AI in Urological Oncology: Prostate Cancer Diagnosis with Magnetic Resonance Imaging

Research output: Contribution to book/Conference proceedings/Anthology/ReportChapter in book/Anthology/ReportContributedpeer-review

Contributors

  • Sherif Mehralivand - , National Institutes of Health (NIH) (Author)
  • Baris Turkbey - , National Institutes of Health (NIH) (Author)

Abstract

The use of medical imaging, especially multiparametric magnetic resonance imaging (MRI), is critical in the diagnosis of prostate cancer. This technique enables better visualization and localization of prostate lesions, which has led to improved prostate cancer detection and classification. However, significant inter-observer variability exists in these imaging reports. AI solutions can assist radiologists during the interpretation process, particularly those with low to intermediate levels of experience. First experiences in prostate lesion detection on multiparametric MRI using deep learning have demonstrated higher detection sensitivities compared to classical machine learning algorithms. This will continue to improve in the future with the availability of larger and diverse prostate MRI datasets. More studies are needed to determine the practical value of AI in this field.

Details

Original languageEnglish
Title of host publicationAI in Clinical Medicine
PublisherWiley, Chichester
Pages298-306
Number of pages9
ISBN (electronic)9781119790686
ISBN (print)9781119790648
Publication statusPublished - 1 Jan 2023
Peer-reviewedYes
Externally publishedYes

Keywords

Sustainable Development Goals

ASJC Scopus subject areas

Keywords

  • Artificial intelligence, Deep learning, Machine learning, Multiparametric magnetic resonance imaging, Prostate, Prostatic neoplasms, Surgical oncology, Urology