EUVIMEDEUVIMED entdecken
Open Access

Pix2Pix generative-adversarial network in improving the quality of T2-weighted prostate magnetic resonance imaging: a multi-reader study

Yeliz Başar iD, Mustafa Said Kartal iD, Mustafa Ege Seker iD, Deniz Alis iD, Delal Seker iD, Müjgan Orman iD, Sabri Şirolu iD, Serpil Kurtcan, Aydan Arslan iD, Nurper Denizoğlu iD, İlkay Öksüz iD, Ercan Karaarslan iD

Diagnostic and Interventional Radiology · 2025

DOI: 10.4274/dir.2025.243102 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Diagnostic and Interventional Radiology, 2025. Verfasst von Yeliz Başar und 11 weiteren Autor:innen. Beteiligte Einrichtung: Acıbadem Healthcare Group, Department of Radiology, İstanbul, Türkiye. Hauptthemen laut MEDLINE: Prostatic Neoplasms, Magnetic Resonance Imaging, Prostate, Deep Learning. Die Arbeit nennt 20 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

PURPOSE: To assess the performance and feasibility of generative deep learning in enhancing the image quality of T2-weighted (T2W) prostate magnetic resonance imaging (MRI). METHODS: Axial T2W images from the prostate imaging: cancer artificial intelligence dataset (n = 1,476, biologically males; n = 1,500 scans) were used, partitioned into training (n = 1300), validation (n = 100), and testing (n = 100) sets. A Pix2Pix model was trained on original and synthetically degraded images, generated using operations such as motion, Gaussian noise, blur, ghosting, spikes, and bias field inhomogeneities to enhance image quality. The efficacy of the model was evaluated by seven radiologists using the prostate imaging quality criteria to assess original, degraded, and improved images. The evaluation also included tests to determine whether the images were original or synthetically improved. Additionally, the model's performance was tested on the in-house external testing dataset of 33 patients. The statistical significance was assessed using the Wilcoxon signedrank test. RESULTS: Results showed that synthetically improved images [median score (interquartile range) 4.71 (1)] were of higher quality than degraded images [3.36 (3), P = 0.0001], with no significant difference from original images [5 (1.14), P > 0.05]. Observers equally identified original and synthetically improved images as original (52% and 53%), proving the model's ability to retain realistic attributes. External testing on a dataset of 33 patients confirmed a significant improvement (P = 0.001) in image quality, from a median score of 4 (2.286)-4.71 (1.715). CONCLUSION: The Pix2Pix model, trained on synthetically degraded data, effectively improved prostate MRI image quality while maintaining realism and demonstrating both applicability to real data and generalizability across various datasets. CLINICAL SIGNIFICANCE: This study critically assesses the efficacy of the Pix2Pix generative-adversarial network in enhancing T2W prostate MRI quality, demonstrating its potential to produce high-quality, realistic images indistinguishable from originals, thereby potentially advancing radiology practice by improving diagnostic accuracy and image reliability.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Prostatic NeoplasmsMagnetic Resonance ImagingProstateDeep LearningImage Interpretation, Computer-AssistedHumansMaleMiddle AgedAgedFeasibility Studies

Stichwörter der Autor:innen

Deep learninggenerative artificial intelligencemagnetic resonance imagingprostateprostate imaging quality

Autor:innen und Institutionen

  • Acıbadem Healthcare Group, Department of Radiology, İstanbul, Türkiye.
  • Cumhuriyet University Faculty of Medicine, Sivas, Türkiye.
  • University of Wisconsin-Madison, School of Medicine, Department of Radiology, Madison, USA.
  • Acıbadem Mehmet Ali Aydınlar University Faculty of Medicine, Department of Radiology, İstanbul, Türkiye.
  • Dicle University Faculty of Engineering, Department of Electrical-Electronics Engineering, Diyarbakır, Türkiye.
  • University of Health Sciences Türkiye, Şişli Hamidiye Etfal Training and Research Hospital, Clinic of Radiology, İstanbul, Türkiye.

ORCID-Profile: Yeliz Başar, Mustafa Said Kartal, Mustafa Ege Seker, Deniz Alis, Delal Seker, Müjgan Orman, Sabri Şirolu, Aydan Arslan, Nurper Denizoğlu, İlkay Öksüz, Ercan Karaarslan

Zitieren

APA

Başar, Y., Kartal, M. S., Seker, M. E., Alis, D., Seker, D., Orman, M., Şirolu, S., Kurtcan, S., Arslan, A., Denizoğlu, N., Öksüz, İ., & Karaarslan, E. (2025). Pix2Pix generative-adversarial network in improving the quality of T2-weighted prostate magnetic resonance imaging: a multi-reader study. Diagnostic and Interventional Radiology. https://doi.org/10.4274/dir.2025.243102

Vancouver

Başar Y, Kartal MS, Seker ME, Alis D, Seker D, Orman M, et al. Pix2Pix generative-adversarial network in improving the quality of T2-weighted prostate magnetic resonance imaging: a multi-reader study. Diagnostic and Interventional Radiology. 2025. doi:10.4274/dir.2025.243102

BibTeX
@article{Baar2025,
  author = {Başar, Yeliz and Kartal, Mustafa Said and Seker, Mustafa Ege and Alis, Deniz and Seker, Delal and Orman, Müjgan and Şirolu, Sabri and Kurtcan, Serpil and Arslan, Aydan and Denizoğlu, Nurper and Öksüz, İlkay and Karaarslan, Ercan},
  title = {Pix2Pix generative-adversarial network in improving the quality of T2-weighted prostate magnetic resonance imaging: a multi-reader study},
  journal = {Diagnostic and Interventional Radiology},
  year = {2025},
  doi = {10.4274/dir.2025.243102}
}
Kostenlos starten →