A fast fully connected neural network-based method for random and scattered coincidence correction in polyhedral brain PET

Yuxuan Zhong; Senhao Wang; Xiulian Chen

doi:10.2478/nuka-2026-0007

.blurhash-client-img { display: none !important; }

A fast fully connected neural network-based method for random and scattered coincidence correction in polyhedral brain PET

Nukleonika

Volume 71 (2026): Issue 3 (September 2026)

By: Yuxuan Zhong , Senhao Wang and Xiulian Chen

Open Access

|Aug 2026

Abstract

Background

Scattered and random coincidences in brain PET imaging introduce significant noise and degrade resolution. Current correction algorithms for these effects are complex and computationally intensive. Fully connected neural network (FCNN) offers strong feature learning capabilities, versatility, and flexibility. Given that the coincidence data in brain PET imaging has relatively few features, FCNNs are well-suited to rapidly establish the relationship between coincidence data information and coincidence event types.

Methods

Eleven digital brain models were simulated using the Geant4 application for tomographic emission (GATE) [1]. Features extracted from the resulting data were used as inputs, with coincidence type as the target. FCNN was trained on this dataset. Quantitative evaluation was performed using contrast recovery coefficient (CRC), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) metrics.

Results

FCNN enables rapid filtering of coincidence data outside the brain region and efficiently corrects both random and scattered coincidences within the brain region. After valid coincidence discrimination, the valid coincidence rate remained at 95.60%, while the scatter fraction decreased from 40.60% to 29.01% and the random fraction dropped from 44.65% to 33.85%. Quantitative evaluation demonstrates that the random and scatter coincidence correction (RSC) method significantly enhances image quality when combined with attenuation-corrected (AC), improving the CRC from 0.946 to 0.969, SNR from 10.24 to 19.13, and CNR from 0.402 to 0.496. Conversely, under non-attenuation-corrected (NAC) conditions, RSC improves CRC (0.945 to 0.971) but reduces SNR (11.53 to 9.73) and CNR (0.569 to 0.521), indicating a trade-off between signal purity and statistical noise.

Conclusion

FCNN effectively reconstructs the true activity distribution of the radioactive source and enhances image contrast, demonstrating promise for clinical real-time brain PET imaging.

References

Strulab, D., Santin, G., Lazaro, D., Breton, V., & Morel, C. (2003). GATE (Geant4 application for tomographic emission): a PET/SPECT general-purpose simulation platform. Nucl. Phys. B-Proc. Suppl., 125, 75–79. DOI: 10.1016/S0920-5632(03)90969-8.
Open DOI Search in Google Scholar Back to article
Tsoumpas, C., Aguiar, P., Nikita, K. S., Ros, D., & Thielemans, K. (2004). Evaluation of the single scatter simulation algorithm implemented in the STIR library. In IEEE Symposium Conference Record Nuclear Science 2004 (Vol. 6, pp. 3361–3365). IEEE. DOI: 10.1109/nssmic.2004.1466455.
Open DOI Search in Google Scholar Back to article
Adam, L. E., Karp, J. S., & Brix, G. (1999). Investigation of scattered radiation in 3D whole-body positron emission tomography using Monte Carlo simulations. Phys. Med. Biol., 44(12), 2879–2895. DOI: 10.1088/0031-9155/44/12/302.
Open DOI Search in Google Scholar Back to article
Watson, C. C., Newport, D. M. E. C., & Casey, M. E. (1996). A single scatter simulation technique for scatter correction in 3D PET. In P. Grangeat & J. -L. Amans (Eds.), Three-dimensional image reconstruction in radiology and nuclear medicine (pp. 255–268). Dordrecht: Springer Netherlands. DOI: 10.1007/978-94-015-8749-5_18.
Open DOI Search in Google Scholar Back to article
Levin, C. S., Dahlbom, M., & Hoffman, E. J. (1995). A Monte Carlo correction for the effect of Compton scattering in 3-D PET brain imaging. IEEE Trans. Nucl. Sci., 42(4), 1181–1185. DOI: 10.1109/nssmic.1994.474776.
Open DOI Search in Google Scholar Back to article
Holdsworth, C. H., Badawi, R. D., Santos, P., Van den Abbeele, A. D., Hoffman, E. J., & El Fakhri, G. (2003). Evaluation of a Monte Carlo scatter correction in clinical 3D PET. In 2023 IEEE Nuclear Science Symposium, 4–11 November 2023, Vancouver, Canada. Conference Record 2023 (IEEE Cat. No. 03CH37515) (Vol. 4, pp. 2540–2544). IEEE. DOI: 10.1109/nssmic.2003.1352408.
Open DOI Search in Google Scholar Back to article
Bailey, D. L., & Meikle, S. R. (1994). A convolution-subtraction scatter correction method for 3D PET. Phys. Med. Biol., 39(3), 411–424. DOI: 10.1088/0031-9155/39/3/009.
Open DOI Search in Google Scholar Back to article
Ma, B., Gaens, M., Caldeira, L., Bert, J., Lohmann, P., Tellmann, L., Lerche, C., Scheins, J., Rota Kops, E., Xu, H., Lenz, M., Pietrzyk, U., & Shah, N. J. (2019). Scatter correction based on GPU-accelerated full Monte Carlo simulation for brain PET/MRI. IEEE Trans. Med. Imaging, 39(1), 140–151. DOI: 10.1109/tmi.2019.2921872.
Open DOI Search in Google Scholar Back to article
Yang, J., Park, D., Gullberg, G. T., & Seo, Y. (2019). Joint correction of attenuation and scatter in image space using deep convolutional neural networks for dedicated brain 18F-FDG PET. Phys. Med. Biol., 64(7), 075019. DOI: 10.1088/1361-6560/ab0606.
Open DOI Search in Google Scholar Back to article
Arabi, H., Bortolin, K., Ginovart, N., Garibotto, V., & Zaidi, H. (2020). Deep learning-guided joint attenuation and scatter correction in multitracer neuroimaging studies. Human Brain Mapp., 41(13), 3667–3679. DOI: 10.1002/hbm.25039.
Open DOI Search in Google Scholar Back to article
Shiri, I., Arabi, H., Geramifar, P., Hajianfar, G., Ghafarian, P., Rahmim, A., Reza Ay, M., & Zaidi, H. (2020). Deep-JASC: joint attenuation and scatter correction in whole-body 18F-FDG PET using a deep residual network. Eur. J. Nucl. Med. Mol. Imaging, 47(11), 2533–2548. DOI: 10.1007/s00259-020-04852-5.
Open DOI Search in Google Scholar Back to article
Liu, F., Jang, H., Kijowski, R., Zhao, G., Bradshaw, T., & McMillan, A. B. (2018). A deep learning approach for 18F-FDG PET attenuation correction. EJNMMI Phys., 5(1), 24. DOI: 10.1186/s40658-018-0225-8.
Open DOI Search in Google Scholar Back to article
Brasse, D., Kinahan, P. E., Lartizien, C., Comtat, C., Casey, M., & Michel, C. (2005). Correction methods for random coincidences in fully 3D whole-body PET: impact on data and image quality. J. Nucl. Med., 46(5), 859–867. DOI: 10.1109/nssmic.2001.1009234.
Open DOI Search in Google Scholar Back to article
Ollinger, J. M., & Fessler, J. A. (2002). Positron-emission tomography. IEEE Signal Process. Mag., 14(1), 43–55. DOI: 10.1109/79.560323.
Open DOI Search in Google Scholar Back to article
Yan, J., Schaefferkoetter, J., Conti, M., & Townsend, D. (2016). A method to assess image quality for low-dose PET: analysis of SNR, CNR, bias and image noise. Cancer Imaging, 16(1), 26. DOI: 10.1186/s40644-016-0086-0.
Open DOI Search in Google Scholar Back to article
Liang, Y., Xiao, Q., Lu, L., & Xie, J. (2021). FCNNLib: A flexible convolution algorithm library for deep learning on FPGAs. IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst., 41(8), 2546–2559. DOI: 10.1109/tcad.2021.3108065.
Open DOI Search in Google Scholar Back to article
Aziz, L., Salam, M. S. B. H., Sheikh, U. U., & Ayub, S. (2020). Exploring deep learning-based architecture, strategies, applications and current trends in generic object detection: A comprehensive review. IEEE Access, 8, 170461–170495. DOI: 10.1109/access.2020.3021508.
Open DOI Search in Google Scholar Back to article
Shawahna, A., Sait, S. M., & El-Maleh, A. (2018). FPGA-based accelerators of deep learning networks for learning and classification: A review. IEEE Access, 7, 7823–7859. DOI: 10.1109/access.2018.2890150.
Open DOI Search in Google Scholar Back to article
Ahmed, S. F., Alam, M. S. B., Hassan, M., Rozbu, M. R., Ishtiak, T., Rafa, N., Mafijur, M., Shawkat, A. B. M., & Gandomi, A. H. (2023). Deep learning modelling techniques: current progress, applications, advantages, and challenges. Artif. Intell. Rev., 56(11), 13521–13617. DOI: 10.1007/s10462-023-10466-8.
Open DOI Search in Google Scholar Back to article
Shi, H., Du, D., Xu, J., Su, Z., & Peng, Q. (2015). Design study of dedicated brain PET with polyhedron geometry. Technol. Health Care, 23(Suppl. 2), S615–S623. DOI: 10.3233/thc-151000.
Open DOI Search in Google Scholar Back to article
Chen, G., Weng, F., Hong, X., Tao, W., Zhao, Z., Peng, Q., & Huang, Q. (2018). Developing a ‘multiPatchPET’ system in GATE for a PET system design with irregular geometries. Phys. Med. Biol., 63(17), 17NT02. DOI: 10.1088/1361-6560/aad8fd.
Open DOI Search in Google Scholar Back to article
Scheins, J. J., Herzog, H., & Shah, N. J. (2011). Fully-3D PET image reconstruction using scanner-independent, adaptive projection data and highly rotation-symmetric voxel assemblies. IEEE Trans. Med. Imaging, 30(3), 879–892. DOI: 10.1109/tmi.2011.2109732.
Open DOI Search in Google Scholar Back to article
Niu, M., Liu, Z., Kuang, Z., Wang, X., Ren, N., Sang, Z., Wu, S., Cong, L., Sun, T., Hu, Z., & Yang, Y. (2022). Ultra-high-resolution depth-encoding small animal PET detectors: using GAGG and LYSO crystal arrays. Med. Phys., 49(5), 3006–3020. DOI: 10.1002/mp.15606.
Open DOI Search in Google Scholar Back to article
Aubert-Broche, B., Griffin, M., Pike, G. B., Evans, A. C., & Collins, D. L. (2006). Twenty new digital brain phantoms for creation of validation image data bases. IEEE Trans. Med. Imaging, 25(11), 1410–1416. DOI: 10.1109/tmi.2006.883453.
Open DOI Search in Google Scholar Back to article
Rousset, O. G., Ma, Y., & Evans, A. C. (1998). Correction for partial volume effects in PET: principle and validation. J. Nucl. Med., 39(5), 904–911.
Search in Google Scholar Back to article
Zubal, I. G., Harrell, C. R., Smith, E. O., Rattner, Z., Gindi, G., & Hoffer, P. B. (1994). Computerized three-dimensional segmented human anatomy. Med. Phys., 21(2), 299–302. DOI: 10.1118/1.597290.
Open DOI Search in Google Scholar Back to article
Cañizares, G., Jiménez-Serrano, S., Lucero, A., Morera-Ballester, C., Muñoz, E., Benlloch, J. M., & González, A. J. (2024). Simulation study of clinical PET scanners with different geometries, including TOF and DOI capabilities. IEEE Trans. Radiat. Plasma Med. Sci., 8(6), 690–699. DOI: 10.1109/TRPMS.2024.3365911.
Open DOI Search in Google Scholar Back to article
Xiong, J., Ma, Z., Yang, H., Yang, C., & Chen, J. (2025). Bayesian-optimized fully connected neural network for enhanced prediction accuracy in concrete compressive strength estimation. Civil Environ. Eng., 21(2), 1292–1303. DOI: 10.2478/cee-2025-0095.
Open DOI Search in Google Scholar Back to article
Anguita, D., Ghelardoni, L., Ghio, A., Oneto, L., & Ridella, S. (2012). The 'K' in K-fold cross validation. In ESANN 2012 Procedings, European Symposium on Artificial Neural Network, Computational Intelligence and Machine Learning, Bruges, Belgium, 25–27 April 2012 (Vol. 102, pp. 441–446). https://hdl.handle.net/11567/539164.
Search in Google Scholar Back to article
Vandenberghe, S., Mikhaylova, E., D'Hoe, E., Mollet, P., & Karp, J. S. (2016). Recent developments in time-of-flight PET. EJNMMI Phys., 3(1), 3. DOI: 10.1186/s40658-016-0138-3.
Open DOI Search in Google Scholar Back to article
Mehranian, A., Kotasidis, F., & Zaidi, H. (2016). Accelerated time-of-flight (TOF) PET image reconstruction using TOF bin subsetization and TOF weighting matrix pre-computation. Phys. Med. Biol., 61(3), 1309–1331. DOI: 10.1088/0031-9155/61/3/1309.
Open DOI Search in Google Scholar Back to article
Filipović, M., Comtat, C., & Stute, S. (2019). Time-off-light (TOF) implementation for PET reconstruction in practice. Phys. Med. Biol., 64(23), 23NT01. DOI: 10.1088/1361-6560/ab4f0b.
Open DOI Search in Google Scholar Back to article
Vidal, F. P., Rocchisani, J. M., Jia, X., & Jiang, S. B. (2015). Reconstruction in positron emission tomography. In X. Jia & S. B. Jiang (Eds.), Graphics processing unit-based high performance computing in radiation therapy (pp. 185–208). CRC Press.
Search in Google Scholar Back to article
Chetih, N., & Messali, Z. (2015). Tomographic image reconstruction using filtered back projection (FBP) and algebraic reconstruction technique (ART). In 2015 3rd International Conference on Control, Engineering & Information Technology (CEIT), 25–27 May 2015 (pp. 1–6). IEEE. DOI: 10.1109/CEIT.2015.7233031.
Open DOI Search in Google Scholar Back to article
Polycarpou, I., Tsoumpas, C., King, A. P., & Marsden, P. K. (2015). Quantitative evaluation of PET respiratory motion correction using MR derived simulated data. IEEE Trans. Nucl. Sci., 62(6), 3110–3116. DOI: 10.1109/TNS.2015.2494593.
Open DOI Search in Google Scholar Back to article
Doss, K. K. M., Mion, P. E., Kao, Y. C. J., Kuo, T. T., & Chen, J. C. (2022). Performance evaluation of a PET of 7T Bruker Micro-PET/MR based on NEMA NU 4-2008 standards. Electronics, 11(14), 2194. DOI: 10.3390/electronics11142194.
Open DOI Search in Google Scholar Back to article
Li, D., Li, X., Qin, Y., Cheng, Y., & Wang, H. (2013). Radar coincidence imaging: an instantaneous imaging technique with stochastic signals. IEEE Trans. Geosci. Remote Sensing, 52(4), 2261–2277. DOI: 10.1109/tgrs.2013.2258929.
Open DOI Search in Google Scholar Back to article
Yan, J., Schaefferkoetter, J., Conti, M., & Townsend, D. (2016). A method to assess image quality for low-dose PET: analysis of SNR, CNR, bias and image noise. Cancer Imaging, 16(1), 26. DOI: 10.1186/s40644-016-0086-0.
Open DOI Search in Google Scholar Back to article
Roy, S., Whitehead, T. D., Quirk, J. D., Salter, A., Ademuyiwa, F. O., Li, S., An, H., & Shoghi, K. I. (2020). Optimal co-clinical radiomics: Sensitivity of radiomic features to tumour volume, image noise and resolution in co-clinical T1-weighted and T2-weighted magnetic resonance imaging. EBioMedicine, 59. DOI: 10.1016/j.ebiom.2020.102963.
Open DOI Search in Google Scholar Back to article
Pyatykh, S., Hesser, J., & Zheng, L. (2012). Image noise level estimation by principal component analysis. IEEE Trans. Image Process., 22(2), 687–699. DOI: 10.1109/tip.2012.2221728.
Open DOI Search in Google Scholar Back to article

Articles in this issue

DOI: https://doi.org/10.2478/nuka-2026-0007 | Journal eISSN: 1508-5791 (formerly 0029-5922) | Journal ISSN: 0029-5922

Journal RSS Feed

Language: English

Page range: 53 - 60

Submitted on: Mar 15, 2026

Accepted on: Jun 15, 2026

Published on: Aug 1, 2026

Published by: Institute of Nuclear Chemistry and Technology

In partnership with: Paradigm Publishing Services

Keywords:

Fully connected neural network (FCNN),

Polyhedral brain PET,

Random coincidence correction,

Scattered coincidence correction,

Valid coincidence discrimination

Related subjects:

Chemistry,

Nuclear chemistry

© 2026 Yuxuan Zhong, Senhao Wang, Xiulian Chen, published by Institute of Nuclear Chemistry and Technology
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Volume 71 (2026): Issue 3 (September 2026)