Automated Machine Learning in the Sonographic Diagnosis of Non-alcoholic Fatty Liver Disease

Gummadi, Sriharsha; Patel, Nirmal; Naringrekar, Haresh; Needleman, Laurence; Lyshchik, Andrej; O’Kane, Patrick; Civan, Jesse; Eisenbrey, John R.

doi:10.37015/audt.2020.200008

Cited by 11 publications

(11 citation statements)

References 16 publications

(16 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…5 Furthermore, previous research in the field of automated NAFLD detection with US utilized high-quality images obtained from traditional, expensive, cart-based US systems, such as the Philips EPIQ7 (Philips Ultrasound, Inc., Bothell, Washington, United States). 12,16 In contrast, the images in this research have a reduced resolution and quality from POCUS with pre-set parameters, compared to traditional diagnostic US hardware. POCUS systems are inexpensive and accessible to PCPs.…”

Section: Resultsmentioning

confidence: 96%

“…Excluding upload and download time, processing in a cloud computer with 2 NVIDIA ® T4 previous studies that utilize B-mode images in ML algorithms, which report AUROCs between 0.71 and 1.0. 5,6,8-12,14-16 However, all previous works used high quality diagnostic US and highly curated datasets acquired by expert radiologists or sonographers 6,[8][9][10][11][12][14][15][16] or researchers formally trained in liver US. 5 Furthermore, previous research in the field of automated NAFLD detection with US utilized high-quality images obtained from traditional, expensive, cart-based US systems, such as the Philips EPIQ7 (Philips Ultrasound, Inc., Bothell, Washington, United States).…”

Section: Resultsmentioning

confidence: 99%

“…The images in this research are acquired by non-experts, who received minimal training, in contrast to expert radiologists in all previously reported studies. 6,8,16 Non-expert HCPs who participated in data acquisition were trained to identify the appearance of basic liver tissue. Correspondingly, our results suggest that it is beneficial for the models to focus on tissue microstructures, rather than full liver anatomy for steatosis detection in our dataset.…”

Section: Resultsmentioning

confidence: 99%

“… – 12 , 14 – 16 However, all previous works used high quality diagnostic US and highly curated datasets acquired by expert radiologists or sonographers 6 , 8 – 12 , 14 – 16 or researchers formally trained in liver US 5 . Furthermore, previous research in the field of automated NAFLD detection with US utilized high-quality images obtained from traditional, expensive, cart-based US systems, such as the Philips EPIQ7 (Philips Ultrasound, Inc., Bothell, Washington, United States) 12 , 16 . In contrast, the images in this research have a reduced resolution and quality from POCUS with pre-set parameters, compared to traditional diagnostic US hardware.…”

Section: Discussionmentioning

confidence: 99%

“…Existing research successfully automates steatosis detection in B-mode US images using machine learning (ML) methods 5 – 13 To our best knowledge, all previous works used high-quality diagnostic US machines, and highly curated datasets, acquired by expert radiologists or sonographers, 6 , 8 – 12 , 14 – 16 or researchers formally trained in liver US 5 . US machines included the GE VivisE9 US System (GE Healthcare INC, Horten, Norway), 14 , 15 Siemens Acuson S1000 (Siemens, Issaquah, Washington, United States), 10 Siemens Acuson S2000 (Siemens, Issaquah, Washington, United States) 5 , 8 etc.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Automated fatty liver disease detection in point-of-care ultrasound B-mode images

Ibrahim,

Blázquez-García,

Lightstone

et al. 2023

J. Med. Imag.

View full text Add to dashboard Cite

.PurposeNon-alcoholic fatty liver disease (NAFLD) is an increasing global health concern, with a prevalence of 25% worldwide. The rising incidence of NAFLD, an asymptomatic condition, reinforces the need for systematic screening strategies in primary care. We present the use of non-expert acquired point-of-care ultrasound (POCUS) B-mode images for the development of an automated steatosis classification algorithm.ApproachWe obtained a Health Insurance Portability and Accountability Act compliant dataset consisting of 478 patients [body mass index 23.60 ± 3.55, age 40.97 ± 10.61], imaged with POCUS by non-expert health care personnel. A U-Net deep learning (DL) model was used for liver segmentation in the POCUS B-mode images, followed by 224 × 224 patch extraction of liver parenchyma. Several DL models including VGG-16, ResNet-50, Inception V3, and DenseNet-121 were trained for binary classification of steatosis. All layers of each tested model were unfrozen, and the final layer was replaced with a custom classifier. Majority voting was applied for patient-level results.ResultsOn a hold-out test set of 81 patients, the final DenseNet-121 model yielded an area under the receiver operator characteristic curve of 90.1%, sensitivity of 95.0%, and specificity of 85.2% for the detection of liver steatosis. Average cross-validation performance in models using patches of liver parenchyma as input outperformed methods using complete B-mode frames.ConclusionsDespite minimal POCUS acquisition training, and low-quality B-mode images, it is possible to detect steatosis using DL algorithms. Implementation of this algorithm in POCUS software may offer an accessible, low-cost steatosis screening technology, for use by non-expert health care personnel.

show abstract

Section: Resultsmentioning

confidence: 96%

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Automated fatty liver disease detection in point-of-care ultrasound B-mode images

Ibrahim,

Blázquez-García,

Lightstone

et al. 2023

J. Med. Imag.

View full text Add to dashboard Cite

show abstract

Ultrasound‐Based Machine Learning Approach for Detection of Nonalcoholic Fatty Liver Disease

Tahmasebi

Wang

Wessner

et al. 2023

J of Ultrasound Medicine

View full text Add to dashboard Cite

Objectives-Current diagnosis of nonalcoholic fatty liver disease (NAFLD) relies on biopsy or MR-based fat quantification. This prospective study explored the use of ultrasound with artificial intelligence for the detection of NAFLD.Methods-One hundred and twenty subjects with clinical suspicion of NAFLD and 10 healthy volunteers consented to participate in this institutional review board-approved study. Subjects were categorized as NAFLD and non-NAFLD according to MR proton density fat fraction (PDFF) findings. Ultrasound images from 10 different locations in the right and left hepatic lobes were collected following a standard protocol. MRI-based liver fat quantification was used as the reference standard with >6.4% indicative of NAFLD. A supervised machine learning model was developed for assessment of NAFLD. To validate model performance, a balanced testing dataset of 24 subjects was used. Sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy with 95% confidence interval were calculated.Results-A total of 1119 images from 106 participants was used for model development. The internal evaluation achieved an average precision of 0.941, recall of 88.2%, and precision of 89.0%. In the testing set AutoML achieved a sensitivity of 72.2% (63.1%-80.1%), specificity of 94.6% (88.7%-98.0%), positive predictive value (PPV) of 93.1% (86.0%-96.7%), negative predictive value of 77.3% (71.6%-82.1%), and accuracy of 83.4% (77.9%-88.0%). The average agreement for an individual subject was 92%.Conclusions-An ultrasound-based machine learning model for identification of NAFLD showed high specificity and PPV in this prospective trial. This approach may in the future be used as an inexpensive and noninvasive screening tool for identifying NAFLD in high-risk patients.

show abstract

Learning from Subjective Ratings Using Auto-Decoded Deep Latent Embeddings

Li¹,

Ren

Yan³

et al. 2021

Medical Image Computing and Computer Assisted Intervention – MICCAI 2021

View full text Add to dashboard Cite

Automated Machine Learning in the Sonographic Diagnosis of Non-alcoholic Fatty Liver Disease

Cited by 11 publications

References 16 publications

Automated fatty liver disease detection in point-of-care ultrasound B-mode images

Automated fatty liver disease detection in point-of-care ultrasound B-mode images

Ultrasound‐Based Machine Learning Approach for Detection of Nonalcoholic Fatty Liver Disease

Learning from Subjective Ratings Using Auto-Decoded Deep Latent Embeddings

Contact Info

Product

Resources

About