prostate uptake in PET and SPECT images
- 2017
-
In: Journal of Nuclear Medicine. - 0161-5505 .- 2159-662X. ; 58:supplement 1
-
Journal article (peer-reviewed)abstract
- Objectives : Uptake of PSMA-targeted tracers and choline in the prostate gland may serve as guidance for management of patients with prostate cancer. Our aim was to develop objectively measured PET/CT and SPECT/CT imaging biomarkers reflecting such uptake. In this study we took the first step by introducing and validating a completely automated algorithm for 3D-segmentation of the prostate gland in CT images. Methods : A group of 100 patients who had undergone 18F-FDG PET/CT scanning was used as training set. A single radiologist performed manual segmentations of the prostate gland in all 100 CT scans using a custom software tool. A multi-atlas-based method was used applied for automated segmentation of the prostate gland. Each of a subset of the training images was registered separately to the test image. By applying the resulting transformations to the manual delineations a rough segmentation of the test image was obtained. This segmentation was refined using a random-forest classifier and the final segmentation was obtained with graph cuts. A separate validation group comprised 46 patients (aged 53-94 years) with biopsy-proven prostate cancer, who had undergone both 18F-fluoromethylcholine PET/CT and 18F-sodiumfluoride PET/CT within a time frame of 3 weeks as part of a previous research project. A diagnostic contrast-enhanced CT scan (64-slice helical, 120 kV, ’smart mA’ maximum 400 mA) was obtained with a CT slice thickness of 3.75 mm. We speculated that the volume of the prostate gland and in particular the fraction of the gland that had abnormally high tracer accumulation, might be useful biomarkers helping to improve management and prognostication in cancer patients. The reproducibility of automated measurements of the prostate gland volume was therefore studied using the two CT scans from each patient in the validation set. Results : The automatically measured prostate gland volumes in the validation set ranged between 13 ml and 90 ml with a mean of 48 ml. The mean difference between the two volume measurements in each patient was 2.4 ml with an SD of 6.6 ml. The difference was less than 10 ml in 41 of the 46 cases. Conclusion : We have demonstrated a reproducible and automated algorithm for 3D-segmentation of the prostate gland in CT images. This is a first step towards objective measurements of prostate gland tracer uptake in PET and SPECT examinations, because PET and SPECT images alone do not allow for accurate segmentation of the prostate gland, which instead depends on proper segmentation based on the corresponding CT scans.
|
|
| 6. |
- Sadik, May, 1970, et al.
(author)
-
Automated quantification of reference levels in liver and mediastinal blood pool for the Deauville therapy response classification using FDG-PET/CT in Hodgkin and non-Hodgkin lymphomas
- 2019
-
In: Clinical Physiology and Functional Imaging. - : Wiley. - 1475-0961 .- 1475-097X. ; 39:1, s. 78-84
-
Journal article (peer-reviewed)abstract
- Background 18F-FDG-PET/CT has become a standard for assessing treatment response in patients with lymphoma. A subjective interpretation of the scan based on the Deauville 5-point scale has been widely adopted. However, inter-observer variability due to the subjectivity of the interpretation is a limitation. Our main goal is to develop an objective and automated method for evaluating response. The first step is to develop and validate an artificial intelligence (AI)-based method, for the automated quantification of reference levels in the liver and mediastinal blood pool in patients with lymphoma. Methods Results The AI-based method was trained to segment the liver and the mediastinal blood pool in CT images from 80 lymphoma patients, who had undergone 18F-FDG-PET/CT, and apply this to a validation group of six lymphoma patients. CT segmentations were transferred to the PET images to obtain automatic standardized uptake values (SUV). The AI-based analysis was compared to corresponding manual segmentations performed by two radiologists. The mean difference for the comparison between the AI-based liver SUV quantifications and those of the two radiologists in the validation group was 0 center dot 02 and 0 center dot 02, respectively, and 0 center dot 02 and 0 center dot 02 for mediastinal blood pool respectively. Conclusions An AI-based method for the automated quantification of reference levels in the liver and mediastinal blood pool shows good agreement with results obtained by experienced radiologists who had manually segmented the CT images. This is a first, promising step towards objective treatment response evaluation in patients with lymphoma based on 18F-FDG-PET/CT.
|
|
| 7. |
- Sadik, May, 1970, et al.
(author)
-
Variability in reference levels for Deauville classifications applied to lymphoma patients examined with 18F-FDG-PET/CT
- 2017
-
In: European Journal of Nuclear Medicine and Molecular Imaging. - 1619-7070 .- 1619-7089. ; 44
-
Journal article (peer-reviewed)abstract
- Aim : To develop and validate a convolutional neural network (CNN) based method for automated quantification of reference levels in liver and mediastinum (blood pool) for the Deauville therapy response classification using FDG-PET/CT in lymphoma patients. Methods : CNNs were trained to segment the liver and the mediastinum, defined as the thoracic part of the aorta, in CT images from 81 consecutive lymphoma patients, who had undergone FDG-PET/CT examinations. Trained image readers segmented the liver and aorta manually in each of the CT images and these segmentations together with the CT images were used to train the CNN. After the training process, the CNN method was applied to a separate validation group consisting of six consecutive lymphoma patients (17-82 years, 3 female). First, the liver and mediastinum were automatically segmented in the CT images. Second, voxels in the corresponding FDG-PET images, which were localized in the liver and mediastinum, were selected and the median standard uptake value (SUV) was calculated. The CNN based analysis was compared to corresponding manual segmentations by two experienced radiologists. The Dice index was used to analyse the overlap between the segmentations by the CNN and the two radiologists. A Dice index of 1.00 indicates perfect matching. Results : The mean Dice indices for the comparison between CNN based liver segmentations and those of the two radiologists in the validation group were 0.95 and 0.95. A corresponding comparison between the two radiologists also resulted in a Dice index of 0.95. The mean liver volumes were 1,752ml, 1,757ml and 1,768ml for the CNN and two radiologists, respectively. The median SUV for the liver was on average 1.8 and the differences between median SUV based on CNN and manual segmentations were less or equal to 0.1. The mean Dice indices for the mediastinum were 0.80, 0.83 (CNN vs radiologists) and 0.86 (comparing the two radiologists). The mean mediastinum (aorta) volumes were 147ml, 140ml and 125ml for the CNN and two radiologists, respectively. The median SUV for the mediastinum was on average 1.4 and the differences between median SUV based on CNN and manual segmentations were less or equal to 0.14. Conclusion : A CNN based method for automated quantification of reference levels in liver and mediastinum show good agreement with results obtained by experienced radiologists, who manually segmented the CT images. This is a first and promising step towards a completely objective treatment response evaluation in patients with lymphoma based on FDG-PET/CT.
|
|
