Skip to main content
Log in

Künstliche Intelligenz (KI) in der Radiologie?

Brauchen wir langfristig noch so viele Radiologen?

Artificial intelligence (AI) in radiology?

Do we need as many radiologists in the future?

  • Leitthema
  • Published:
Der Urologe Aims and scope Submit manuscript

Zusammenfassung

Wir befinden uns mitten in einer digitalen Revolution der Medizin. Da stellt sich die Frage, ob besonders Fächer wie die Radiologie, die sich oberflächlich betrachtet mit der Bildinterpretation beschäftigt, durch diese Revolution besonders verändern werden. Insbesondere ist zu diskutieren, ob nicht die zukünftige Erledigung zunächst einfacherer, dann komplexerer Bildanalyseaufgaben durch Computersysteme zu einem verminderten Bedarf an Radiologen führen wird. Was die Radiologie besonders auszeichnet ist ihre Schlüsselposition zwischen Hochtechnologie und medizinischer Versorgung. In diesem Beitrag wird diskutiert, dass nicht nur die Radiologie, sondern jedes medizinische Fach von Neuerungen durch die digitale Revolution betroffen sein wird, dass eine Redefinition der bildgebenden Fächer sinnvoll erscheint und dass im Rahmen immer größerer Bilddatenmengen die Ankunft der künstlichen Intelligenz (KI) in der Radiologie zu begrüßen ist, um überhaupt mit der derzeitigen Anzahl an Radiologen noch in Zukunft die Menge an Bilddaten beherrschen zu können. Weiterhin hat die zunehmende Befundlast in den letzten Jahren dazu geführt, dass Radiologen einen stark durch die Befundungsarbeit dominierten Arbeitsalltag erleben. Die KI kann in den genannten Bereichen zu einer Verbesserung der Effizienz sowie der Balance beitragen. Hinsichtlich der Facharztausbildung ist eine Integration informationstechnischer Inhalte in das Curriculum zu erwarten. Die Radiologie fungiert damit als Pionier des Einzugs der KI in die Medizin. Es ist zu erwarten, dass zu dem Zeitpunkt, an dem Radiologen durch KI wesentlich ersetzt werden können, ein Ersatz menschlicher Beiträge auch in anderen medizinischen und nicht-medizinischen Fächern weit fortgeschritten sein wird.

Abstract

We are in the middle of a digital revolution in medicine. This raises the question of whether subjects such as radiology, which is superficially concerned with the interpretation of images, will be particularly changed by this revolution. In particular, it should be discussed whether in the future the completion of initially simpler, then more complex image analysis tasks by computer systems may lead to a reduced need for radiologists. What distinguishes radiology in particular is its key position between advanced technology and medical care. This article discusses that not only radiology but every medical discipline will be affected by innovations due to the digital revolution, and that a redefinition of medical specialties focusing on imaging and visual interpretation makes sense and that the arrival of artificial intelligence (AI) in radiology is to be welcomed in the context of ever larger amounts of image data—to at all be able to handle the increasing amount of image data in the future at the current number of radiologists. In this respect, the balance between research and teaching in comparison to patient care is more difficult to maintain in the academic environment. AI can help improve efficiency and balance in the areas mentioned. With regard to specialist training, information technology topics are expected to be integrated into the radiological curriculum. Radiology acts as a pioneer designing the entry of AI into medicine. It is to be expected that by the time radiologists can be substantially replaced by AI, the replacement of human contributions in other medical and non-medical fields will also be well advanced.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Abb. 1
Abb. 2

Literatur

  1. Bonekamp D, Schlemmer HP (2021) Machine learning and multiparametric MRI for early diagnosis of prostate cancer. Urologe A 60(5):576–591

    Article  CAS  Google Scholar 

  2. Schelb P et al (2019) Classification of cancer at prostate MRI: deep learning versus clinical PI-RADS assessment. Radiology 293(3):607–617

    Article  Google Scholar 

  3. Zhang KS et al (2021) Improvement of PI-RADS-dependent prostate cancer classification by quantitative image assessment using radiomics or mean ADC. Magn Reson Imaging 82:9–17

    Article  CAS  Google Scholar 

  4. Netzer N et al (2021) Fully automatic deep learning in Bi-institutional prostate magnetic resonance imaging: effects of cohort size and heterogeneity. Invest Radiol 56(12):799–808

    Article  CAS  Google Scholar 

  5. Pesapane F, Codari M, Sardanelli F (2018) Artificial intelligence in medical imaging: threat or opportunity? Radiologists again at the forefront of innovation in medicine. Eur Radiol Exp 2(1):1–10

    Article  Google Scholar 

  6. Langlotz CP (2019) Will artificial intelligence replace radiologists? Radiological Society of North America, https://doi.org/10.1148/ryai.2019190058

    Book  Google Scholar 

  7. Reardon S (2019) Rise of robot radiologists. Nature 576(7787):S54–S54

    Article  CAS  Google Scholar 

  8. Kobayashi Y, Ishibashi M, Kobayashi H (2019) How will “democratization of artificial intelligence” change the future of radiologists? Jpn J Radiol 37(1):9–14

    Article  CAS  Google Scholar 

  9. Tang A et al (2018) Canadian association of radiologists white paper on artificial intelligence in radiology. Can Assoc Radiol J 69(2):120–135

    Article  Google Scholar 

  10. Canadian Association of Radiologists (CAR) Artificial Intelligence Working Group (2019) Canadian association of radiologists white paper on ethical and legal issues related to artificial intelligence in radiology. Can Assoc Radiol J 70(2):107–118

    Article  Google Scholar 

  11. Larson DB et al (2020) Ethics of using and sharing clinical imaging data for artificial intelligence: a proposed framework. Radiology 295(3):675–682

    Article  Google Scholar 

  12. Jungmann F et al (2021) Attitudes toward artificial intelligence among radiologists, IT specialists, and industry. Acad Radiol 28(6):834–840

    Article  Google Scholar 

  13. van Hoek J et al (2019) A survey on the future of radiology among radiologists, medical students and surgeons: students and surgeons tend to be more skeptical about artificial intelligence and radiologists may fear that other disciplines take over. Eur J Radiol 121:108742

    Article  Google Scholar 

  14. Pesapane F et al (2020) Myths and facts about artificial intelligence: why machine-and deep-learning will not replace interventional radiologists. Med Oncol 37(5):1–9

    Article  Google Scholar 

  15. Obermeyer Z, Emanuel EJ (2016) Predicting the future—big data, machine learning, and clinical medicine. N Engl J Med 375(13):1216–1219

    Article  Google Scholar 

  16. Hassabis D (2017) Artificial intelligence: chess match of the century. Nature 544(7651):413–414

    Article  CAS  Google Scholar 

  17. Gibney E (2016) Google AI algorithm masters ancient game of Go. Nature 529(7587):445

    Article  CAS  Google Scholar 

  18. Hannun A et al (2014) Deep speech: scaling up end-to-end speech recognition (arXiv:1412.5567)

    Google Scholar 

  19. LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521(7553):436–444

    Article  CAS  Google Scholar 

  20. Russell S, Dewey D, Tegmark M (2015) Research priorities for robust and beneficial artificial intelligence: an open letter. Ai Magazine 36(4):105–114

    Article  Google Scholar 

  21. Schlemmer HP et al (2021) Imaging of prostate cancer. Dtsch Arztebl Int. https://doi.org/10.3238/arztebl.m2021.0309

    Article  PubMed  PubMed Central  Google Scholar 

  22. Fehr D et al (2015) Automatic classification of prostate cancer Gleason scores from multiparametric magnetic resonance images. Proc Natl Acad Sci USA 112(46):E6265–E6273

    Article  CAS  Google Scholar 

  23. Wang J et al (2017) Machine learning-based analysis of MR radiomics can help to improve the diagnostic performance of PI-RADS v2 in clinically relevant prostate cancer. Eur Radiol 27(10):4082–4090

    Article  Google Scholar 

  24. Bonekamp D et al (2018) Radiomic machine learning for characterization of prostate lesions with MRI: comparison to ADC values. Radiology 289(1):128–137

    Article  Google Scholar 

  25. Cameron A et al (2015) MAPS: a quantitative radiomics approach for prostate cancer detection. IEEE Trans Biomed Eng 63(6):1145–1156

    Article  Google Scholar 

  26. Armato SG et al (2018) PROSTATEx challenges for computerized classification of prostate lesions from multiparametric magnetic resonance images. J Med Imaging 5(4):44501

    Article  Google Scholar 

  27. Radtke JP et al (2016) Multiparametric magnetic resonance imaging (MRI) and MRI-transrectal ultrasound fusion biopsy for index tumor detection: correlation with radical prostatectomy specimen. Eur Urol 70(5):846–853

    Article  Google Scholar 

  28. Hosseinzadeh M et al (2021) Deep learning-assisted prostate cancer detection on bi-parametric MRI: minimum training data size requirements and effect of prior knowledge. Eur Radiol. https://doi.org/10.1007/s00330-021-08320-y

    Article  PubMed  PubMed Central  Google Scholar 

  29. Reinke A et al (2021) Common limitations of image processing metrics: a picture story (arXiv:2104.05642)

    Google Scholar 

  30. Schelb P et al (2021) Simulated clinical deployment of fully automatic deep learning for clinical prostate MRI assessment. Eur Radiol 31(1):302–313

    Article  Google Scholar 

  31. Schelb P et al (2021) Comparison of prostate MRI lesion segmentation agreement between multiple radiologists and a fully automatic deep learning system. Rofo 193(5):559–573

    Article  Google Scholar 

  32. Tavakoli AA et al (2021) Measured multipoint ultra-high b‑value diffusion MRI in the assessment of MRI-detected prostate lesions. Invest Radiol 56(2):94–102

    Article  CAS  Google Scholar 

  33. Wang X et al (2021) Comparison of single-scanner single-protocol quantitative ADC measurements to ADC ratios to detect clinically significant prostate cancer. Eur J Radiol 136:109538

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to David Bonekamp.

Ethics declarations

Interessenkonflikt

H.-P. Schlemmer gibt an, Konsultationen/Honorare: Siemens, Curagita, Profound, Bayer; Board membership: Curagita, Grants: BMBF, Deutsche Krebshilfe, Dietmar-Hopp-Stiftung, Roland-Ernst-Stiftung, BMWK. D. Bonekamp ist Sprecher für Bayer Vital und erhält Grant Support durch das BMWK.

Für diesen Beitrag wurden von den Autoren keine Studien an Menschen oder Tieren durchgeführt. Für die aufgeführten Studien gelten die jeweils dort angegebenen ethischen Richtlinien.

Additional information

figure qr

QR-Code scannen & Beitrag online lesen

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bonekamp, D., Schlemmer, HP. Künstliche Intelligenz (KI) in der Radiologie?. Urologe 61, 392–399 (2022). https://doi.org/10.1007/s00120-022-01768-w

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00120-022-01768-w

Schlüsselwörter

Keywords

Navigation