Robustness of¹⁸F-FDG PET Radiomic Features in Lung Cancer: Impact of Advanced Reconstruction Algorithm

Journal of Nuclear Medicine Technology, Journal Year: 2025, Volume and Issue: unknown, P. jnmt.124.268252 - jnmt.124.268252

Published: Feb. 5, 2025

¹⁸F-FDG PET radiomics is emerging as a promising tool to identify imaging biomarkers for quantifying intratumor heterogeneity in lung cancer. However, the robustness of radiomic features (RFs) influenced by factors such image reconstruction algorithms, tumor segmentation, and discretization. Although impact these on RFs has been explored, specific influence advanced block sequential regularized expectation maximization (BSREM) algorithm remains unclear. This study investigated potential variations associated with different when using BSREM. Methods: Retrospective data from 120 cancer patients were reconstructed twice BSREM conventional ordered-subset methods. For each set, 3 segmentations performed, including manual, 40% threshold, Nestle Two discretization methods absolute relative settings applied dataset before RF extraction. Stable robust assessed coefficient variance intraclass correlation coefficient, respectively. Results: High instability was exhibited 19%, 33%, 36% RFs, variation more than 20% reconstruction, discretization, Conversely, 60%, 35% demonstrated against factors, an 0.90. The comparative evaluation revealed significantly greater most subtypes under varying segmentation conditions (P < 0.05). Conclusion: stability are enhanced if rather method. Study results suggest that method could offer benefits providing consistent PET-based analysis improving diagnostic prognostic value.

Language: Английский

---

Deep learning image enhancement algorithms in PET/CT imaging: a phantom and sarcoma patient radiomic evaluation

European Journal of Nuclear Medicine and Molecular Imaging, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 27, 2025

PET/CT imaging data contains a wealth of quantitative information that can provide valuable contributions to characterising tumours. A growing body work focuses on the use deep-learning (DL) techniques for denoising PET data. These models are clinically evaluated prior use, however, image assessment provides potential further evaluation. This uses radiomic features compare two manufacturer enhancement algorithms, one which has been commercialised, against 'gold-standard' reconstruction in phantom and sarcoma patient set (N=20). All studies retrospective clinical [ 18 F]FDG dataset were acquired either GE Discovery 690 or 710 scanner with volumes segmented by an experienced nuclear medicine radiologist. The modular heterogeneous used this was filled F]FDG, five repeat acquisitions scanner. DL-enhanced images compared algorithms trained emulate input images. difference between sets tested significance 93 international biomarker standardisation initiative (IBSI) standardised features. Comparing 'gold-standard', 4.0% 9.7% measured significantly different (pcritical < 0.0005) respectively (averaged over DL algorithms). Larger differences observed comparing algorithm 29.8% 43.0% measuring found be similar generated using target method more than 80% not all comparisons across unseen result offers insight into performance demonstrate applications harmonisation radiomics evaluation algorithms.

Language: Английский

---

New Parametric 2D Curves for Modeling Prostate Shape in Magnetic Resonance Images

Symmetry, Journal Year: 2024, Volume and Issue: 16(6), P. 755 - 755

Published: June 17, 2024

Geometric shape models often help to extract specific contours in digital images (the segmentation process) with major precision. Motivated by this idea, we introduce two for the representation of prostate axial plane magnetic resonance images. In more detail, are parametric closed curves plane. The analytic study includes geometric role parameters describing curves, symmetries, invariants, special cases, elliptic Fourier descriptors, conditions simple and area enclosed surfaces. were validated shapes fitting delineated a radiologist measuring errors mean distance, Hausdorff distance Dice similarity coefficient. Validation was also conducted comparing our deformed superellipse model used literature. Our equivalent metrics model; however, they have advantage straightforward formulation depend on fewer parameters, implying reduced computational time process. Due validation, may be applied developing innovative performing methods or improving existing ones.

Language: Английский

---

Comparison of quantitative whole body PET parameters on [68Ga]Ga-PSMA-11 PET/CT using ordered Subset Expectation Maximization (OSEM) vs. bayesian penalized likelihood (BPL) reconstruction algorithms in men with metastatic castration-resistant prostate cancer

Cancer Imaging, Journal Year: 2024, Volume and Issue: 24(1)

Published: May 6, 2024

Abstract Background PSMA PET/CT is a predictive and prognostic biomarker for determining response to [ 177 Lu]Lu-PSMA-617 in patients with metastatic castration resistant prostate cancer (mCRPC). Thresholds defined date may not be generalizable newer image reconstruction algorithms. Bayesian penalized likelihood (BPL) algorithm novel that improve contrast whilst preventing introduction of noise. The aim this study compare the quantitative parameters obtained using BPL Ordered Subset Expectation Maximization (OSEM) Methods Fifty consecutive mCRPC who underwent 68 Ga]Ga-PSMA-11 OSEM assess suitability therapy were selected. was then used retrospectively reconstruct same PET raw data. Quantitative volumetric measurements such as tumour standardised uptake value (SUV)max, SUVmean Molecular Tumour Volume (MTV-PSMA) calculated on both methods. Results compared (Bland-Altman, Pearson correlation coefficient) including subgroups low high-volume disease burdens (MTV-PSMA cut-off 40 mL). SUVmax higher, MTV-PSMA lower reconstructed images group, mean difference 8.4 (17.5%), 0.7 (8.2%) − 21.5 mL (-3.4%), respectively. There strong between SUVmax, SUVmean, values (Pearson r 0.98, 0.99, 1.0, respectively). No reclassified from high volume or vice versa when switching reconstruction. Conclusions produced by methods are strongly correlated. Differences proportional small which biomarker. Our suggests acceptable without clinical impact findings. For longitudinal comparison, committing method would preferred ensure consistency.

Language: Английский

---

Radiomics reproducibility in computed tomography through changes of ROI size, resolution, and hounsfield unit: A phantom study

Radiography, Journal Year: 2024, Volume and Issue: 30(6), P. 1629 - 1636

Published: Oct. 1, 2024

Language: Английский

---

Prostate-Specific Membrane Antigen-Positron Emission Tomography-Guided Radiomics and Machine Learning in Prostate Carcinoma

Cancers, Journal Year: 2024, Volume and Issue: 16(19), P. 3369 - 3369

Published: Oct. 1, 2024

Positron emission tomography (PET) using radiolabeled prostate-specific membrane antigen targeting PET-imaging agents has been increasingly used over the past decade for imaging and directing prostate carcinoma treatment. Here, we summarize available literature data on radiomics machine learning these in carcinoma. Gleason scores derived from biopsy after resection are discordant a large number of patients. Available studies suggest that applied to PSMA-radioligand avid primary might be better performing than biopsy-based Gleason-scoring could serve as an alternative non-invasive GS characterization. Furthermore, it may allow prediction biochemical recurrence with net benefit clinical utilization. Machine based PET/CT features was also shown able differentiate benign malignant increased tracer uptake PSMA-targeting radioligand examinations, thus paving way fully automated image reading nuclear medicine. As treatment outcome following 177Lu-PSMA therapy overall survival, limited have reported promising results images this purpose. Its added value parameters warrants further exploration larger datasets

Language: Английский

---