Multimodal Ultrasound Radiomics in Liver Disease: Current Status and Future Directions

doi:10.26599/AUDT.2025.250101

Advanced Ultrasound in Diagnosis and Therapy ›› 2025, Vol. 9 ›› Issue (4): 388-408.doi: 10.26599/AUDT.2025.250101

收稿日期:2025-10-15 修回日期:2025-10-28 接受日期:2025-11-05 出版日期:2025-12-30 发布日期:2025-11-06

Multimodal Ultrasound Radiomics in Liver Disease: Current Status and Future Directions

Zhong Xian^a, Xie Xiaoyan^a^,^*()

^aDepartment of Medical Ultrasonics, Institute of Diagnostic and Interventional Ultrasound, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China

Received:2025-10-15 Revised:2025-10-28 Accepted:2025-11-05 Online:2025-12-30 Published:2025-11-06
Contact: Department of Medical Ultrasonics, Institute of Diagnostic and Interventional Ultrasound, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou 510080, China (Xiaoyan Xie), e-mail: xiexyan@mail.sysu.edu.cn (XY X).

摘要/Abstract

Abstract:

Multimodal ultrasound, including B-mode imaging, contrast-enhanced ultrasound (CEUS), and ultrasound-based elastography, has demonstrated significant value in evaluating both diffuse liver diseases such as fibrosis and steatosis, and focal liver lesions such as hepatocellular carcinoma (HCC). Radiomics, including both handcrafted radiomics and deep learning approaches, has emerged as a promising strategy to enhance ultrasound-based liver disease assessment. Recent studies have applied radiomics across multimodal ultrasound, achieving notable success in grading fatty liver disease, staging fibrosis, and improving diagnosis, risk stratification, and prognostic prediction in HCC. Multimodal ultrasound provides complementary information on liver morphology, perfusion, and stiffness, while fusion strategies further enhance diagnostic accuracy and robustness. Future efforts should focus on standardized, large-scale multicenter validation, methodological improvements in multimodal integration, and the incorporation of explainable artificial intelligence to support clinical translation. Ultimately, despite ongoing challenges related to data heterogeneity, reproducibility, interpretability, and clinical validation, multimodal ultrasound radiomics holds strong promise for noninvasive, individualized, and clinically meaningful liver disease management.

Key words: Handcrafted radiomics, Deep learning, Contrast-enhanced ultrasound, Shear wave elastography, Liver disease

. [J]. Advanced Ultrasound in Diagnosis and Therapy, 2025, 9(4): 388-408.

Zhong Xian, Xie Xiaoyan. Multimodal Ultrasound Radiomics in Liver Disease: Current Status and Future Directions[J]. Advanced Ultrasound in Diagnosis and Therapy, 2025, 9(4): 388-408.

图/表 7

Reference	Study design	Imaging modality	No. of patients/ images	Radiomics technique	Detailed method	Task	Reference standard	Results
Reference	Study design	Imaging modality	No. of patients/ images	Radiomics technique	Detailed method	Task	Reference standard	AUC	Accuracy	Sensitivity	Specificity
HR, handcrafted radiomics; DL, deep learning; FLD, fatty liver disease; CAP, controlled attenuation parameter; CNN, convolutional neural networks; MRI-PDFF, magnetic resonance imaging derived proton density fat fraction; ORF, original radio frequency signal; CDFI, Color doppler flow imaging
Wu 2022 [38]	Retrospective, single center	B-mode	321 images from 235 patients	HR	Different classifiers	FLD diagnosis	CAP and biopsy	0.75	70	66.4	NA
Cao 2020 [33]	Retrospective, single center	B-mode	240 images from 240 patients	DL	CNN	FLD diagnosis and severity classification	US evaluation by radiologists	0.933-0.958
Byra 2018 [40]	Retrospective, single center	B-mode	550 images from 55 patients	DL	Inception-ResNet-v2+SVM	FLD diagnosis	Liver biopsy	0.977	96.3	100	88.2
Reddy 2018 [41]	Retrospective, single center	B-mode	157 images	DL	VGG16 Transfer Learning	FLD diagnosis	US evaluation by radiologists	0.96	90.6	95	85
Kim 2021 [42]	Retrospective, single center	B-mode	180 images from 90 cases	DL	VGG19	FLD diagnosis	MRI–PDFF	0.87	80.1	NA	80.5
Han 2020 [43]	Prospective, single center	ORF	204 patients	DL	One-dimensional CNN	FLD diagnosis and fat fraction estimation	MRI–PDFF	0.98	96	97	94
Chou 2021 [44]	Retrospective, single center	B-mode	21855 images from 2070 patients	DL	ResNet-50 v2	FLD diagnosis and severity classification	US evaluation by radiologists	0.971-0.996
Tahmasebi 2023 [45]	Prospective, single center	B-mode	1435 images from 130 patients	DL	Google AutoML	FLD diagnosis	MRI–PDFF	NA	83.4	72.2	94.6
Rhyou 2021 [21]	Retrospective, multicenter	B-mode	3200 images from 1902 patients'	DL	Cascaded Neural Network	FLD diagnosis and severity classification	US evaluation by radiologists	NA	99.6	99.1	100
Yang 2023 [46]	Retrospective, single center	B-mode	1856 images from 928 patients	DL	Two-section neural network	FLD diagnosis and severity classification	US evaluation by radiologists	0.84-0.93	76.3
Liu 2024 [47]	Prospective, single center	B-mode +CDFI	710 images from 710 patients	DL+HR	VGG16 model and two HR features	FLD diagnosis and severity classification	NA	0.836-0.947	77.5

Reference	Study design	Imaging modality	No. of patients/images	Radiomics technique	Detailed method	Task	Reference standard	Results
Reference	Study design	Imaging modality	No. of patients/images	Radiomics technique	Detailed method	Task	Reference standard	AUC	Accuracy	Sensitivity	Specificity
HR, handcrafted radiomics; DL, deep learning; TE, transient elastography; CNN, convolutional neural networks; GAN, generative adversarial network; ORF, original radio frequency signal; SWE, shear wave elastography; SVM, support vector machines; LSM, liver stiffness measurement; STQ, sound touch quantification; STE, sound touch elastography
Meng 2017 [48]	Retrospective, single center	B-mode	279 patients	DL	Transfer Learning and FCNet	Liver fibrosis classification	NA		93.9
Lee 2020 [49]	Retrospective, single center	B-mode	3446 patients	DL	DCNN	Liver fibrosis classification	Pathology or TE	0.857 for cirrhosis prediction	88.3	77.8	93.7
Feng 2021 [50]	Retrospective, single center	B-mode	286 patients	DL	Pyramid-structured CNN	Liver fibrosis classification	Liver biopsy	0.99	95.7	96.5
Ruan 2021 [51]	Retrospective, multicenter	B-mode	508 patients	DL	Multi-scale texture network (MSTNet)	Liver fibrosis classification	Liver biopsy	0.92 (≥F2) 0.89 (F4)		85.1 (≥F2) 87.8 (F4)	87.6 (≥F2) 78.1 (F4)
Duan 2022 [52]	Retrospective, multicenter	B-mode	434 patients	DL+HR	DL features generated by GAN and HR features	Liver fibrosis screening	Liver biopsy	0.861 for cirrhosis prediction
Joo 2023 [53]	Retrospective, multicenter	B-mode	955 patients	DL	Transfer learning of different models	Liver fibrosis classification	Liver biopsy	0.8592
Park 2024 [54]	Retrospective, multicenter	B-mode	933 patients	DL	EfficientNet	Liver fibrosis classification	Liver surgery and biopsy	0.94-0.96 for F0-F4	93-96	79-89	95-98
Ai 2024 [55]	Retrospective, single center	ORF	237 patients	DL	2D CNN for segmentation and 1D CNN for classification	Liver fibrosis classification	Liver biopsy	0.957 (≥F1) 0.729 (≥F3) 0.876 (≥F4)
Zhang 2025 [56]	Retrospective, multicenter	high-frequency B-mode images	1500 patients	DL	InceptionNeXt	Liver fibrosis classification	Liver biopsy	0.81 (≥S2) 0.93 (≥S3) 0.87 (S4)	74 (≥S2) 85 (≥S3) 84 (S4)	84 (≥S2) 85 (≥S3) 79 (S4)	65 (≥S2) 86 (≥S3) 85 (S4)
Chen 2017 [57]	Retrospective, multicenter	Real-time tissue elastography	513 patients	HR	Random Forest	Liver fibrosis classification	Liver biopsy		82	75	86
Gatos 2016 [58]	Retrospective, single center	SWE	85 patients	HR	Color to stiffness mapping and SVM	Liver fibrosis classification	Liver biopsy	0.85	87	83.3	89.1
Gatos 2017 [59]	Retrospective, single center	SWE	126 patients	HR	Stiffness value clustering and SVM	Liver fibrosis classification	Liver biopsy	0.87	87.3	93.5	81.2
Gatos 2019 [60]	Retrospective, single center	SWE	200 patients	DL	A wavelet transform and fuzzy c-means clustering algorithm to detect areas of high and low stiffness temporal stability	Liver fibrosis classification	Liver biopsy	0.89-0.95	82.5-92.5
Kagadis 2020 [61]	Retrospective, single center	SWE	200 patients	DL	Transfer learning of different models	Liver fibrosis classification	Liver biopsy	0.979-0.990	87.2-97.4
Wang 2018 [33]	Prospective, multicenter	SWE	398 patients	DL	Deep learning radiomics of elastography (DLRE)	Liver fibrosis classification	Liver biopsy	0.85 (≥F2) 0.98 (≥F3) 0.97 (F4)		69.1 (≥F2) 90.4 (≥F3) 96.9 (F4)	90.9 (≥F2) 98.3 (≥F3) 88.0 (F4)
Meng 2023 [62]	Retrospective, single center	B-mode, SWE, clinical parameters	618 patients	HR	SVM	Predicting the risk of fibrosis Progression NAFLD	LSM values	0.95		86.2
Li 2019 [35]	Prospective, single center	B-mode, ORF and contrast-enhanced micro-flow (CEMF)	144 patients	HR	Multiparametric model	Diagnosis of significant liver fibrosis	Liver surgery and biopsy	0.78-0.85		87.5-93.8	69.2-76.9
Liu 2022 [63]	Retrospective, single center	B-mode, liver stiffness values, and clinical parameters	284 patients	DL	deep learning-based data integration network (DI-Net)	Diagnosis of significant liver fibrosis	Liver surgery and biopsy	0.901	81.3	81.6	80.8
Liu 2023 [34]	Retrospective, single center	B-mode, Contrast-enhanced microflow (CEMF) cines and clinical parameters	218 patients	DL	Data integration-based deep learning (DIDL)	Diagnosis of significant liver fibrosis	Liver surgery and biopsy	0.901	81.3	81.6	80.9
Gao 2021 [64]	Retrospective, single center	B-mode, STQ, STE	168 patients	DL	Multi-modal fusion network with AL (MMFN-AL)	Liver fibrosis classification	Liver biopsy	0.891	70.59
Xue 2020 [8]	Retrospective, multicenter	B-mode, SWE	466 patients	DL	Transfer learning of Inception-V3	Liver fibrosis classification	Liver surgery	0.93 (≥S2) 0.93 (≥S3) 0.95 (S4)		90.0 (≥S2) 89.9 (≥S3) 90.1 (S4)	87.8 (≥S2) 87.9 (≥S3) 94.3 (S4)
Lu 2021 [65]	Retrospective, multicenter	B-mode, SWE, clinical parameters	807 patients	DL	Multichannel deep learning radiomics models (DLRE 2.0)	Diagnosis of significant liver fibrosis	Liver biopsy	0.95		90.6	90.1
Chen 2024 [66]	Retrospective, multicenter	B-mode, SWE	5894 patients	DL	ResNet152 for liver stiffness prediction and sequential algorithm for liver fibrosis screening	Liver fibrosis screening	Liver biopsy		85	54	94

Reference	Study design	Imaging modality	No. of patients/ images	Radiomics technique	Detailed method	Task	Reference standard	Results
Reference	Study design	Imaging modality	No. of patients/ images	Radiomics technique	Detailed method	Task	Reference standard	AUC	Accuracy	Sensitivity	Specificity
HR, handcrafted radiomics; DL, deep learning; FLL, focal liver lesions; HCC, hepatocellular carcinoma; ICC, intrahepatic cholangiocarcinoma; FFS, focal fat sparing; FFI, focal fat infiltration; MLC, metastatic liver cancer; TIC, time-intensity curve; CEUS, contrast-enhanced ultrasound; CNN, convolutional neural networks; FNH, focal nodular hyperplasia; SWE, shear wave elastography; SWV, shear wave velocity
Hwang 2015 [67]	Retrospective, NA	B-mode	115 patients	HR	HR features with artificial neural network	FLL diagnosis (cysts, hemangiomas, malignancies)	NA
Mao 2021 [68]	Retrospective, single center	B-mode	114 patients	HR	Different classifiers	Classification of primary and metastatic liver cancer	Pathology	0.816	84.3	76.8	88
Peng 2020 [69]	Retrospective, single center	B-mode and clinical variables	531 patients	HR	Different classifiers	Differentiation among HCC and ICC/cHCC-ICC	Pathology	0.728-0.775
Qin 2020 [70]	Retrospective, single center	B-mode	254 patients	HR	Different classifiers	Identification of primary tumorous sources of liver metastases	Pathology	0.750-0.768	70.6; 79.2; 64.3	73.9; 70.0; 75.0	67.9; 85.7; 50.0
Peng 2022 [71]	Retrospective, single center	B-mode	589 patients	HR	Different classifiers	Differentiating infected focal liver lesions from malignancy	Pathology or follow up/imaging	0.745-0.836	66.7-78.4	65.4-87.7	45.2-67.7
Schmauch 2019 [72]	NA,	B-mode	544 patients	DL	algorithm was trained using an attention with annotations	FLL diagnosis (benign and malignant)	NA	0.812-0.922
Xi 2021 [73]	Retrospective, single center	B-mode	596 patients	DL	Resnet	FLL diagnosis (benign and malignant)	MRI or histopathology	0.85	80	91	62
Tiyarattanachai 2021 [74]	Retrospective, multicenter	B-mode	22472 lesions	DL	RetinaNet for both detection and diagnosis	FLL detection and diagnosis (HCC, cyst, hemangioma, FFS and FFI)	Pathology and/or MRI/CT		95.3	84.9	97.1
Chen 2024 [75]	Retrospective, single center	B-mode	465 patients	DL	Different DCNN	Differentiation among HCC, ICC, and cHCC-ICC	Pathology	0.92	86	84.59	92.65
Yang 2020 [76]	Retrospective, multicenter	B-mode, clinical variables	2143 patients	DL	Resnet 18	FLL diagnosis (benign and malignant)	Pathology	0.924	84.7	86.5	85.5
Yang 2023 [77]	Retrospective, multicenter	B-mode	6784 patients	DL	Resnet 50	Identification of hepatic echinococcosis	Pathological or clinical diagnosis	0.913-0.982 (echinococcosis) 0.900-0.986 (alveolar echinococcosis)	71.9-84; 90.7-94.6	92.1-100; 77.1-97.1	66.6-80.5; 92.4-94.2
Du 2025 [78]	Retrospective, multicenter	B-mode and clinical parameters	1052 patients	HR	XGBoost	classification of ≤3 cm HCC	Pathology or CT/MRI with follow-up	0.899	85.9	92.8	77.9
Streba 2012 [79]	Prospective, single center	CEUS	112 patients	HR and DL	TIC features	FLL diagnosis (HCC, MLC, hepatic hemangiomas, local fatty changes)	Pathology or CT/MRI or follow up	0.89	87.1	93.2	89.7
Gatos 2015 [80]	Retrospective, single center	CEUS	52 patients	HR	TIC features	FLL detection and diagnosis (benign and malignant)	Pathology or CT/MRI	0.89	90.3	93.1	86.9
Kondo 2017 [81]	Retrospective, single center	CEUS	98 patients	HR	TIC features	FLL diagnosis (benign and malignant, differentiation of benign, HCC, or MLC)	Pathology or CT/MRI with follow-up		91.8	94	87.1
Turco 2022 [82]	Retrospective, single center	CEUS	72 patients	HR	TIC features and textural features	FLL diagnosis (benign and malignant)	Pathology and/or MRI/CT	0.84	84	76	92
Guo 2018 [83]	Retrospective, single center	CEUS	93 patients	HR	Deep canonical correlation analysis (DCCA) and multiple kernel learning (MKL)	FLL diagnosis (benign and malignant)	Pathology or CT/MRI		90.4	93.6	86.9
Li 2024 [84]	Retrospective, single center	CEUS, clinical features	159 patients	HR	Textual features from 6 CEUS images	Differentiation among HCC and ICC/cHCC-ICC in LR-M patients	Pathology	0.912	85.4	95.2	77.8
Wang 2024 [85]	Prospective, multicenter	CEUS, clinical features	116 patients	HR	HR features from one Kupffer phase image	Differentiating well-differentiated hepatocellular carcinoma (w-HCC) from atypical FLL	Pathology	0.912		82.4	85
Pan 2019 [86]	Retrospective, single center	CEUS	242 tumors	DL	3D-CNN	FLL diagnosis (HCC and FNH)	Pathology	0.97	93.1	94.5	93.6
Caleanu 2021 [87]	Retrospective, single center	CEUS	91 patients	DL	Different DCNN	FLL diagnosis (HCC, MLC, hemangiomas, FNH)	NA		88
Ding 2025 [88]	Retrospective, multicenter	CEUS	3725 lesions	DL	Module-Disease, Module-Biomarker and Module-Clinic were built and aggregated	FLL diagnosis (HCC, MLC, ICC, hepatic hemangioma, hepatic abscess and others)	Pathology for malignancy or MRI with follow-up for benign lesions	0.86		85	97
Yao 2018 [89]	Retrospective, single center	B-mode, SWE, SWV	177 patients	HR	Sparse representation theory (SRT)	FLL diagnosis (benign and malignant)	Pathology	0.94	88	91	86
Hu 2024 [90]	Retrospective, single center	B-mode, CEUS	527 patients	HR	HR features extracted from both B-mode images and three-phase CEUS images	FLL diagnosis (benign and malignant)	Pathology for malignant lesions; CEUS and follow-up for benign lesions	0.914		95.2	72.2
Su 2024 [91]	Retrospective, single center	B-mode, CEUS, clinical features	280 patients	HR	HR features from both B-mode and CEUS images	Differentiation of HCC and ICC	Pathology	0.97	91	89	93
Hu 2021 [92]	Retrospective, single center	B-mode, CEUS	574 patients	DL	ResNet on four-phase images	FLL diagnosis (benign and malignant)	Pathology for malignant lesions; CEUS and follow-up for benign lesions	0.934	91	92.7	85.1
Liu 2022 [93]	Retrospective, multicenter	B-mode, CEUS, clinical features	303 patients	DL	Four Stream (B-mode, CEUS and their optic flow) 3D convolutional neural network	FLL diagnosis (benign and malignant)	Pathology	0.957	94	96.6	90.5

Reference	Study design	Imaging modality	No. of patients/images	Radiomics technique	Detailed method	Task	Reference standard	Results
Reference	Study design	Imaging modality	No. of patients/images	Radiomics technique	Detailed method	Task	Reference standard	AUC	Accuracy	Sensitivity	Specificity
ORF, original radio frequency signal; HR, handcrafted radiomics; DL, deep learning; CEUS, contrast-enhanced ultrasound; MVI, microvascular invasion; CNN, convolutional neural networks; SWE, shear wave elastography
Dong 2019 [94]	Prospective, single center	ORF	42 patients	HR	Signal analysis and processing technology in feature extraction	MVI prediction	Pathology	0.95	92.8	85.7	100
Dong 2020 [95]	Retrospective, single center	B-mode, clinical variables	322 patients	HR	Gross- and peri-tumoral region HR	MVI prediction	Pathology	0.744	63.4	89.2	48.4
Hu 2019 [96]	Retrospective, single center	B-mode, clinical variables	482 patients	HR	Radiomics features combined with clinical variables	MVI prediction	Pathology	0.731
Dong 2022 [97]	Prospective, single center	CEUS, clinical variables	100 patients	HR	Radiomics features from peritumoral liver tissues of Kupffer phase	MVI prediction	Pathology	0.804	75	87.5	69.1
Zhang 2021 [98]	Retrospective, single center	B-mode, CEUS, clinical variables	313 patients	HR	HR features from different CEUS phase	MVI prediction	Pathology	0.788	72.7	75.5	70.8
Zhang 2022 [99]	Retrospective, single center	CEUS, clinical variables	436 patients	DL	GRU for temporal features and CNN for spatial features	MVI prediction	Pathology	0.865	78.8	83.3	81
Qin 2023 [100]	Retrospective, single center	CEUS	252 patients	DL	ResNet50+SE	MVI prediction	Pathology	0.856	77.2	52.4	93.9
Qin 2025 [101]	Retrospective, single center	CEUS	164 patients	DL	Transformer+Resnet	MVI prediction	Pathology	0.859	71.4	86.2	80
Wang 2025 [102]	Retrospective, single center	CEUS	318 patients	DL	Graph Convolutional Network	MVI prediction	Pathology	0.928	89.3	85.3	91.7
Zheng 2023 [103]	Retrospective, single center	CEUS, MRI	85 patients	HR	Comparison between CEUS and MRI radiomics model	MVI prediction	Pathology	0.86		100	85.7
Zhang 2024 [104]	Retrospective, multicenter	B-mode, CEUS, clinical variables	576 patients	HR, DL	Comparison of DL and HR models with different conditions	MVI prediction	Pathology	0.738	71	60.6	76.5
Li 2025 [105]	Experiments on rabbits	3D ultrasound	9 rabbits	HR	3D US images are fused with whole-slide images to localize MVI regions	MVI prediction	Pathology	0.91	86	76	92
Ren 2021 [106]	Retrospective, multicenter	B-mode, clinical variables	193 patients	HR	combination of HR features and clinical variables	Differentiation prediction	Pathology	0.849	81.8	75	85.7
Qin 2023 [107]	Retrospective, single center	CEUS	272 patients	DL+HR	Combination of both DL and HR features	Differentiation prediction	Pathology	0.932	91.5	93.8	90
Li 2022 [108]	Retrospective, single center	CEUS	54 patients	HR	Radiomics features from Kupffer phase	Differentiation prediction	Pathology	0.878-0.938	97.1 97.1	93.3 83.3	98.1 100
Qian 2023 [109]	Retrospective, single center	B-mode	118 patients	HR	HR features from intratumoral and peritumoral regions	Ki-67 prediction	Pathology	0.87	80.6	73.7	88.2
Zhang 2024 [110]	Retrospective, single center	B-mode, CEUS, clinical variables	310 patients	HR	HR features from different CEUS phase	Ki-67 prediction	Pathology	0.856	76.8	79.3	74.1
Yao 2018 [89]	Retrospective, single center	B-mode, SWE, SWV	177 patients	HR	Sparse representation theory (SRT)	PD-1, Ki-67, MVI prediction	Pathology	0.94-0.98	92; 93; 95	100; 95; 91	88; 89; 100
Qian 2024 [111]	Retrospective, single center	B-mode	153 patients	HR	Different classifiers	Prediction of differentiation, CK-7, Ki-67 and p53	Pathology	0.762-0.922	82.1; 87.5; 75.0; 70.3	78.2; 100; 85.7 100
Bu 2025 [112]	Retrospective, single center	B-mode, clinical parameters	154 patients	HR	Different classifiers	Predicting TP53 mutation	Pathology	0.846	82.3	80.6	83.9
Liang 2025 [113]	Retrospective, single center	B-mode, CEUS, clinical variables	434 patients	HR	HR features from B-mode, arterial, portal venous, and delayed phases	CK-19 prediction	Pathology	0.927	86.9	89.3	86.3

Reference	Study design	Imaging modality	No. of patients/images	Radiomics technique	Detailed method	Task	Reference standard	Results
Reference	Study design	Imaging modality	No. of patients/images	Radiomics technique	Detailed method	Task	Reference standard	AUC	C-index	Accuracy	Sensitivity	Specificity
ER, early recurrence; LR, late recurrence; RFS, recurrence free survival; MWA, microwave ablation; HR, handcrafted radiomics; DL, deep learning; CEUS, contrast-enhanced ultrasound; HCC, hepatocellular carcinoma; SR, surgical resection; RFA, radiofrequency ablation; PHLF, post-hepatectomy liver failure; ISGLS, international study group of liver surgery
Huang 2021 [114]	Retrospective, single center	B-mode, CEUS	215 patients	HR	Combined features from both tumoral and peritumoral area in different CEUS phase	ER prediction after resection or ablation	follow up	0.845		84.2	86.7	82.6
Cao 2024 [115]	Retrospective, single center	B-mode, clinical variables	127 patients	HR	Logistic regression	ER prediction following surgical resection	follow up	0.925			77.8	100
Wu 2022 [116]	Retrospective, single center	B-mode	513 patients	DL+HR	Comparison of DL and HR features	ER, LR, RFS after MWA and differentiation prediction	follow up		0.695 (ER); 0.715 (LR); 0.721 (RFS)
Zhang 2022 [117]	Retrospective, single center	B-mode, CEUS,	172 patients	DL+HR	Combination of DL and HR	ER prediction following surgical resection	follow up	0.889		78.4	90	66.7
Huang 2022 [118]	Retrospective, single center	B-mode, CEUS, clinical features	414 patients	DL+HR	Combination of both HR and DL features	ER and survival prediction after resection	follow up	0.57 (ER)	0.759 (OS)	59 (ER)	62 (ER)	56 (ER)
Ma 2021 [119]	Retrospective, single center	B-mode, SWE, clinical features	318 patients	DL+HR	Combining CEUS, US radiomics, and clinical factors	ER and LR prediction after ablation	follow up	0.84 (ER)	0.77 (LR)	81 (ER)	69 (ER)	93 (ER)
Liu 2020 [120]	Retrospective, single center	CEUS, clinical variables	419 patients	DL	Cross Stratification Using R-RFA and R-SR in Swapped Groups	PFS prediction of early HCC after SR and RFA	follow up	0.726 (RFA); 0.741 (SR)
Zhong 2023 [121]	Prospective, single center	SWE, clinical variables	345 patients	HR	multi-scale radiomics model	PHLF prediction	ISGLS	0.822		75	70.4	77.3
Xue 2024 [122]	Retrospective, multi-center	B-mode, CEUS, clinical features	532 patients	DL, HR	Resnet 50 trained with varying granularity with progressive training strategy	PHLF prediction	ISGLS	0.860-1		88-90.5	75-100	87.5-100
Jiang 2025 [123]	Retrospective, single center	SWE, clinical variables	633 patients	HR	Different classifier	Postoperative complications and ER prediction after resection	Comprehensive Complication Index (CCI) and follow up	0.832 (complications); 0.844 (ER)		78; 66.3	78.6; 60.4	74.6: 98.5
Liu 2019 [124]	Retrospective, single center	B-mode, CEUS	130 patients	DL, HR	Comparison of R-DLCEUS, R-TIC, and R-Bmode models	Prediction of responses to TACE	mRECIST	0.93 (R-DLCEUS)		90	89.3	92.3
Jin 2021 [125]	Retrospective, single center	B-mode, SWE, clinical features	434 patients	DL	Combination of 2D-SWE and B-mode DL features, sex and age	HCC occurrence in chronic hepatitis B patients	follow up	0.9			83.3	96.3

参考文献 127

[1]	Asrani SK , Devarbhavi H , Eaton J , Kamath PS . Burden of liver diseases in the world. J Hepatol 2019; 70: 151-171. doi: 10.1016/j.jhep.2018.09.014
[2]	EASL Clinical Practice Guidelines on the management of hepatocellular carcinoma. J Hepatol 2025; 82:315-374.
[3]	Lambin P , Leijenaar RTH , Deist TM , Peerlings J , de Jong EEC , van Timmeren J , et al . Radiomics: the bridge between medical imaging and personalized medicine. Nat Rev Clin Oncol 2017; 14: 749-762. doi: 10.1038/nrclinonc.2017.141
[4]	Zhang D , Zhang XY , Duan YY , Dietrich CF , Cui XW , Zhang CX . An overview of ultrasound-derived radiomics and deep learning in liver. Med Ultrason 2023; 25: 445-452. doi: 10.11152/mu-4080
[5]	Quaia E , Calliada F , Bertolotto M , Rossi S , Garioni L , Rosa L , et al . Characterization of focal liver lesions with contrast-specific US modes and a sulfur hexafluoride-filled microbubble contrast agent: diagnostic performance and confidence. Radiology 2004; 232: 420-430. doi: 10.1148/radiol.2322031401
[6]	Fang C , Sidhu PS . Ultrasound-based liver elastography: current results and future perspectives. Abdominal radiology (New York) 2020; 45: 3463-3472. doi: 10.1007/s00261-020-02717-x
[7]	Zhang Y , Cui J , Wan W , Liu J . Multimodal imaging under artificial intelligence algorithm for the diagnosis of liver cancer and its relationship with expressions of EZH2 and p57. Comput Intell Neurosci 2022; 2022: 4081654
[8]	Xue LY , Jiang ZY , Fu TT , Wang QM , Zhu YL , Dai M , et al . Transfer learning radiomics based on multimodal ultrasound imaging for staging liver fibrosis. Eur Radiol 2020; 30: 2973-2983. doi: 10.1007/s00330-019-06595-w
[9]	Wei X , Wang Y , Wang L , Gao M , He Q , Zhang Y , et al. Simultaneous grading diagnosis of liver fibrosis, inflammation, and steatosis using multimodal quantitative ultrasound and artificial intelligence framework. Med Biol Eng Comput 2024.
[10]	Lambin P , Rios-Velazquez E , Leijenaar R , Carvalho S , Van Stiphout RG , Granton P , et al . Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer 2012; 48: 441-446. doi: 10.1016/j.ejca.2011.11.036
[11]	Nohara Y , Matsumoto K , Soejima H , Nakashima NJCM , Biomedicine Pi . Explanation of machine learning models using shapley additive explanation and application for real data in hospital Comput Methods Programs Biomed 2022; 214:106584.
[12]	Traverso A , Wee L , Dekker A , Gillies RJIJoROBP . Repeatability and reproducibility of radiomic features: a systematic review Int J Radiat Oncol Biol Phys 2018; 102:1143-1158.
[13]	Lecun Y , Bottou L , Bengio Y , Haffner P . Gradient-based learning applied to document recognition. Proceedings of the IEEE 1998; 86: 2278-2324. doi: 10.1109/5.726791
[14]	Hochreiter S , Schmidhuber J . Long Short-Term Memory. Neural Computation 1997; 9: 1735-1780. doi: 10.1162/neco.1997.9.8.1735
[15]	Wu Z , Pan S , Chen F , Long G , Zhang C , Yu PS . A comprehensive survey on graph neural networks. IEEE Trans Neural Netw Learn Syst 2021; 32: 4-24. doi: 10.1109/TNNLS.2020.2978386
[16]	Vaswani A , Shazeer N , Parmar N , Uszkoreit J , Jones L , Gomez AN , et al. Attention is all you need. 2017;30.
[17]	Dosovitskiy A , Beyer L , Kolesnikov A , Weissenborn D , Zhai X , Unterthiner T , et al. An image is worth 16x16 words: Transformers for image recognition at scale. 2020.
[18]	Salahuddin Z , Woodruff HC , Chatterjee A , Lambin PJCib , medicine. Transparency of deep neural networks for medical image analysis: A review of interpretability methods Comput Biol Med 2022; 140:105111.
[19]	Li MD , Cheng MQ , Chen LD , Hu HT , Zhang JC , Ruan SM , et al . Reproducibility of radiomics features from ultrasound images: influence of image acquisition and processing. Eur Radiol 2022; 32: 5843-5851. doi: 10.1007/s00330-022-08662-1
[20]	Salem N , Malik H , Shams A . Medical image enhancement based on histogram algorithms. Procedia Computer Science 2019; 163: 300-311. doi: 10.1016/j.procs.2019.12.112
[21]	Rhyou SY , Yoo JC . Cascaded deep learning neural network for automated liver steatosis diagnosis using ultrasound images. Sensors (Basel, Switzerland) 2021; 21: 5304. doi: 10.3390/s21165304
[22]	Mali SA , Ibrahim A , Woodruff HC , Andrearczyk V , Müller H , Primakov S , et al . Making radiomics more reproducible across scanner and imaging protocol variations: a review of harmonization methods. J Pers Med 2021; 11: 842. doi: 10.3390/jpm11090842
[23]	Zhao Z , Qin Y , Shao K , Liu Y , Zhang Y , Li H , et al. Radiomics harmonization in ultrasound images for cervical cancer lymph node metastasis prediction using Cycle-GAN. Technol Cancer Res Treat 2024; 23:15330338241302237.
[24]	Johnson WE , Li C , Rabinovic A . Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics (Oxford, England) 2007; 8: 118-127. doi: 10.1093/biostatistics/kxj037
[25]	Duarte-Salazar CA , Castro-Ospina AE , Becerra MA , Delgado-Trejos E . Speckle noise reduction in ultrasound images for improving the metrological evaluation of biomedical applications: an overview. IEEE Access 2020; 8: 15983-15999. doi: 10.1109/ACCESS.2020.2967178
[26]	Dietrich CF , Nolsøe CP , Barr RG , Berzigotti A , Burns PN , Cantisani V , et al . Guidelines and good clinical practice recommendations for contrast-enhanced ultrasound (CEUS) in the liver-update 2020 WFUMB in cooperation with EFSUMB, AFSUMB, AIUM, and FLAUS. Ultrasound Med Biol 2020; 46: 2579-2604. doi: 10.1016/j.ultrasmedbio.2020.04.030
[27]	Tian H , Cai W , Ding W , Liang P , Yu J , Huang Q . Long-term liver lesion tracking in contrast-enhanced ultrasound videos via a siamese network with temporal motion attention. Front Physiol 2023; 14: 1180713. doi: 10.3389/fphys.2023.1180713
[28]	Li MD , Hu HT , Ruan SM , Cheng MQ , Chen LD , Huang ZR , et al . ADMNet: adaptive-weighting dual mapping for online tracking with respiratory motion estimation in contrast-enhanced ultrasound. IEEE Trans Image Process 2024; 33: 58-68. doi: 10.1109/TIP.2023.3333195
[29]	Duan Y , Shi S , Long H , Zhong X , Tan Y , Liu G , et al . A bi-modal temporal segmentation network for automated segmentation of focal liver lesions in dynamic contrast-enhanced ultrasound. Ultrasound Med Biol 2025; 51: 759-767. doi: 10.1016/j.ultrasmedbio.2024.12.014
[30]	Dietrich CF , Bamber J , Berzigotti A , Bota S , Cantisani V , Castera L , et al . EFSUMB guidelines and recommendations on the clinical use of liver ultrasound elastography, update 2017 (long version). Ultraschall Med 2017; 38: e16-e47
[31]	Ferraioli G , Filice C , Castera L , Choi BI , Sporea I , Wilson SR , et al . WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 3: liver. Ultrasound Med Biol 2015; 41: 1161-1179. doi: 10.1016/j.ultrasmedbio.2015.03.007
[32]	Wang H , Zheng P , Wang X , Sang L . Effect of Q-Box size on liver stiffness measurement by two-dimensional shear wave elastography. J Clin Ultrasound 2021; 49: 978-983. doi: 10.1002/jcu.23075
[33]	Wang K , Lu X , Zhou H , Gao Y , Zheng J , Tong M , et al . Deep learning radiomics of shear wave elastography significantly improved diagnostic performance for assessing liver fibrosis in chronic hepatitis B: a prospective multicentre study. Gut 2019; 68: 729-741. doi: 10.1136/gutjnl-2018-316204
[34]	Liu Z , Li W , Zhu Z , Wen H , Li MD , Hou C , et al . A deep learning model with data integration of ultrasound contrast-enhanced micro-flow cines, B-mode images, and clinical parameters for diagnosing significant liver fibrosis in patients with chronic hepatitis B. Eur Radiol 2023; 33: 5871-5881. doi: 10.1007/s00330-023-09436-z
[35]	Li W , Huang Y , Zhuang BW , Liu GJ , Hu HT , Li X , et al . Multiparametric ultrasomics of significant liver fibrosis: A machine learning-based analysis. Eur Radiol 2019; 29: 1496-1506. doi: 10.1007/s00330-018-5680-z
[36]	Stahlschmidt SR , Ulfenborg B , Synnergren J . Multimodal deep learning for biomedical data fusion: a review. Brief Bioinform 2022; 23: bbab569. doi: 10.1093/bib/bbab569
[37]	Younossi ZM , Golabi P , Paik JM , Henry A , Van Dongen C , Henry L . The global epidemiology of nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH): a systematic review. Hepatology 2023; 77: 1335-1347. doi: 10.1097/HEP.0000000000000004
[38]	Wu CH , Hung CL , Lee TY , Wu CY , Chu WCC , editors. Fatty liver diagnosis using deep learning in ultrasound image. 2022 IEEE International Conference on Digital Health (ICDH) 2022 10-16 July 2022.
[39]	Cao W , An X , Cong L , Lyu C , Zhou Q , Guo R . Application of deep learning in quantitative analysis of 2-dimensional ultrasound imaging of nonalcoholic fatty liver disease. J Ultrasound Med 2020; 39: 51-59. doi: 10.1002/jum.15070
[40]	Byra M , Styczynski G , Szmigielski C , Kalinowski P , Michałowski Ł , Paluszkiewicz R , et al . Transfer learning with deep convolutional neural network for liver steatosis assessment in ultrasound images. Int J Comput Assist Radiol Surg 2018; 13: 1895-1903. doi: 10.1007/s11548-018-1843-2
[41]	Reddy DS , Bharath R , Rajalakshmi P , editors. A novel computer-aided diagnosis framework using deep learning for classification of fatty liver disease in ultrasound imaging. 2018 IEEE 20th International Conference on e-Health Networking, Applications and Services (Healthcom) 2018 17-20 Sept. 2018.
[42]	Kim T , Lee DH , Park EK , Choi S . Deep learning techniques for fatty liver using multi-view ultrasound images scanned by different scanners: development and validation study. JMIR Med Inform 2021; 9: e30066. doi: 10.2196/30066
[43]	Han A , Byra M , Heba E , Andre MP , Erdman JW , Jr R , et al . Noninvasive diagnosis of nonalcoholic fatty liver disease and quantification of liver fat with radiofrequency ultrasound data using one-dimensional convolutional neural networks. Radiology 2020; 295: 342-350
[44]	Chou TH , Yeh HJ , Chang CC , Tang JH , Kao WY , Su IC , et al . Deep learning for abdominal ultrasound: a computer-aided diagnostic system for the severity of fatty liver. J Chin Med Assoc 2021; 84: 842-850. doi: 10.1097/JCMA.0000000000000585
[45]	Tahmasebi A , Wang S , Wessner CE , Vu T , Liu JB , Forsberg F , et al . Ultrasound-based machine learning approach for detection of nonalcoholic fatty liver disease. J Ultrasound Med 2023; 42: 1747-1756. doi: 10.1002/jum.16194
[46]	Yang Y , Liu J , Sun C , Shi Y , Hsing JC , Kamya A , et al . Nonalcoholic fatty liver disease (NAFLD) detection and deep learning in a Chinese community-based population. Eur Radiol 2023; 33: 5894-5906. doi: 10.1007/s00330-023-09515-1
[47]	Liu Y , Yu W , Wang P , Huang Y , Li J , Li P . Deep learning with ultrasound images enhance the diagnosis of nonalcoholic fatty liver. Ultrasound Med Biol 2024; 50: 1724-1730. doi: 10.1016/j.ultrasmedbio.2024.07.014
[48]	Meng D , Zhang L , Cao G , Cao W , Zhang G , Hu B . Liver fibrosis classification based on transfer learning and fcnet for ultrasound images. IEEE Access 2017; 5: 5804-5810
[49]	Lee JH , Joo I , Kang TW , Paik YH , Sinn DH , Ha SY , et al . Deep learning with ultrasonography: automated classification of liver fibrosis using a deep convolutional neural network. Eur Radiol 2020; 30: 1264-1273. doi: 10.1007/s00330-019-06407-1
[50]	Feng X , Chen X , Dong C , Liu Y , Liu Z , Ding R , et al . Multi-scale information with attention integration for classification of liver fibrosis in B-mode US image. Comput Methods Programs Biomed 2022; 215: 106598. doi: 10.1016/j.cmpb.2021.106598
[51]	Ruan D , Shi Y , Jin L , Yang Q , Yu W , Ren H , et al . An ultrasound image-based deep multi-scale texture network for liver fibrosis grading in patients with chronic HBV infection. Liver Int 2021; 41: 2440-2454. doi: 10.1111/liv.14999
[52]	Duan YY , Qin J , Qiu WQ , Li SY , Li C , Liu AS , et al . Performance of a generative adversarial network using ultrasound images to stage liver fibrosis and predict cirrhosis based on a deep-learning radiomics nomogram. Clin Radiol 2022; 77: e723-e731. doi: 10.1016/j.crad.2022.06.003
[53]	Joo Y , Park HC , Lee OJ , Yoon C , Choi MH , Choi C . Classification of liver fibrosis from heterogeneous ultrasound image. IEEE Access 2023; 11: 9920-9930. doi: 10.1109/ACCESS.2023.3240216
[54]	Park HC , Joo Y , Lee OJ , Lee K , Song TK , Choi C , et al . Automated classification of liver fibrosis stages using ultrasound imaging. BMC Med Imaging 2024; 24: 36. doi: 10.1186/s12880-024-01209-4
[55]	Ai H , Huang Y , Tai DI , Tsui PH , Zhou Z . Ultrasonic assessment of liver fibrosis using one-dimensional convolutional neural networks based on frequency spectra of radiofrequency signals with deep learning segmentation of liver regions in B-mode images: a feasibility study. Sensors (Basel, Switzerland) 2024; 24: 5513. doi: 10.3390/s24175513
[56]	Zhang L , Tan Z , Li C , Mou L , Shi YL , Zhu XX , et al . A deep learning model based on high-frequency ultrasound images for classification of different stages of liver fibrosis. Liver Int 2025; 45: e70148. doi: 10.1111/liv.70148
[57]	Chen Y , Luo Y , Huang W , Hu D , Zheng RQ , Cong SZ , et al . Machine-learning-based classification of real-time tissue elastography for hepatic fibrosis in patients with chronic hepatitis B. Comput Biol Med 2017; 89: 18-23. doi: 10.1016/j.compbiomed.2017.07.012
[58]	Gatos I , Tsantis S , Spiliopoulos S , Karnabatidis D , Theotokas I , Zoumpoulis P , et al . A new computer aided diagnosis system for evaluation of chronic liver disease with ultrasound shear wave elastography imaging. Medical physics 2016; 43: 1428-1436. doi: 10.1118/1.4942383
[59]	Gatos I , Tsantis S , Spiliopoulos S , Karnabatidis D , Theotokas I , Zoumpoulis P , et al . A machine-learning algorithm toward color analysis for chronic liver disease classification, employing ultrasound shear wave elastography. Ultrasound Med Biol 2017; 43: 1797-1810. doi: 10.1016/j.ultrasmedbio.2017.05.002
[60]	Gatos I , Tsantis S , Spiliopoulos S , Karnabatidis D , Theotokas I , Zoumpoulis P , et al . Temporal stability assessment in shear wave elasticity images validated by deep learning neural network for chronic liver disease fibrosis stage assessment. Medical physics 2019; 46: 2298-2309. doi: 10.1002/mp.13521
[61]	Kagadis GC , Drazinos P , Gatos I , Tsantis S , Papadimitroulas P , Spiliopoulos S , et al . Deep learning networks on chronic liver disease assessment with fine-tuning of shear wave elastography image sequences. Phys Med Biol 2020; 65: 215027. doi: 10.1088/1361-6560/abae06
[62]	Meng F , Wu Q , Zhang W , Hou S . Application of interpretable machine learning models based on ultrasonic radiomics for predicting the risk of fibrosis progression in diabetic patients with nonalcoholic fatty liver disease. Diabetes Metab Syndr Obes 2023; 16: 3901-3913. doi: 10.2147/DMSO.S439127
[63]	Liu Z , Wen H , Zhu Z , Li Q , Liu L , Li T , et al . Diagnosis of significant liver fibrosis in patients with chronic hepatitis B using a deep learning-based data integration network. Hepatol Int 2022; 16: 526-536. doi: 10.1007/s12072-021-10294-4
[64]	Gao L , Zhou R , Dong C , Feng C , Li Z , Wan X , et al., editors. Multi-modal active learning for automatic liver fibrosis diagnosis based on ultrasound shear wave elastography. 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI) 2021 13-16 April 2021.
[65]	Lu X , Zhou H , Wang K , Jin J , Meng F , Mu X , et al . Comparing radiomics models with different inputs for accurate diagnosis of significant fibrosis in chronic liver disease. Eur Radiol 2021; 31: 8743-8754. doi: 10.1007/s00330-021-07934-6
[66]	Chen LD , Huang ZR , Yang H , Cheng MQ , Hu HT , Lu XZ , et al . US-based sequential algorithm integrating an ai model for advanced liver fibrosis screening. Radiology 2024; 311: e231461. doi: 10.1148/radiol.231461
[67]	Hwang YN , Lee JH , Kim GY , Jiang YY , Kim SM . Classification of focal liver lesions on ultrasound images by extracting hybrid textural features and using an artificial neural network. Biomed Mater Eng 2015;26 Suppl 1:S1599-611.
[68]	Mao B , Ma J , Duan S , Xia Y , Tao Y , Zhang L . Preoperative classification of primary and metastatic liver cancer via machine learning-based ultrasound radiomics. Eur Radiol 2021; 31: 4576-4586. doi: 10.1007/s00330-020-07562-6
[69]	Peng Y , Lin P , Wu L , Wan D , Zhao Y , Liang L , et al . Ultrasound-based radiomics analysis for preoperatively predicting different histopathological subtypes of primary liver cancer. Frontiers in oncology 2020; 10: 1646. doi: 10.3389/fonc.2020.01646
[70]	Qin H , Wu YQ , Lin P , Gao RZ , Li X , Wang XR , et al . Ultrasound image-based radiomics: an innovative method to identify primary tumorous sources of liver metastases. J Ultrasound Med 2021; 40: 1229-1244. doi: 10.1002/jum.15506
[71]	Peng JB , Peng YT , Lin P , Wan D , Qin H , Li X , et al . Differentiating infected focal liver lesions from malignant mimickers: value of ultrasound-based radiomics. Clin Radiol 2022; 77: 104-113. doi: 10.1016/j.crad.2021.10.009
[72]	Xi IL , Wu J , Guan J , Zhang PJ , Horii SC , Soulen MC , et al . Deep learning for differentiation of benign and malignant solid liver lesions on ultrasonography. Abdom Radiol (NY) 2021; 46: 534-543. doi: 10.1007/s00261-020-02564-w
[73]	Schmauch B , Herent P , Jehanno P , Dehaene O , Saillard C , Aubé C , et al . Diagnosis of focal liver lesions from ultrasound using deep learning. Diagn Interv Imaging 2019; 100: 227-233. doi: 10.1016/j.diii.2019.02.009
[74]	Chen J , Zhang W , Bao J , Wang K , Zhao Q , Zhu Y , et al . Implications of ultrasound-based deep learning model for preoperatively differentiating combined hepatocellular-cholangiocarcinoma from hepatocellular carcinoma and intrahepatic cholangiocarcinoma. Abdom Radiol (NY) 2024; 49: 93-102
[75]	Tiyarattanachai T , Apiparakoon T , Marukatat S , Sukcharoen S , Geratikornsupuk N , Anukulkarnkusol N , et al . Development and validation of artificial intelligence to detect and diagnose liver lesions from ultrasound images. PLoS One 2021; 16: e0252882. doi: 10.1371/journal.pone.0252882
[76]	Yang Q , Wei J , Hao X , Kong D , Yu X , Jiang T , et al . Improving B-mode ultrasound diagnostic performance for focal liver lesions using deep learning: A multicentre study. EBioMedicine 2020; 56: 102777. doi: 10.1016/j.ebiom.2020.102777
[77]	Yang Y , Cairang Y , Jiang T , Zhou J , Zhang L , Qi B , et al . Ultrasound identification of hepatic echinococcosis using a deep convolutional neural network model in China: a retrospective, large-scale, multicentre, diagnostic accuracy study. Lancet Digit Health 2023; 5: e503-e514. doi: 10.1016/S2589-7500(23)00091-2
[78]	Du Z , Fan F , Ma J , Liu J , Yan X , Chen X , et al . Development and validation of an ultrasound-based interpretable machine learning model for the classification of ≤ 3 cm hepatocellular carcinoma: a multicentre retrospective diagnostic study. EClinicalMedicine 2025; 81: 103098. doi: 10.1016/j.eclinm.2025.103098
[79]	Streba CT , Ionescu M , Gheonea DI , Sandulescu L , Ciurea T , Saftoiu A , et al . Contrast-enhanced ultrasonography parameters in neural network diagnosis of liver tumors. World J Gastroenterol 2012; 18: 4427-4434. doi: 10.3748/wjg.v18.i32.4427
[80]	Gatos I , Tsantis S , Spiliopoulos S , Skouroliakou A , Theotokas I , Zoumpoulis P , et al . A new automated quantification algorithm for the detection and evaluation of focal liver lesions with contrast-enhanced ultrasound. Medical physics 2015; 42: 3948-3959. doi: 10.1118/1.4921753
[81]	Kondo S , Takagi K , Nishida M , Iwai T , Kudo Y , Ogawa K , et al . Computer-aided diagnosis of focal liver lesions using contrast-enhanced ultrasonography with perflubutane microbubbles. IEEE Trans Med Imaging 2017; 36: 1427-1437. doi: 10.1109/TMI.2017.2659734
[82]	Turco S , Tiyarattanachai T , Ebrahimkheil K , Eisenbrey J , Kamaya A , Mischi M , et al . Interpretable machine learning for characterization of focal liver lesions by contrast-enhanced ultrasound. IEEE Trans Ultrason Ferroelectr Freq Control 2022; 69: 1670-1681. doi: 10.1109/TUFFC.2022.3161719
[83]	Guo LH , Wang D , Qian YY , Zheng X , Zhao CK , Li XL , et al . A two-stage multi-view learning framework based computer-aided diagnosis of liver tumors with contrast enhanced ultrasound images. Clin Hemorheol Microcirc 2018; 69: 343-354. doi: 10.3233/CH-170275
[84]	Li L , Liang X , Yu Y , Mao R , Han J , Peng C , et al . Radiomics-based machine learning classification strategy for characterization of hepatocellular carcinoma on contrast-enhanced ultrasound in high-risk patients with li-rads category m nodules. Indian J Radiol Imaging 2024; 34: 405-415. doi: 10.1055/s-0043-1777993
[85]	Wang Z , Yao J , Jing X , Li K , Lu S , Yang H , et al . A combined model based on radiomics features of Sonazoid contrast-enhanced ultrasound in the Kupffer phase for the diagnosis of well-differentiated hepatocellular carcinoma and atypical focal liver lesions: a prospective, multicenter study. Abdom Radiol (NY) 2024; 49: 3427-3437. doi: 10.1007/s00261-024-04253-4
[86]	Pan F , Huang Q , Li X , editors. Classification of liver tumors with CEUS based on 3D-CNN. 2019 IEEE 4th International Conference on Advanced Robotics and Mechatronics (ICARM) 2019 3-5 July 2019.
[87]	Căleanu CD , Sîrbu CL , Simion G . Deep neural architectures for contrast enhanced ultrasound (ceus) focal liver lesions automated diagnosis. Sensors (Basel, Switzerland) 2021; 21: 4126. doi: 10.3390/s21124126
[88]	Ding W , Meng Y , Ma J , Pang C , Wu J , Tian J , et al . Contrast-enhanced ultrasound-based AI model for multi-classification of focal liver lesions. Journal of Hepatology 2025; 83: 426-439. doi: 10.1016/j.jhep.2025.01.011
[89]	Yao Z , Dong Y , Wu G , Zhang Q , Yang D , Yu JH , et al . Preoperative diagnosis and prediction of hepatocellular carcinoma: Radiomics analysis based on multi-modal ultrasound images. BMC Cancer 2018; 18: 1089. doi: 10.1186/s12885-018-5003-4
[90]	Hu HT , Li MD , Zhang JC , Ruan SM , Wu SS , Lin XX , et al . Ultrasomics differentiation of malignant and benign focal liver lesions based on contrast-enhanced ultrasound. BMC medical imaging 2024; 24: 242. doi: 10.1186/s12880-024-01426-x
[91]	Su LY , Xu M , Chen YL , Lin MX , Xie XY . Ultrasomics in liver cancer: developing a radiomics model for differentiating intrahepatic cholangiocarcinoma from hepatocellular carcinoma using contrast-enhanced ultrasound. World J Radiol 2024; 16: 247-255. doi: 10.4329/wjr.v16.i7.247
[92]	Hu HT , Wang W , Chen LD , Ruan SM , Chen SL , Li X , et al . Artificial intelligence assists identifying malignant versus benign liver lesions using contrast-enhanced ultrasound. J Gastroenterol Hepatol 2021; 36: 2875-2883. doi: 10.1111/jgh.15522
[93]	Liu L , Tang C , Li L , Chen P , Tan Y , Hu X , et al . Deep learning radiomics for focal liver lesions diagnosis on long-range contrast-enhanced ultrasound and clinical factors. Quant Imaging Med Surg 2022; 12: 3213-3226. doi: 10.21037/qims-21-1004
[94]	Dong Y , Wang QM , Li Q , Li LY , Zhang Q , Yao Z , et al . Preoperative prediction of microvascular invasion of hepatocellular carcinoma: radiomics algorithm based on ultrasound original radio frequency signals. Frontiers in oncology 2019; 9: 1203. doi: 10.3389/fonc.2019.01203
[95]	Dong Y , Zhou L , Xia W , Zhao XY , Zhang Q , Jian JM , et al . Preoperative prediction of microvascular invasion in hepatocellular carcinoma: initial application of a radiomic algorithm based on grayscale ultrasound images. Frontiers in oncology 2020; 10: 353. doi: 10.3389/fonc.2020.00353
[96]	Hu HT , Wang Z , Huang XW , Chen SL , Zheng X , Ruan SM , et al . Ultrasound-based radiomics score: a potential biomarker for the prediction of microvascular invasion in hepatocellular carcinoma. Eur Radiol 2019; 29: 2890-2901. doi: 10.1007/s00330-018-5797-0
[97]	Dong Y , Zuo D , Qiu YJ , Cao JY , Wang HZ , Yu LY , et al . Preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma based on kupffer phase radiomics features of sonazoid contrast-enhanced ultrasound (SCEUS): A prospective study. Clinical hemorheology and microcirculation 2022; 81: 97-107. doi: 10.3233/CH-211363
[98]	Zhang D , Wei Q , Wu GG , Zhang XY , Lu WW , Lv WZ , et al . Preoperative prediction of microvascular invasion in patients with hepatocellular carcinoma based on radiomics nomogram using contrast-enhanced ultrasound. Frontiers in oncology 2021; 11: 709339. doi: 10.3389/fonc.2021.709339
[99]	Zhang Y , Wei Q , Huang Y , Yao Z , Yan C , Zou X , et al . Deep learning of liver contrast-enhanced ultrasound to predict microvascular invasion and prognosis in hepatocellular carcinoma. Frontiers in oncology 2022; 12: 878061. doi: 10.3389/fonc.2022.878061
[100]	Qin X , Zhu J , Tu Z , Ma Q , Tang J , Zhang C . Contrast-enhanced ultrasound with deep learning with attention mechanisms for predicting microvascular invasion in single hepatocellular carcinoma. Academic radiology 2023;30 Suppl 1:S73-S80.
[101]	Qin Q , Pang J , Li J , Gao R , Wen R , Wu Y , et al . Transformer model based on Sonazoid contrast-enhanced ultrasound for microvascular invasion prediction in hepatocellular carcinoma. Medical physics 2025; 52: e17895. doi: 10.1002/mp.17895
[102]	Wang Y , Xie W , Li C , Xu Q , Du Z , Zhong Z , et al . Automated microvascular invasion prediction of hepatocellular carcinoma via deep relation reasoning from dynamic contrast-enhanced ultrasound. Computerized medical imaging and graphics: the official journal of the Computerized Medical Imaging Society 2025; 124: 102606. doi: 10.1016/j.compmedimag.2025.102606
[103]	Zheng R , Zhang X , Liu B , Zhang Y , Shen H , Xie X , et al . Comparison of non-radiomics imaging features and radiomics models based on contrast-enhanced ultrasound and Gd-EOB-DTPA-enhanced MRI for predicting microvascular invasion in hepatocellular carcinoma within 5 cm. Eur Radiol 2023; 33: 6462-6472. doi: 10.1007/s00330-023-09789-5
[104]	Zhang W , Guo Q , Zhu Y , Wang M , Zhang T , Cheng G , et al . Cross-institutional evaluation of deep learning and radiomics models in predicting microvascular invasion in hepatocellular carcinoma: validity, robustness, and ultrasound modality efficacy comparison. Cancer imaging : the official publication of the International Cancer Imaging Society 2024; 24: 142
[105]	Li L , Wang S , Chen J , Wu C , Chen Z , Ye F , et al . Radiomics diagnosis of microvascular invasion in hepatocellular carcinoma using 3D ultrasound and whole-slide image fusion. Small methods 2025; 9: e2401617. doi: 10.1002/smtd.202401617
[106]	Ren S , Qi Q , Liu S , Duan S , Mao B , Chang Z , et al . Preoperative prediction of pathological grading of hepatocellular carcinoma using machine learning-based ultrasomics: A multicenter study. Eur J Radiol 2021; 143: 109891. doi: 10.1016/j.ejrad.2021.109891
[107]	Qin X , Hu X , Xiao W , Zhu C , Ma Q , Zhang C . Preoperative evaluation of hepatocellular carcinoma differentiation using contrast-enhanced ultrasound-based deep-learning radiomics model. Journal of hepatocellular carcinoma 2023; 10: 157-168. doi: 10.2147/JHC.S400166
[108]	Li C , Xu J , Liu Y , Wu M , Dai W , Song J , et al . Kupffer phase radiomics signature in sonazoid-enhanced ultrasound is an independent and effective predictor of the pathologic grade of hepatocellular carcinoma. Journal of oncology 2022; 2022: 6123242
[109]	Qian H , Shen Z , Zhou D , Huang Y . Intratumoral and peritumoral radiomics model based on abdominal ultrasound for predicting Ki-67 expression in patients with hepatocellular cancer. Frontiers in oncology 2023; 13: 1209111. doi: 10.3389/fonc.2023.1209111
[110]	Zhang D , Zhang XY , Lu WW , Liao JT , Zhang CX , Tang Q , et al . Predicting Ki-67 expression in hepatocellular carcinoma: nomogram based on clinical factors and contrast-enhanced ultrasound radiomics signatures. Abdominal radiology (New York) 2024; 49: 1419-1431. doi: 10.1007/s00261-024-04191-1
[111]	Qian H , Huang Y , Xu L , Fu H , Lu B . Role of peritumoral tissue analysis in predicting characteristics of hepatocellular carcinoma using ultrasound-based radiomics. Sci Rep 2024; 14: 11538. doi: 10.1038/s41598-024-62457-6
[112]	Bu D , Duan S , Ren S , Ma Y , Liu Y , Li Y , et al . Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinoma. Front Med (Lausanne) 2025; 12: 1565618
[113]	Liang L , Pang JS , Gao RZ , Que Q , Wu YQ , Peng JB , et al . Development and validation of a combined radiomic and clinical model based on contrast-enhanced ultrasound for preoperative prediction of CK19-positive hepatocellular carcinoma. Abdominal radiology (New York) 2025; 50: 3516-3529. doi: 10.1007/s00261-025-04799-x
[114]	Huang H , Ruan SM , Xian MF , Li MD , Cheng MQ , Li W , et al . Contrast-enhanced ultrasound-based ultrasomics score: a potential biomarker for predicting early recurrence of hepatocellular carcinoma after resection or ablation. The British journal of radiology 2022; 95: 20210748. doi: 10.1259/bjr.20210748
[115]	Cao K , Wang X , Xu C , Wu L , Li L , Yuan Y , et al . Ultrasound-based radiomics analysis for assessing risk factors associated with early recurrence following surgical resection of hepatocellular carcinoma. Ultrasound Med Biol 2024; 50: 1964-1972. doi: 10.1016/j.ultrasmedbio.2024.09.002
[116]	Wu JP , Ding WZ , Wang YL , Liu S , Zhang XQ , Yang Q , et al . Radiomics analysis of ultrasound to predict recurrence of hepatocellular carcinoma after microwave ablation. International journal of hyperthermia: the official journal of European Society for Hyperthermic Oncology, North American Hyperthermia Group 2022; 39: 595-604. doi: 10.1080/02656736.2022.2062463
[117]	Zhang H , Huo F . Prediction of early recurrence of HCC after hepatectomy by contrast-enhanced ultrasound-based deep learning radiomics. Frontiers in oncology 2022; 12: 930458. doi: 10.3389/fonc.2022.930458
[118]	Huang Z , Shu Z , Zhu RH , Xin JY , Wu LL , Wang HZ , et al . Deep learning-based radiomics based on contrast-enhanced ultrasound predicts early recurrence and survival outcome in hepatocellular carcinoma. World journal of gastrointestinal oncology 2022; 14: 2380-2392. doi: 10.4251/wjgo.v14.i12.2380
[119]	Ma QP , He XL , Li K , Wang JF , Zeng QJ , Xu EJ , et al . Dynamic contrast-enhanced ultrasound radiomics for hepatocellular carcinoma recurrence prediction after thermal ablation. Molecular imaging and biology 2021; 23: 572-585. doi: 10.1007/s11307-021-01578-0
[120]	Liu F , Liu D , Wang K , Xie X , Su L , Kuang M , et al . Deep learning radiomics based on contrast-enhanced ultrasound might optimize curative treatments for very-early or early-stage hepatocellular carcinoma patients. Liver Cancer 2020; 9: 397-413. doi: 10.1159/000505694
[121]	Zhong X , Salahuddin Z , Chen Y , Woodruff HC , Long H , Peng J , et al . An interpretable radiomics model based on two-dimensional shear wave elastography for predicting symptomatic post-hepatectomy liver failure in patients with hepatocellular carcinoma. Cancers 2023; 15: 5303. doi: 10.3390/cancers15215303
[122]	Xue L , Zhu J , Fang Y , Xie X , Cheng G , Zhang Y , et al . Preoperative ultrasound radomics to predict posthepatectomy liver failure in patients with hepatocellular carcinoma. J Ultrasound Med 2024; 43: 2269-2280. doi: 10.1002/jum.16559
[123]	Jiang D , Ren J , Qian Y , Gu Y , Wang R , Yu H , et al . Prediction models after hepatectomy for hepatocellular carcinoma-based ultrasonic radiomics: an observational study. European journal of medical research 2025; 30: 722. doi: 10.1186/s40001-025-02977-7
[124]	Liu D , Liu F , Xie X , Su L , Liu M , Xie X , et al . Accurate prediction of responses to transarterial chemoembolization for patients with hepatocellular carcinoma by using artificial intelligence in contrast-enhanced ultrasound. Eur Radiol 2020; 30: 2365-2376. doi: 10.1007/s00330-019-06553-6
[125]	Jin J , Yao Z , Zhang T , Zeng J , Wu L , Wu M , et al . Deep learning radiomics model accurately predicts hepatocellular carcinoma occurrence in chronic hepatitis B patients: a five-year follow-up. American journal of cancer research 2021; 11: 576-589
[126]	Došilović FK , Brčić M , Hlupić N , editors. Explainable artificial intelligence: a survey. 2018 41st International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO) 2018 21-25 May 2018.
[127]	Lambin P , Woodruff HC , Mali SA , Zhong X , Kuang S , Lavrova E , et al . Radiomics quality score 2. 0: towards radiomics readiness levels and clinical translation for personalized medicine. Nature Reviews Clinical Oncology 2025; 22: 831-846

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

Multimodal Ultrasound Radiomics in Liver Disease: Current Status and Future Directions

RichHTML

PDF (PC)

PDF (Mobile)

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 127

相关文章 0

Metrics

本文评价

推荐阅读 0