External validation of a machine learning based algorithm to differentiate hepatic mucinous cystic neoplasms from benign hepatic cysts
Carregando...
Citações na Scopus
1
Tipo de produção
article
Data de publicação
2023
Título da Revista
ISSN da Revista
Título do Volume
Editora
SPRINGER
Autores
FURTADO, Felipe S. S.
BADENES-ROMERO, Alvaro
HESAMI, Mina
MOSTAFAVI, Leila
NAJMI, Zahra
MOJTAHED, Amirkasra
ANDERSON, Mark A. A.
CATALANO, Onofrio A. A.
Citação
ABDOMINAL RADIOLOGY, v.48, n.7, p.2311-2320, 2023
Resumo
Purpose To externally validate an algorithm for non-invasive differentiation of hepatic mucinous cystic neoplasms (MCN) from benign hepatic cysts (BHC), which differ in management. Methods Patients with cystic liver lesions pathologically confirmed as MCN or BHC between January 2005 and March 2022 from multiple institutions were retrospectively included. Five readers (2 radiologists, 3 non-radiologist physicians) independently reviewed contrast-enhanced CT or MRI examinations before tissue sampling and applied the 3-feature classification algorithm described by Hardie et al. to differentiate between MCN and BHC, which had a reported accuracy of 93.5%. The classification was then compared to the pathology results. Interreader agreement between readers across different levels of experience was evaluated with Fleiss' Kappa. Results The final cohort included 159 patients, median age of 62 years (IQR [52.0, 70.0]), 66.7% female (106). Of all patients, 89.3% (142) had BHC, and the remaining 10.7% (17) had MCN on pathology. Agreement for class designation between the radiologists was almost perfect (Fleiss' Kappa 0.840, p < 0.001). The algorithm had an accuracy of 98.1% (95% CI [94.6%, 99.6%]), a positive predictive value of 100.0% (95% CI [76.8%, 100.0%]), a negative predictive value of 97.9% (95% CI [94.1%, 99.6%]), and an area under the receiver operator characteristic curve (AUC) of 0.911 (95% CI [0.818, 1.000]). Conclusion The evaluated algorithm showed similarly high diagnostic accuracy in our external, multi-institutional validation cohort. This 3-feature algorithm is easily and rapidly applied and its features are reproducible among radiologists, showing promise as a clinical decision support tool. [GRAPHICS] .
Palavras-chave
Neoplasms, cystic, mucinous, and serous, Magnetic resonance imaging, Cystadenocarcinoma, Cystadenoma, Algorithms, Liver
Referências
- Anderson MA, 2023, ABDOM RADIOL, V48, P79, DOI 10.1007/s00261-021-03303-5
- Anderson MA, 2021, CLIN IMAG, V76, P46, DOI 10.1016/j.clinimag.2021.01.036
- Arnaoutakis DJ, 2015, ANN SURG, V261, P361, DOI 10.1097/SLA.0000000000000543
- Austin PC, 2017, STAT METHODS MED RES, V26, P796, DOI 10.1177/0962280214558972
- Bakoyiannis A, 2013, WORLD J GASTROENTERO, V19, P7603, DOI 10.3748/wjg.v19.i43.7603
- Bleeker SE, 2003, J CLIN EPIDEMIOL, V56, P826, DOI 10.1016/S0895-4356(03)00207-5
- Boyum JH, 2021, AM J ROENTGENOL, V216, P403, DOI 10.2214/AJR.20.22768
- Boyum JH, 2016, MAYO CLIN PROC, V91, P329, DOI 10.1016/j.mayocp.2015.11.015
- Chi H, 2017, EUR J GASTROEN HEPAT, V29, P36, DOI 10.1097/MEG.0000000000000731
- Choi HK, 2010, J CLIN GASTROENTEROL, V44, P289, DOI 10.1097/MCG.0b013e3181b5c789
- Dong Y, 2017, SCAND J GASTROENTERO, V52, P365, DOI 10.1080/00365521.2016.1259652
- Dua MM, 2016, DIGEST DIS SCI, V61, P1835, DOI 10.1007/s10620-015-3943-y
- Halimu C, 2019, PROCEEDINGS OF THE 3RD INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND SOFT COMPUTING (ICMLSC 2019), P1, DOI 10.1145/3310986.3311023
- Hardie AD, 2022, ACAD RADIOL, V29, P1149, DOI 10.1016/j.acra.2021.08.025
- Kim HJ, 2014, AM J ROENTGENOL, V202, P83, DOI 10.2214/AJR.12.9170
- Kim JY, 2010, AM J ROENTGENOL, V195, P1142, DOI 10.2214/AJR.09.4026
- Kovacs MD, 2018, CLIN IMAG, V49, P44, DOI 10.1016/j.clinimag.2017.10.022
- LANDIS JR, 1977, BIOMETRICS, V33, P159, DOI 10.2307/2529310
- Lantinga MA, 2013, WORLD J GASTROENTERO, V19, P3543, DOI 10.3748/wjg.v19.i23.3543
- Le Don K, 2019, Proc (Bayl Univ Med Cent), V32, P124, DOI 10.1080/08998280.2018.1520623
- Marrero JA, 2014, AM J GASTROENTEROL, V109, P1328, DOI 10.1038/ajg.2014.213
- Mavilia MG, 2018, J CLIN TRANSL HEPATO, V6, P208, DOI 10.14218/JCTH.2017.00069
- McHugh ML, 2012, BIOCHEM MEDICA, V22, P276, DOI 10.11613/bm.2012.031
- Myers RP, 2008, LIVER INT, V28, P705, DOI 10.1111/j.1478-3231.2008.01691.x
- Park HJ, 2018, AM J ROENTGENOL, V211, P67, DOI 10.2214/AJR.17.19261
- R Core Team, 2020, R LANG ENV STAT COMP
- Robin X, 2011, BMC BIOINFORMATICS, V12, DOI 10.1186/1471-2105-12-77
- Rodriguez RM, 2019, J SURG CASE REP, DOI 10.1093/jscr/rjz364
- RStudio Team, 2019, RSTUDIO INT DEV R
- Seo JK, 2010, EUR J GASTROEN HEPAT, V22, P989, DOI 10.1097/MEG.0b013e328337c971
- Shah S, 2021, POL J RADIOL, V86, pE440, DOI 10.5114/pjr.2021.108257
- Soni Sweta, 2021, Med Pharm Rep, V94, P366, DOI 10.15386/mpr-1543
- Steyerberg EW, 2001, J CLIN EPIDEMIOL, V54, P774, DOI 10.1016/S0895-4356(01)00341-9
- Steyerberg EW, 2016, J CLIN EPIDEMIOL, V69, P245, DOI 10.1016/j.jclinepi.2015.04.005
- Vachha B, 2011, AM J ROENTGENOL, V196, pW355, DOI 10.2214/AJR.10.5292
- Xu H-X, 2012, Ultraschall Med, V33, pE241, DOI 10.1055/s-0031-1299276
- Yan WJ, 2020, RADIOL-ARTIF INTELL, V2, DOI 10.1148/ryai.2020190195
- Yin J, 2020, J X-RAY SCI TECHNOL, V28, P683, DOI 10.3233/XST-200675
- Zech JR, 2018, PLOS MED, V15, DOI 10.1371/journal.pmed.1002683