Dataset Information

Image enhancement of whole-body oncology [¹⁸F]-FDG PET scans using deep neural networks to reduce noise.

ABSTRACT:

Purpose

To enhance the image quality of oncology [¹⁸F]-FDG PET scans acquired in shorter times and reconstructed by faster algorithms using deep neural networks.

Methods

List-mode data from 277 [¹⁸F]-FDG PET/CT scans, from six centres using GE Discovery PET/CT scanners, were split into ¾-, ½- and ¼-duration scans. Full-duration datasets were reconstructed using the convergent block sequential regularised expectation maximisation (BSREM) algorithm. Short-duration datasets were reconstructed with the faster OSEM algorithm. The 277 examinations were divided into training (n = 237), validation (n = 15) and testing (n = 25) sets. Three deep learning enhancement (DLE) models were trained to map full and partial-duration OSEM images into their target full-duration BSREM images. In addition to standardised uptake value (SUV) evaluations in lesions, liver and lungs, two experienced radiologists scored the quality of testing set images and BSREM in a blinded clinical reading (175 series).

Results

OSEM reconstructions demonstrated up to 22% difference in lesion SUV_max, for different scan durations, compared to full-duration BSREM. Application of the DLE models reduced this difference significantly for full-, ¾- and ½-duration scans, while simultaneously reducing the noise in the liver. The clinical reading showed that the standard DLE model with full- or ¾-duration scans provided an image quality substantially comparable to full-duration scans with BSREM reconstruction, yet in a shorter reconstruction time.

Conclusion

Deep learning-based image enhancement models may allow a reduction in scan time (or injected activity) by up to 50%, and can decrease reconstruction time to a third, while maintaining image quality.

SUBMITTER: Mehranian A

PROVIDER: S-EPMC8803788 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:The Centers for Medicare and Medicaid Services coverage includes 3 posttherapy 18F-FDG PET/CT scans per patient and per tumor type. Any additional follow-up 18F-FDG PET/CT scans will be reimbursed at the discretion of a local Medicare administrator, if deemed medically necessary. This study aimed to investigate common clinical indications for performing a fourth or additional follow-up 18F-FDG PET/CT scans that could affect the management of patients. Methods: This was a retrospective institutional review of 433 oncology patients (203 men; mean age, 55 y), including a total of 1,659 fourth or subsequent follow-up PET/CT scans after completion of primary treatment. Twelve indications for performing a fourth or subsequent follow-up PET/CT scan were determined, and the impact of each of the 12 indications on patients' management was evaluated. Results: The primary tumors were breast cancer (92 patients, 426 scans), non-Hodgkin lymphoma (77 patients, 208 scans), Hodgkin disease (41 patients, 182 scans), colorectal cancer (70 patients, 286 scans), melanoma (69 patients, 271 scans), and lung cancer (84 patients, 286 scans). The indications were categorized in 4 groups: PET/CT for diagnosis of tumor recurrence (303/1,659, 18.3%), PET/CT before starting therapy for tumor recurrence (64/1,659, 3.9%), PET/CT to assess therapy response for tumor recurrence (507/1,659, 30.6%), and follow-up PET/CT after completion of treatment for tumor recurrence (785/1,659, 47.3%). Overall, fourth and subsequent follow-up 18F-FDG PET/CT scans resulted in change in management in 31.6% of the scans (356 of 1,128) when the scans were obtained for medical necessities (indications 1-11), and in 5.6% of the scans (30/531) when the scans were obtained without any medical necessity (indication 12). Conclusion: The fourth and subsequent PET/CT scans obtained after completion of primary treatment led to a change in management in 31.6% of the scans when acquired for appropriate clinical reasons. Performing follow-up PET/CT without appropriate medical reason had a low impact on patients' management and should be avoided.

Project description:Knowledge of the within-subject variability of 18F-FDG PET/MRI measurements is necessary for proper interpretation of quantitative PET or MRI metrics in the context of therapeutic efficacy assessments with integrated PET/MRI scanners. The goal of this study was to determine the test-retest repeatability of these metrics on PET/MRI, with comparison to similar metrics acquired by PET/CT. Methods: This prospective study enrolled subjects with pathology-proven pelvic malignancies. Baseline imaging consisted of PET/CT immediately followed by PET/MRI, using a single 370-MBq 18F-FDG dose. Repeat imaging was performed within 7 d using an identical imaging protocol, with no oncologic therapy between sessions. PET imaging on both scanners consisted of a list-mode acquisition at a single pelvic station. The MRI consisted of 2-point Dixon imaging for attenuation correction, standard sequences for anatomic correlation, and diffusion-weighted imaging. PET data were statically reconstructed using various frame durations and minimizing uptake time differences between sessions. SUV metrics were extracted for both PET/CT and PET/MRI in each imaging session. Apparent diffusion coefficient (ADC) metrics were extracted for both PET/MRI sessions. Results: The study cohort consisted of 14 subjects (13 female, 1 male) with various pelvic cancers (11 cervical, 2 rectal, 1 endometrial). For SUVmax, the within-subject coefficient of variation (wCV) appeared higher for PET/CT (8.5%-12.8%) than PET/MRI (6.6%-8.7%) across all PET reconstructions, though with no significant repeatability differences (all P values ≥ 0.08) between modalities. For lean body mass-adjusted SUVpeak, the wCVs appeared similar for PET/CT (9.9%-11.5%) and PET/MRI (9.2%-11.3%) across all PET reconstructions, again with no significant repeatability differences (all P values ≥ 0.14) between modalities. For PET/MRI, the wCV for ADCmedian of 3.5% appeared lower than the wCVs for SUVmax (6.6%-8.7%) and SULpeak (9.2%-11.3%), though without significant repeatability differences (all P values ≥ 0.23). Conclusion: For solid tumors of the pelvis, the repeatability of the evaluated SUV and ADC metrics on 18F-FDG PET/MRI is both acceptably high and similar to previously published values for 18F-FDG PET/CT and MRI, supporting the use of 18F-FDG PET/MRI for quantitative oncologic treatment response assessments.

Project description:IntroductionTumor hypoxia and angiogenesis are implicated in tumor growth and metastases, and anti-angiogenic therapies have an important role in treating patients with metastatic colorectal cancer. However, the prevalence of hypoxia has not been fully evaluated in colorectal liver metastases, and hypoxic response to anti-angiogenic therapy has not been clearly established. The aims of the study were to evaluate the changes seen on 18F-FMISO and 18F-FDG PET scans in patients treated with anti-angiogenic therapy, and to correlate these measures of hypoxia and metabolism with clinical outcomes, and blood biomarkers of angiogenesis.MethodsPatients with metastatic colorectal carcinoma planned for treatment with bevacizumab and chemotherapy received routine staging investigations prior to any treatment, including a FDG PET scan. A FMISO PET scan was performed within 4 weeks of staging tests, with blood specimens collected at that time for serum VEGF and osteopontin measurement. Follow-up FDG and FMISO scans were performed after 1 cycle of treatment. Results were compared to response (RECIST), progression free survival (PFS), and overall survival (OS).ResultsA total of 15 patients were recruited into this prospective trial, of which 13 patients were evaluable for assessment of treatment follow-up. Baseline FDG uptake was higher than FMISO uptake, and there was a significant decrease in FDG uptake (SUVmax and TGV) but not FMISO uptake (SUVmax and TNR) after treatment. There was a positive correlation between FDG and FMISO SUVmax on both baseline and post-treatment PET scans. Blood biomarkers of serum VEGF and osteopontin were significantly correlated with the FDG and FMISO PET parameters.ConclusionsThis study shows that hypoxia in metastatic colorectal cancer, assessed by FMISO PET, shows minor changes following initial treatment with anti-angiogenic therapy, but is associated with therapeutic response. FDG PET uptake changes (SUVmax, TLG) are also associated with response to anti-angiogenic therapy. These findings demonstrate the interplay between tumor metabolism and hypoxic regulation following anti-angiogenic treatment of metastatic colorectal cancer.

Project description:Active surveillance for patients with esophageal cancer and a clinically complete response (cCR) after neoadjuvant chemoradiotherapy (nCRT) is being studied. Active surveillance requires accurate clinical response evaluations. 18F-FDG PET/CT might be able to detect local tumor recurrence after nCRT as soon as the esophagus recovers from radiation-induced esophagitis. The aims of this study were to assess the value of serial 18F-FDG PET/CT scans for detecting local recurrence in patients beyond 3 mo after nCRT and to determine when radiation-induced esophagitis has resolved. Methods: This retrospective multicenter study included patients who had cCR after nCRT, who initially declined surgery, and who subsequently underwent active surveillance. Clinical response evaluations included 18F-FDG PET/CT, endoscopic biopsies, and endoscopic ultrasound with fine-needle aspiration at regular intervals. SUVmax normalized for lean body mass (SULmax) was measured at the primary tumor site. The percentage change in SULmax (Δ%SULmax) between the last follow-up scan and the scan at 3 mo after nCRT was calculated. Tumor recurrence was defined as biopsy-proven vital tumor at the initial tumor site. Results: Of 41 eligible patients, 24 patients had recurrent disease at a median of 6.5 mo after nCRT and 17 patients remained cancer free during a median follow-up of 24 mo after nCRT. Five of 24 patients with tumor recurrence had sudden intense SULmax increases of greater than 180%. In 19 of 24 patients with tumor recurrence, SULmax gradually increased (median Δ%SULmax, +18%), whereas SULmax decreased (median Δ%SULmax, -12%) in patients with ongoing cCR (P < 0.001, independent-samples t test). In patients with ongoing cCR, SULmax was lowest at 11 mo after nCRT. Conclusion: Serial 18F-FDG PET/CT might be a useful tool for detecting tumor recurrence during active surveillance. In patients with ongoing cCR, the lowest SULmax was reached at 11 mo after nCRT, suggesting that radiation-induced esophagitis had mostly resolved by that time. These findings warrant further evaluation in a larger cohort.

Project description:BackgroundWe assessed and compared image quality obtained with clinical 18F-FDG whole-body oncologic PET protocols used in three different, state-of-the-art digital PET/CT and two conventional PMT-based PET/CT devices. Our goal was to evaluate an improved trade-off between administered activity (patient dose exposure/signal-to-noise ratio) and acquisition time (patient comfort) while preserving diagnostic information achievable with the recently introduced digital detector technology compared to previous analogue PET technology.MethodsWe performed list-mode (LM) PET acquisitions using a NEMA/IEC NU2 phantom, with activity concentrations of 5?kBq/mL and 25?kBq/mL for the background (9.5?L) and sphere inserts, respectively. For each device, reconstructions were obtained varying the image statistics (10, 30, 60, 90, 120, 180, and 300?s from LM data) and the number of iterations (range 1 to 10) in addition to the employed local clinical protocol setup. We measured for each reconstructed dataset: the quantitative cross-calibration, the image noise on the uniform background assessed by the coefficient of variation (COV), and the recovery coefficients (RCs) evaluated in the hot spheres. Additionally, we compared the characteristic time-activity-product (TAP) that is the product of scan time per bed position × mass-activity administered (in min·MBq/kg) across datasets.ResultsGood system cross-calibration was obtained for all tested datasets with <?6% deviation from the expected value was observed. For all clinical protocol settings, image noise was compatible with clinical interpretation (COV?<?15%). Digital PET showed an improved background signal-to-noise ratio as compared to conventional PMT-based PET. RCs were comparable between digital and PMT-based PET datasets. Compared to PMT-based PET, digital systems provided comparable image quality with lower TAP (from ~?40% less and up to 70% less).ConclusionsThis study compared the achievable clinical image quality in three state-of-the-art digital PET/CT devices (from different vendors) as well as in two conventional PMT-based PET. Reported results show that a comparable image quality is achievable with a TAP reduction of ~?40% in digital PET. This could lead to a significant reduction of the administered mass-activity and/or scan time with direct benefits in terms of dose exposure and patient comfort.

Dataset Information

Image enhancement of whole-body oncology [¹⁸F]-FDG PET scans using deep neural networks to reduce noise.

Purpose

Methods

Results

Conclusion

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets

Dataset Information

Image enhancement of whole-body oncology [18F]-FDG PET scans using deep neural networks to reduce noise.

Purpose

Methods

Results

Conclusion

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets

Image enhancement of whole-body oncology [¹⁸F]-FDG PET scans using deep neural networks to reduce noise.