Retrospective Study
Copyright ©The Author(s) 2021. Published by Baishideng Publishing Group Inc. All rights reserved.
World J Gastroenterol. Sep 21, 2021; 27(35): 5978-5988
Published online Sep 21, 2021. doi: 10.3748/wjg.v27.i35.5978
Diagnosis of focal liver lesions with deep learning-based multi-channel analysis of hepatocyte-specific contrast-enhanced magnetic resonance imaging
Róbert Stollmayer, Bettina K Budai, Ambrus Tóth, Ildikó Kalina, Erika Hartmann, Péter Szoldán, Viktor Bérczi, Pál Maurovich-Horvat, Pál N Kaposi
Róbert Stollmayer, Bettina K Budai, Ambrus Tóth, Ildikó Kalina, Viktor Bérczi, Pál Maurovich-Horvat, Pál N Kaposi, Department of Radiology, Medical Imaging Centre, Faculty of Medicine, Semmelweis University, Budapest 1083, Hungary
Erika Hartmann, Department of Transplantation and Surgery, Faculty of Medicine, Semmelweis University, Budapest 1082, Hungary
Péter Szoldán, MedInnoScan Research and Development Ltd., Budapest 1112, Hungary
Author contributions: Stollmayer R designed and performed the research and wrote the paper; Budai BK and Tóth A contributed to data collection and analysis; Kalina I and Hartmann E provided clinical advice; Szoldán P contributed to the analysis; Bérczi V and Maurovich-Horvat P supervised the report; Kaposi PN designed the research, contributed to the analysis and supervised the report; all authors have read and approved the final manuscript.
Institutional review board statement: The present study has been approved by the institutional ethics committee of Semmelweis University (Semmelweis University Regional and Institutional Committee of Science and Research Ethics) according to the World Medical Association guidelines and Declaration of Helsinki, revised in 2000 in Edinburgh, No. SE-RKEB 136/2019.
Informed consent statement: As this is a retrospective study, in compliance with the Hungarian legal code, the need for written patient consent was waived by the ethics committee. Patients were not required to give informed consent to the study because the analysis used only anonymized clinical data that were obtained after each patient agreed to treatment and gave written informed consent to the MRI scan in compliance with our institutional protocol.
Conflict-of-interest statement: The authors have no financial relationships to disclose.
Data sharing statement: Additional anonymized data are available upon request from the corresponding author.
Open-Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution-NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See:
Corresponding author: Bettina K Budai, MD, Department of Radiology, Medical Imaging Centre, Faculty of Medicine, Semmelweis University, Korányi Sándor st. 2., Budapest 1083, Hungary.
Received: April 30, 2021
Peer-review started: April 30, 2021
First decision: June 23, 2021
Revised: July 7, 2021
Accepted: August 25, 2021
Article in press: August 25, 2021
Published online: September 21, 2021

The nature of input data is an essential factor when training neural networks. Research concerning magnetic resonance imaging (MRI)-based diagnosis of liver tumors using deep learning has been rapidly advancing. Still, evidence to support the utilization of multi-dimensional and multi-parametric image data is lacking. Due to higher information content, three-dimensional input should presumably result in higher classification precision. Also, the differentiation between focal liver lesions (FLLs) can only be plausible with simultaneous analysis of multi-sequence MRI images.


To compare diagnostic efficiency of two-dimensional (2D) and three-dimensional (3D)-densely connected convolutional neural networks (DenseNet) for FLLs on multi-sequence MRI.


We retrospectively collected T2-weighted, gadoxetate disodium-enhanced arterial phase, portal venous phase, and hepatobiliary phase MRI scans from patients with focal nodular hyperplasia (FNH), hepatocellular carcinomas (HCC) or liver metastases (MET). Our search identified 71 FNH, 69 HCC and 76 MET. After volume registration, the same three most representative axial slices from all sequences were combined into four-channel images to train the 2D-DenseNet264 network. Identical bounding boxes were selected on all scans and stacked into 4D volumes to train the 3D-DenseNet264 model. The test set consisted of 10-10-10 tumors. The performance of the models was compared using area under the receiver operating characteristic curve (AUROC), specificity, sensitivity, positive predictive values (PPV), negative predictive values (NPV), and f1 scores.


The average AUC value of the 2D model (0.98) was slightly higher than that of the 3D model (0.94). Mean PPV, sensitivity, NPV, specificity and f1 scores (0.94, 0.93, 0.97, 0.97, and 0.93) of the 2D model were also superior to metrics of the 3D model (0.84, 0.83, 0.92, 0.92, and 0.83). The classification metrics of FNH were 0.91, 1.00, 1.00, 0.95, and 0.95 using the 2D and 0.90, 0.90, 0.95, 0.95, and 0.90 using the 3D models. The 2D and 3D networks' performance in the diagnosis of HCC were 1.00, 0.80, 0.91, 1.00, and 0.89 and 0.88, 0.70, 0.86, 0.95, and 0.78, respectively; while the evaluation of MET lesions resulted in 0.91, 1.00, 1.00, 0.95, and 0.95 and 0.75, 0.90, 0.94, 0.85, and 0.82 using the 2D and 3D networks, respectively.


Both 2D and 3D-DenseNets can differentiate FNH, HCC and MET with good accuracy when trained on hepatocyte-specific contrast-enhanced multi-sequence MRI volumes.

Keywords: Artificial intelligence, Multi-parametric magnetic resonance imaging, Hepatocyte-specific contrast, Densely connected convolutional network, Hepatocellular carcinoma, Focal nodular hyperplasia

Core Tip: Our study aimed to assess the performance of two-dimensional (2D) and three-dimensional (3D) densely connected convolutional neural networks (DenseNets) in the classification of focal liver lesions (FLLs) based on multi-parametric magnetic resonance imaging (MRI) with hepatocyte-specific contrast. We used multi-channel data input to train our networks and found that both 2D and 3D-DenseNets can differentiate between focal nodular hyperplasias, hepatocellular carcinomas or liver metastases with excellent accuracy. We conclude that DensNets can reliably classify FLLs based on multi-parametric and hepatocyte-specific post-contrast MRI. Meanwhile, multi-channel input is advantageous when the number of clinical cases available for model training is limited.