Anterior|Próximo

Desempenho do GPT-4 para decisão automatizada de biópsia prostática com base em mpMRI: um estudo multicêntrico de evidências

  • role： Contribute equally, First author同等贡献者, 第一作者
  • Affiliation：

    Department of Urology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

    Institute of Urology, Beijing Municipal Health Commission, Beijing 101313, China

Ming-Jun Shi12,  
  • role： Contribute equally同等贡献者
  • Affiliation：

    Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

    Department of Ultrasound, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Zhi-Xiang Wang34,  
  • role： Contribute equally同等贡献者
  • Affiliation：

    Department of Radiology, Beijing Chaoyang Hospital, Capital Medical University, Beijing 100020, China

Shuang-Kun Wang5,  
  • Affiliation：

    Department of Urology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Xuan-Hao Li1,  
  • Affiliation：

    Department of Pathology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Yan-Lin Zhang6,  
  • Affiliation：

    Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Ying Yan3,  
  • Affiliation：

    Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Ran An3,  
  • Affiliation：

    Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Li-Ning Dong3,  
  • Affiliation：

    Department of Urology, Peking University Third Hospital, Beijing 100083, China

Lei Qiu7,  
  • Affiliation：

    Department of Radiology, Beijing Chaoyang Hospital, Capital Medical University, Beijing 100020, China

Tian Tian5,  
  • Affiliation：

    Department of Urology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

    Department of Urology, Beijing Fuxing Hospital, Capital Medical University, Beijing 100039, China

Jia-Xin Liu18,  
  • Affiliation：

    Department of Urology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Hong-Chen Song1,  
  • Affiliation：

    Department of Radiology, Beijing Chaoyang Hospital, Capital Medical University, Beijing 100020, China

Ya-Fan Wang5,  
  • Affiliation：

    Department of Radiology, Beijing Chaoyang Hospital, Capital Medical University, Beijing 100020, China

Che Deng5,  
  • Affiliation：

    Department of Urology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

    Department of Urology, Beijing Miyun District Traditional Chinese Medicine Hospital, Beijing 101500, China

Zi-Bing Cao19,  
  • Affiliation：

    Department of Urology, Peking University Third Hospital, Beijing 100083, China

Hong-Yin Wang7,  
  • Affiliation：

    Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Zheng Wang3,  
  • Affiliation：

    Division of Science and Technology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

Wei Wei10,  
  • role： Corresponding author通信作者
  • Affiliation：

    Department of Urology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

    Institute of Urology, Beijing Municipal Health Commission, Beijing 101313, China

  • Email：songjian1974@aliyun.com
  • Introduction：* Jian Song songjian1974@aliyun.com
Jian Song12*,  
  • role： Corresponding author通信作者
  • Affiliation：

    Department of Urology, Peking University Third Hospital, Beijing 100083, China

  • Email：lujian@bjmu.edu.cn
  • Introduction：Jian Lu lujian@bjmu.edu.cn
Jian Lu7*,  
  • role： Corresponding author通信作者
  • Affiliation：

    Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

  • Email：weixuan315@163.com
  • Introduction：Xuan Wei weixuan315@163.com
Xuan Wei3*,  
  • role： Corresponding author通信作者
  • Affiliation：

    Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China

  • Email：cjr.wzhch@vip.163.com
  • Introduction：Zhen-Chang Wang cjr.wzhch@vip.163.com
Zhen-Chang Wang3*

Resumo

A ressonância magnética multiparamétrica (mpMRI) avançou significativamente a detecção do câncer de próstata (PCa), mas as decisões sobre biópsia invasiva com escores moderados do sistema de relatórios e dados de imagem da próstata (PI-RADS) permanecem ambíguas. Para explorar a capacidade de tomada de decisão do Generative Pretrained Transformer-4 (GPT-4) para recomendações automatizadas de biópsia prostática, incluímos 2299 indivíduos que se submeteram a biópsia prostática entre 2018 e 2023 em 3 grandes centros médicos, com mpMRI disponível antes da biópsia e registros clínico-histopatológicos documentados. O GPT-4 gerou relatórios estruturados com prompts fornecidos. O desempenho do GPT-4 foi quantificado usando matrizes de confusão, e sensibilidade, especificidade, bem como a área sob a curva foram calculadas. Diversos procedimentos artificiais de avaliação foram realizados. Testes de soma de postos de Wilcoxon, teste exato de Fisher e testes de Kruskal-Wallis foram utilizados para comparações. Utilizando o maior tamanho amostral da população chinesa, pacientes com escores moderados de PI-RADS (escores 3 e 4) representaram 39,7% (912/2299), definidos como subconjunto de interesse (SOI). As taxas de detecção de câncer de próstata clinicamente significativo correspondentes aos escores PI-RADS 2-5 foram 9,4%, 27,3%, 49,2% e 80,1%, respectivamente. Quase 47,5% (433/912) dos pacientes SOI foram comprovadamente submetidos a biópsias prostáticas desnecessárias. Com a assistência do GPT-4, 20,8% (190/912) da população SOI poderia evitar biópsias desnecessárias, e teve desempenho ainda melhor [28,8% (118/410)] no subgrupo mais heterogêneo do escore PI-RADS 3. Mais de 90,0% dos relatórios gerados pelo GPT-4 foram abrangentes e fáceis de entender, mas menos satisfatórios quanto à precisão (82,8%). O GPT-4 também demonstrou potencial cognitivo para lidar com problemas complexos. Além disso, o método da cadeia de pensamento nos permitiu entender melhor a lógica de decisão por trás do GPT-4. Por fim, desenvolvemos a plataforma ProstAIGuide para facilitar o acesso de médicos e pacientes. Este estudo multicêntrico destaca a utilidade clínica do GPT-4 para a tomada de decisões de biópsia prostática e avança nosso entendimento das mais recentes implementações de inteligência artificial em vários cenários médicos.

Palavras-chave

Biópsia de próstata; Generative Pretrained Transformer-4 (GPT-4); Tomada de decisão; câncer de próstata; Ressonância magnética multiparamétrica (mpMRI)

References

  1. 1..
    Bergengren O,Pekala KR,Matsoukas K,Fainberg J,Mungovan SF,Bratt O, et al. 2022 update on prostate cancer epidemiology and risk factors—a systematic review. Eur Urol. 2023;84(2):191-206.
  2. 2..
    Culp MB,Soerjomataram I,Efstathiou JA,Bray F,Jemal A. Recent global patterns in prostate cancer incidence and mortality rates. Eur Urol.2020;77(1):38-52.
  3. 3..
    Xu H,Li YF,Yi XY,Zheng XN,Yang Y,Wang Y, et al. ADP-dependent glucokinase controls metabolic fitness in prostate cancer progression. Mil Med Res. 2023;10(1):64.
  4. 4..
    Constancio V,Lobo J,Sequeira JP,Henrique R,Jeronimo C. Prostate cancer epigenetics - from pathophysiology to clinical application. Nat Rev Urol.2025. .
  5. 5..
    Massanova M,Vere R,Robertson S,Crocetto F,Barone B,Dutto L, et al. Clinical and prostate multiparametric magnetic resonance imaging findings as predictors of general and clinically significant prostate cancer risk:a retrospective single-center study. Curr Urol. 2023;17(3):147-52.
  6. 6..
    Cornford P,Van Den Bergh RCN,Briers E,Van Den Broeck T,Brunckhorst O,Darraugh J, et al. EAU-EANM-ESTRO-ESUR-ISUP-SIOG guidelines on prostate cancer-2024 update. Part I: screening, diagnosis, and local treatment with curative intent. Eur Urol. 2024;86(2):148-63.
  7. 7..
    Ginsburg K,Cole AI,Silverman ME,Livingstone J,Smith DW,Heilbrun LK, et al. Should all prostate needle biopsy gleason score 4 + 4 = 8 prostate cancers be high risk? Implications for shared decision-making and patient counselling. Urol Oncol. 2020;38(3):78.e1-e6.
  8. 8..
    Lowenstein LM,Basourakos SP,Williams MD,Troncoso P,Gregg JR,Thompson TC, et al. Active surveillance for prostate and thyroid cancers:evolution in clinical paradigms and lessons learned. Nat Rev Clin Oncol.2019;16(3):168-84.
  9. 9..
    Borghesi M,Ahmed H,Nam R,Schaeffer E,Schiavina R,Taneja S, et al. Complications after systematic, random, and image-guided prostate biopsy. Eur Urol. 2017;71(3):353-65.
  10. 10..
    Devetzis K,Kum F,Popert R. Recent advances in systematic and targeted prostate biopsies. Res Rep Urol. 2021;13:799-809.
  11. 11..
    Van Poppel H,Hogenhout R,Albers P,Van Den Bergh RCN,Barentsz JO,Roobol MJ. Early detection of prostate cancer in 2020 and beyond: facts and recommendations for the European Union and the European Commission. Eur Urol. 2021;79(3):327-9.
  12. 12..
    Fütterer JJ,Briganti A,De Visschere P,Emberton M,Giannarini G,Kirkham A, et al. Can clinically significant prostate cancer be detected with mul-tiparametric magnetic resonance imaging? A systematic review of the literature. Eur Urol. 2015;68(6):1045-53.
  13. 13..
    Padhani AR,Weinreb J,Rosenkrantz AB,Villeirs G,Turkbey B,Barentsz J. Prostate imaging-reporting and data system steering committee: PI-RADS v2 status update and future directions. Eur Urol. 2019;75(3):385-96.
  14. 14..
    Barentsz JO,Richenberg J,Clements R,Choyke P,Verma S,Villeirs G, et al. ESUR prostate MR guidelines 2012. Eur Radiol. 2012;22(4):746-57.
  15. 15..
    Turkbey B,Rosenkrantz AB,Haider MA,Padhani AR,Villeirs G,Macura KJ, et al. Prostate Imaging Reporting and Data System Version 2.1: 2019 update of prostate imaging reporting and data system version 2. Eur Urol. 2019;76(3):340-51.
  16. 16..
    Weinreb JC,Barentsz JO,Choyke PL,Cornud F,Haider MA,Macura KJ, et al. PI-RADS prostate imaging - reporting and data system: 2015, version 2. Eur Urol. 2016;69(1):16-40.
  17. 17..
    Zhao LT,Liu ZY,Xie WF,Shao LZ,Lu J,Tian J, et al. What benefit can be obtained from magnetic resonance imaging diagnosis with artificial intelligence in prostate cancer compared with clinical assessments?. Mil Med Res. 2023;10(1):29.
  18. 18..
    Padhani AR,Barentsz J,Villeirs G,Rosenkrantz AB,Margolis DJ,Turkbey B, et al. PI-RADS Steering Committee: the PI-RADS multiparametric MRI and MRI-directed biopsy pathway. Radiology. 2019;292(2):464-74.
  19. 19..
    Renard-Penna R,Mozer P,Cornud F,Barry-Delongchamps N,Bruguière E,Portalez D, et al. Prostate imaging reporting and data system and likert scoring system: multiparametric MR imaging validation study to screen patients for initial biopsy. Radiology. 2015;275(2):458-68.
  20. 20..
    Barkovich EJ,Shankar PR,Westphalen AC. A systematic review of the existing prostate imaging reporting and data system version 2 (PI-RADSv2) literature and subset meta-analysis of PI-RADSv2 categories stratified by Gleason scores. AJR Am J Roentgenol. 2019;212(4):847-54.
  21. 21..
    Morash C. What do you do with PI-RADS-3?. Can Urol Assoc J.2021;15(4):122.
  22. 22..
    Nicola R,Bittencourt LK. PI-RADS 3 lesions: a critical review and discussion of how to improve management. Abdom Radiol (NY).2023;48(7):2401-5.
  23. 23..
    Boeken T,Feydy J,Lecler A,Soyer P,Feydy A,Barat M, et al. Artificial intelligence in diagnostic and interventional radiology: where are we now? Diagn Interv Imaging. 2023;104(1):1-5.
  24. 24..
    Bhayana R,Bleakney RR,Krishna S. GPT-4 in radiology: improvements in advanced reasoning. Radiology. 2023;307(5):e230987.
  25. 25..
    Cheng K,Li Z,Li C,Xie R,Guo Q,He Y, et al. The potential of GPT-4 as an AI-powered virtual assistant for surgeons specialized in joint arthro-plasty. Ann Biomed Eng. 2023;51(7):1366-70.
  26. 26..
    Kanjee Z,Crowe B,Rodman A. Accuracy of a generative artificial intelligence model in a complex diagnostic challenge. JAMA.2023;330(1):78-80.
  27. 27..
    Lee P,Bubeck S,Petro J. Benefits, limits, and risks of GPT-4 as an AI chatbot for medicine. N Engl J Med. 2023;388(13):1233-9.
  28. 28..
    Epstein JI. Prostate cancer grading: a decade after the 2005 modified system. Mod Pathol. 2018;31(S1):S47-63.
  29. 29..
    Pierorazio PM,Walsh PC,Partin AW,Epstein JI. Prognostic Gleason grade grouping: data based on the modified Gleason scoring system. BJU Int. 2013;111(5):753-60.
  30. 30..
    Patel HD,Koehne EL,Shea SM,Fang AM,Gerena M,Gorbonos A, et al. A prostate biopsy risk calculator based on MRI: development and comparison of the prospective loyola university multiparametric MRI(PLUM) and prostate biopsy collaborative group (PBCG) risk calculators. BJU Int. 2023;131(2):227-35.
  31. 31..
    Wang NN,Zhou SR,Chen L,Tibshirani R,Fan RE,Ghanouni P, et al. The stanford prostate cancer calculator: development and external validation of online nomograms incorporating PI-RADS scores to predict clinically significant prostate cancer. Urol Oncol. 2021;39(12):831.e19-e27.
  32. 32..
    Wei J,Wang X,Schuurmans D,Bosma M,Ichter B,Xia F, et al. Chain-of-thought prompting elicits reasoning in large language models. arXiv.2022.
  33. 33..
    Elkhoury FF,Felker ER,Kwan L,Sisk AE,Delfin M,Natarajan S, et al. Com-parison of targeted vs systematic prostate biopsy in men who are biopsy naive: the prospective assessment of image registration in the diagnosis of prostate cancer (PAIREDCAP) study. JAMA Surg. 2019;154(9):811-8.
  34. 34..
    Pepe P,Pennisi M. Targeted biopsy in men high risk for prostate cancer: 68Ga-PSMA PET/CT versus mpMRI. Clin Genitourin Canc.2023;21(6):639-42.
  35. 35..
    Mian BM,Tikkinen KaO,Kibel AS. Transrectal versus transperineal prostate biopsy: weighing the trade-offs. Lancet Oncol. 2025;26(5):533-5.
  36. 36..
    Pepe P,Pennisi M. Morbidity following transperineal prostate biopsy: our experience in 8.500 men. Arch Ital Urol Andro. 2022;94(2):155-9.
  37. 37..
    Ankerst DP,Straubinger J,Selig K,Guerrios L,De Hoedt A,Hernandez J, et al. A contemporary prostate biopsy risk calculator based on multiple heterogeneous cohorts. Eur Urol. 2018;74(2):197-203.
  38. 38..
    Ankerst DP,Hoefler J,Bock S,Goodman PJ,Vickers A,Hernandez J, et al. Prostate cancer prevention trial risk calculator 2.0 for the prediction of low- vs high-grade prostate cancer. Urology. 2014;83(6):1362-7.
  39. 39..
    Roobol MJ,Steyerberg EW,Kranse R,Wolters T,Van Den Bergh RCN,Bangma CH, et al. A risk-based strategy improves prostate-specific antigen-driven detection of prostate cancer. Eur Urol. 2010;57(1):79-85.
  40. 40..
    Orbe Villota PM,Leiva Centeno JA,Lugones J,Minuzzi PG,Varea SM. Comparison between the european randomized study for screening of prostate cancer (ERSPC) and prostate biopsy collaborative group (PBCG)risk calculators: prediction of clinically significant prostate cancer risk in a cohort of patients from argentina. Actas Urol Esp. 2024;48(3):210-7.
  41. 41..
    Shen J,Zhang CJP,Jiang B,Chen J,Song J,Liu Z, et al. Artificial intelligence versus clinicians in disease diagnosis: systematic review. JMIR Med Inf. 2019;7(3):e10010-e.
  42. 42..
    Wadden JJ. Defining the undefinable: the black box problem in health-care artificial intelligence. J Med Ethics. 2021;48(10):764-8.
  43. 43..
    Gugliandolo SG,Pepa M,Isaksson LJ,Marvaso G,Raimondi S,Botta F, et al. MRI-based radiomics signature for localized prostate cancer: a new clinical tool for cancer aggressiveness prediction? Sub-study of prospective phase II trial on ultra-hypofractionated radiotherapy (AIRC IG-13218). Eur Radiol. 2021;31(2):716-28.
  44. 44..
    Ponsiglione A,Gambardella M,Stanzione A,Green R,Cantoni V,Nappi C, et al. Radiomics for the identification of extraprostatic extension with prostate MRI: a systematic review and meta-analysis. Eur Radiol.2024;34(6):3981-91.
  45. 45..
    Woźnicki P,Westhoff N,Huber T,Riffel P,Froelich MF,Gresser E, et al. Multiparametric MRI for prostate cancer characterization: combined use of radiomics model with PI-RADS and clinical parameters. Cancers.2020;12(7):1767.
  46. 46..
    Vietri MT,D’elia G,Caliendo G,Resse M,Casamassimi A,Passariello L, et al. Hereditary prostate cancer: genes related, target therapy and prevention. Int J Mol Sci. 2021;22(7):3753.
  47. 47..
    Giri VN,Knudsen KE,Kelly WK,Abida W,Andriole GL,Bangma CH, et al. Role of genetic testing for inherited prostate cancer risk: phila-delphia prostate cancer consensus conference 2017. J Clin Oncol.2018;36(4):414-24.
  48. 48..
    Tuffaha H,Edmunds K,Fairbairn D,Roberts MJ,Chambers S,Smith DP, et al. Guidelines for genetic testing in prostate cancer: a scoping review. Prostate Cancer Prostatic Dis. 2024;27(4):594-603.
  49. 49..
    Casanova-Salas I,Athie A,Boutros PC,Del Re M,Miyamoto DT,Pienta KJ, et al. Quantitative and qualitative analysis of blood-based liquid biopsies to inform clinical decision-making in prostate cancer. Eur Urol.2021;79(6):762-71.
  50. 50..
    Hoffmann TJ,Graff RE,Madduri RK,Rodriguez AA,Cario CL,Feng K, et al. Genome-wide association study of prostate-specific antigen levels in 392, 522 men identifies new loci and improves prediction across ancestry groups. Nat Genet. 2025;57(2):334-44.

Leia o texto completo

The above content is generated by Large Model Translation. The translated content is for reference only. We do not assume any commercial or legal responsibilty for any consequences arising from the use of our website