Dataset Information

Artificial intelligence enables whole-body positron emission tomography scans with minimal radiation exposure.

ABSTRACT:

Purpose

To generate diagnostic ¹⁸F-FDG PET images of pediatric cancer patients from ultra-low-dose ¹⁸F-FDG PET input images, using a novel artificial intelligence (AI) algorithm.

Methods

We used whole-body ¹⁸F-FDG-PET/MRI scans of 33 children and young adults with lymphoma (3-30 years) to develop a convolutional neural network (CNN), which combines inputs from simulated 6.25% ultra-low-dose ¹⁸F-FDG PET scans and simultaneously acquired MRI scans to produce a standard-dose ¹⁸F-FDG PET scan. The image quality of ultra-low-dose PET scans, AI-augmented PET scans, and clinical standard PET scans was evaluated by traditional metrics in computer vision and by expert radiologists and nuclear medicine physicians, using Wilcoxon signed-rank tests and weighted kappa statistics.

Results

The peak signal-to-noise ratio and structural similarity index were significantly higher, and the normalized root-mean-square error was significantly lower on the AI-reconstructed PET images compared to simulated 6.25% dose images (p < 0.001). Compared to the ground-truth standard-dose PET, SUV_max values of tumors and reference tissues were significantly higher on the simulated 6.25% ultra-low-dose PET scans as a result of image noise. After the CNN augmentation, the SUV_max values were recovered to values similar to the standard-dose PET. Quantitative measures of the readers' diagnostic confidence demonstrated significantly higher agreement between standard clinical scans and AI-reconstructed PET scans (kappa = 0.942) than 6.25% dose scans (kappa = 0.650).

Conclusions

Our CNN model could generate simulated clinical standard ¹⁸F-FDG PET images from ultra-low-dose inputs, while maintaining clinically relevant information in terms of diagnostic accuracy and quantitative SUV measurements.

SUBMITTER: Wang YJ

PROVIDER: S-EPMC8266729 | biostudies-literature | 2021 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Artificial intelligence enables whole-body positron emission tomography scans with minimal radiation exposure.

Wang Yan-Ran Joyce YJ Baratto Lucia L Hawk K Elizabeth KE Theruvath Ashok J AJ Pribnow Allison A Thakor Avnesh S AS Gatidis Sergios S Lu Rong R Gummidipundi Santosh E SE Garcia-Diaz Jordi J Rubin Daniel D Daldrup-Link Heike E HE

European journal of nuclear medicine and molecular imaging 20210201 9

<h4>Purpose</h4>To generate diagnostic <sup>18</sup>F-FDG PET images of pediatric cancer patients from ultra-low-dose <sup>18</sup>F-FDG PET input images, using a novel artificial intelligence (AI) algorithm.<h4>Methods</h4>We used whole-body <sup>18</sup>F-FDG-PET/MRI scans of 33 children and young adults with lymphoma (3-30 years) to develop a convolutional neural network (CNN), which combines inputs from simulated 6.25% ultra-low-dose <sup>18</sup>F-FDG PET scans and simultaneously acquired M ...[more]

PMID: 33527176

Similar Datasets

Project description:BackgroundPatients are referred to functional coronary artery disease (CAD) testing based on their pre-test probability (PTP) to search for myocardial ischemia. The recommended prediction tools incorporate three variables (symptoms, age, sex) and are easy to use, but have a limited diagnostic accuracy. Hence, a substantial proportion of non-invasive functional tests reveal no myocardial ischemia, leading to unnecessary radiation exposure and costs. Therefore, preselection of patients before ischemia testing needs to be improved using a more predictive and personalised approach.AimsUsing multiple variables (symptoms, vitals, ECG, biomarkers), artificial intelligence-based tools can provide a detailed and individualised profile of each patient. This could improve PTP assessment and provide a more personalised diagnostic approach in the framework of predictive, preventive and personalised medicine (PPPM).MethodsConsecutive patients (n = 2417) referred for Rubidium-82 positron emission tomography were evaluated. PTP was calculated using the ESC 2013/2019 and ACC 2012/2021 guidelines, and a memetic pattern-based algorithm (MPA) was applied incorporating symptoms, vitals, ECG and biomarkers. Five PTP categories from very low to very high PTP were defined (i.e., < 5%, 5-15%, 15-50%, 50-85%, > 85%). Ischemia was defined as summed difference score (SDS) ≥ 2.ResultsIschemia was present in 37.1%. The MPA model was most accurate to predict ischemia (AUC: 0.758, p < 0.001 compared to ESC 2013, 0.661; ESC 2019, 0.673; ACC 2012, 0.585; ACC 2021, 0.667). Using the < 5% threshold, the MPA's sensitivity and negative predictive value to rule out ischemia were 99.1% and 96.4%, respectively. The model allocated patients more evenly across PTP categories, reduced the proportion of patients in the intermediate (15-85%) range by 29% (ACC 2012)-51% (ESC 2019), and was the only tool to correctly predict ischemia prevalence in the very low PTP category.ConclusionThe MPA model enhanced ischemia testing according to the PPPM framework:The MPA model improved individual prediction of ischemia significantly and could safely exclude ischemia based on readily available variables without advanced testing ("predictive").It reduced the proportion of patients in the intermediate PTP range. Therefore, it could be used as a gatekeeper to prevent patients from further unnecessary downstream testing, radiation exposure and costs ("preventive").Consequently, the MPA model could transform ischemia testing towards a more personalised diagnostic algorithm ("personalised").Supplementary informationThe online version contains supplementary material available at 10.1007/s13167-023-00341-5.

Project description:Fluorine-18-fluoro-2-deoxy-D-glucose (FDG) is widely used as positron-emission-tomography (PET) radiotracer for the detection and staging of human cancer. Tumor uptake of FDG varies substantially between different cancer types and between patients with the same tumor type. The molecular basis for this heterogeneity is unknown. Using cancer cell lines and primary human tumors of distinct histologic origins, we here show that increased FDG uptake is universally associated with coordinate upregulation of genes within the glycolysis, pentose-phosphate, and other related metabolic pathways. In primary human breast cancers, this FDG signature shows significant overlap with established breast cancer signatures for the “basal-like” disease subtype and “poor prognosis”. FDG high breast cancer showed significantly more gene copy number alterations genome wide than FDG low cancers. About 50 % of primary breast cancers with high FDG uptake and FDG gene expression signature show DNA copy gain encompassing the c-myc gene locus and express gene sets regulated by the transcription factor MYC. Our data shows that FDG-PET marks a distinct subset of “basal-like” human breast cancer which is characterized by MYC and prognostically unfavorable gene expression signatures, suggesting that FDG-PET imaging may be useful to risk-stratify patients with locally advanced breast cancer. Our general strategy to identify molecular determinants of FDG-retention through a genome-wide approach consisted of FDG uptake measurements in cancer cell lines and primary human tumors, the selection of samples with particularly high versus low FDG-retention, and subsequent pathway-based analysis of RNA expression data collected from these samples. Cell line RNA was extracted from a 10 cm plate seeded simultaneously and at the same density as the wells for FDG-uptake assays. For primary human tumor samples, RNA was extracted from macrodissected frozen tumor tissue. All cell line RNA samples and the astrocytoma RNA aliquots were hybridized to Affymetrix U133 Plus 2.0 arrays. All primary breast cancer RNA samples were hybridized to Affymetrix U133A arrays.

Project description:Background[18F]F-FDG, [68Ga]Ga-PSMA-11, and [68Ga]Ga-FAPI-04 have achieved good results in multiple clinical trials and clinical practice, but the imaging of these tracers is limited to traditional short-axis positron emission tomography/computed tomography (PET/CT). Therefore, we aimed to use total-body PET/CT dynamic scanning to describe whole-body biodistribution of these three tracers and to calculate more precise radiation doses.MethodsTotal-body PET/CT (uExplorer, United Imaging Healthcare) dynamic scanning was performed on 54 patients, including 30 patients with [18F]F-FDG, 10 patients with [68Ga]Ga-PSMA-11, and 14 patients with [68Ga]Ga-FAPI-04. A 60-minute dynamic scanning of whole body was performed simultaneously after bedside bolus injection of the corresponding tracers. The dynamic sequence of 92 frames was quantitatively analyzed by the Pmod4.0 software. Whole body biodistribution was calculated as time-activity curves (TACs) describing dynamic uptake patterns in the subject's major organs, followed by calculation of tracer kinetics and cumulative organ activity. Finally, combined with the OLINDA/EXM software, effective doses of the different tracers and individual organ doses were calculated.ResultsIn a systematic TAC analysis of three tracers, we identified distinct biodistribution patterns in major organs. [68Ga]Ga-PSMA-11 showed a trend of rapid increasing and slow decreasing in liver, spleen, muscle, and bone. In the heart, stomach, brain, and lung, tracer decreased rapidly after rapid increasing. Similarly, tracer uptake in the kidney and urinary bladder increased gradually. [68Ga]Ga-FAPI-04 showed a rapid increasing and rapid decreasing trend in brain, lung, liver, spleen, bone, heart, kidney, and stomach. The mean effective dose of [68Ga]Ga-PSMA-11 was 1.47E-02 mSv/MBq, and the mean effective doses of [18F]F-FDG and [68Ga]Ga-FAPI-04 were comparable (2.52E-02 mSv/MBq and 2.23E-02 mSv/MBq). The mean effective dose of [18F]F-FDG was lower than that reported in the literature measured by previous short-axis PET, while both [68Ga]Ga-PSMA-11 and [68Ga]Ga-FAPI-04 had higher value than previously reported value.Conclusions[18F]F-FDG, [68Ga]Ga-PSMA-11 and [68Ga]Ga-FAPI-04 have good biodistribution in human organs. Real-time high-sensitivity dynamic scanning with total-body PET/CT is a very effective way to accurately calculate biodistribution and effective dose of positron-labeled radiopharmaceuticals.

Dataset Information

Artificial intelligence enables whole-body positron emission tomography scans with minimal radiation exposure.

Purpose

Methods

Results

Conclusions

Publications

Artificial intelligence enables whole-body positron emission tomography scans with minimal radiation exposure.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets