Abstract
BackgroundThe imaging features of focal liver lesions (FLLs) are diverse and complex. Diagnosing FLLs with imaging alone remains challenging. We developed and validated an interpretable deep learning model for the classification of seven categories of FLLs on multisequence MRI and compared the differential diagnosis between the proposed model and radiologists.MethodsIn all, 557 lesions examined by multisequence MRI were utilised in this retrospective study and divided into training–validation (n = 444) and test (n = 113) datasets. The area under the receiver operating characteristic curve (AUC) was calculated to evaluate the performance of the model. The accuracy and confusion matrix of the model and individual radiologists were compared. Saliency maps were generated to highlight the activation region based on the model perspective.ResultsThe AUC of the two- and seven-way classifications of the model were 0.969 (95% CI 0.944–0.994) and from 0.919 (95% CI 0.857–0.980) to 0.999 (95% CI 0.996–1.000), respectively. The model accuracy (79.6%) of the seven-way classification was higher than that of the radiology residents (66.4%, p = 0.035) and general radiologists (73.5%, p = 0.346) but lower than that of the academic radiologists (85.4%, p = 0.291). Confusion matrices showed the sources of diagnostic errors for the model and individual radiologists for each disease. Saliency maps detected the activation regions associated with each predicted class.ConclusionThis interpretable deep learning model showed high diagnostic performance in the differentiation of FLLs on multisequence MRI. The analysis principle contributing to the predictions can be explained via saliency maps.
Highlights
With the development of imaging technology, most focal liver lesions (FLLs) can be detected accurately by magnetic resonance imaging (MRI) [1]
The inclusion criteria were as follows: (1) participants underwent unenhanced and enhanced liver MRI inspection; (2) participants had one of the following common FLLs, including liver cyst, cavernous haemangioma (HEM), hepatic abscess (HEP), focal nodular hyperplasia (FNH), hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (ICC) and hepatic metastasis (MET); and (3) up to one imaging study per patient was included, and up to six lesions were used in each study
Most malignant tumours were confirmed by histopathology, while other malignancies and benign tumours were diagnosed by follow-up reports that were supported by two radiologists for 3–12 months
Summary
With the development of imaging technology, most focal liver lesions (FLLs) can be detected accurately by MRI [1]. Diagnosing FLLs with imaging alone remains a challenge. Wang et al Insights into Imaging (2021) 12:173 diverse and complex, and different lesion features overlap. Atypical characteristics in some common lesions make the diagnosis challenging, including atypical morphologic features, atypical location or lesions that may mimic other primary liver tumours [2]. Maximising the imaging diagnosis accuracy of FLLs is paramount in avoiding unnecessary biopsies [3] and optimal patient management. The imaging features of focal liver lesions (FLLs) are diverse and complex. Diagnosing FLLs with imag‐ ing alone remains challenging. We developed and validated an interpretable deep learning model for the classifica‐ tion of seven categories of FLLs on multisequence MRI and compared the differential diagnosis between the pro‐ posed model and radiologists
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
