Development and Validation of a Deep Neural Network for Accurate Identification of Endoscopic Images From Patients With Ulcerative Colitis and Crohn's Disease

Ruan, Guangcong; Qi, Jing; Cheng, Yi; Liu, Rongbei; Zhang, Bingqiang; Mao, Zhi; Chen, Jun-Rong; Xiao, Fang; Shen, Xinyong; Fan, Ling; Li, Qin; Li, Ning; Qiu, Zhujing; Xiao, Zhifeng; Lv, Linling; Chen, Minjia; Ying, Senhong; Chen, Lu; Tian, Yuting; Li, Guanhu; Zhang, Zhou; He, Mi; Qiao, Liang; Zhang, Zhu; Chen, Dongfeng; Cao, Qian; Nian, Yongjian; Wei, Yanling

doi:10.3389/fmed.2022.854677

Cited by 18 publications

(7 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the UC-CD cases, we observed that classification metrics are not able to perfectly distinguish between the two, which is in line with what previous machine learning studies suggest. Indeed, among the different datasets and ML techniques used in the literature to diagnose UC vs CD, such as using RNA sequencing data [32] or endoscopic images [33], none has been able to perform as a perfect classifier. Additionally, we note that in the case of HC and NH, all the accuracies are relatively high for all methods, suggesting the strong capability of ML methods to distinguish between HC and NH.…”

Section: Resultsmentioning

confidence: 99%

Machine learning based gut microbiota pattern and response to fiber as a diagnostic tool for chronic inflammatory diseases

Boodaghidizaji

Ardekani

2023

Preprint

View full text Add to dashboard Cite

Gut microbiota has been implicated in the pathogenesis of multiple gastrointestinal (GI) and systemic metabolic and inflammatory disorders where disrupted gut microbiota composition and function (dysbiosis) has been found in multiple studies. Thus, human microbiome data has a potential to be a great source of information for the diagnosis and disease characteristics (phenotypes, disease course, therapeutic response) of diseases with dysbiotic microbiota community. However, multiple attempts to leverage gut microbiota taxonomic data for diagnostic and disease characterization have failed due to significant inter-individual variability of microbiota community and overlap of disrupted microbiota communities among multiple diseases. One potential approach is to look at the microbiota community pattern and response to microbiota modifiers like dietary fiber in different disease states. This approach is now feasible by availability of machine learning that is able to identify hidden patterns in the human microbiome and predict diseases. Accordingly, the aim of our study was to test the hypothesis that application of machine learning algorithms can distinguish stool microbiota pattern and microbiota response to fiber between diseases where overlapping dysbiotic microbiota have been previously reported. Here, we have applied machine learning algorithms to distinguish between Parkinson's disease, Crohn's disease (CD), ulcerative colitis (UC), human immune deficiency virus (HIV), and healthy control (HC) subjects in the presence and absence of fiber treatments. We have shown that machine learning algorithms can classify diseases with accuracy as high as 95%. Furthermore, machine learning methods applied to the microbiome data to predict UC vs CD led to prediction accuracy as high as 90%.

show abstract

Section: Resultsmentioning

confidence: 99%

Machine learning based gut microbiota pattern and response to fiber as a diagnostic tool for chronic inflammatory diseases

Boodaghidizaji

Ardekani

2023

Preprint

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 99%

“…ResNet, a more frequently used CNN architecture in IBD, has been used for differential diagnosis of CD, grading CD ulcerations, and grading UC for MES, but with variable performance and limited reproducibility. [29][30][31][32][33] Specifically, the accuracy of ResNet predicting endoscopic remission was 72.02%, compared with 84.52% of InceptionV3, 73.81% of VGG19, and 87.50% of DenseNet, respectively. To the best of our knowledge, although it has been the only study reported on the comparison of different CNN architectures in grading endoscopic images of UC so far, the limitations were marked with small samples, unsatisfactory performance, class imbalance among more moderate/severe cases, selection bias, and lack of multiclass classifcation.…”

Section: Mesmentioning

confidence: 97%

CB‐HRNet: A Class‐Balanced High‐Resolution Network for the evaluation of endoscopic activity in patients with ulcerative colitis

Wang

Zhang

Liu

et al. 2023

Clinical Translational Sci

View full text Add to dashboard Cite

Endoscopic evaluation is the key to the management of ulcerative colitis (UC). However, there is interobserver variability in interpreting endoscopic images among gastroenterologists. Furthermore, it is time‐consuming. Convolutional neural networks (CNNs) can help overcome these obstacles and has yielded preliminary positive results. We aimed to develop a new CNN‐based algorithm to improve the performance for evaluation tasks of endoscopic images in patients with UC. A total of 12,163 endoscopic images from 308 patients with UC were collected from January 2014 to December 2021. The training set and test set images were randomly divided into 37,515 and 3191 after excluding possible interference and data augmentation. Mayo Endoscopic Subscores (MES) were predicted by different CNN‐based models with different loss functions. Their performances were evaluated by several metrics. After comparing the results of different CNN‐based models with different loss functions, High‐Resolution Network with Class‐Balanced Loss achieved the best performances in all MES classification subtasks. It was especially great at determining endoscopic remission in UC, which achieved a high accuracy of 95.07% and good performances in other evaluation metrics with sensitivity 92.87%, specificity 95.41%, kappa coefficient 0.8836, positive predictive value 93.44%, negative predictive value 95.00% and area value under the receiver operating characteristic curve 0.9834, respectively. In conclusion, we proposed a new CNN‐based algorithm, Class‐Balanced High‐Resolution Network (CB‐HRNet), to evaluate endoscopic activity of UC with excellent performance. Besides, we made an open‐source dataset and it can be a new benchmark in the task of MES classification.

show abstract

“…Although most clinical studies on lesion detection in endoscopic images have been conducted using the open-source AI algorithm, the excellent results are highly likely to be obtained only by a few Japanese companies that oliogopolize the gastrointestinal endoscopy market. Ruan et al have reported a high identification accuracy of a deep learning diagnostic system for inflammatory bowel disease ( 20 ). All endoscopic examinations were performed using an Olympus CV-290SL model.…”

Section: Discussionmentioning

confidence: 99%

Reduced detection rate of artificial intelligence in images obtained from untrained endoscope models and improvement using domain adaptation algorithm

Park

Hwang²,

Kim³

et al. 2022

Front. Med.

View full text Add to dashboard Cite

A training dataset that is limited to a specific endoscope model can overfit artificial intelligence (AI) to its unique image characteristics. The performance of the AI may degrade in images of different endoscope model. The domain adaptation algorithm, i.e., the cycle-consistent adversarial network (cycleGAN), can transform the image characteristics into AI-friendly styles. We attempted to confirm the performance degradation of AIs in images of various endoscope models and aimed to improve them using cycleGAN transformation. Two AI models were developed from data of esophagogastroduodenoscopies collected retrospectively over 5 years: one for identifying the endoscope models, Olympus CV-260SL, CV-290 (Olympus, Tokyo, Japan), and PENTAX EPK-i (PENTAX Medical, Tokyo, Japan), and the other for recognizing the esophagogastric junction (EGJ). The AIs were trained using 45,683 standardized images from 1,498 cases and validated on 624 separate cases. Between the two endoscope manufacturers, there was a difference in image characteristics that could be distinguished without error by AI. The accuracy of the AI in recognizing gastroesophageal junction was >0.979 in the same endoscope-examined validation dataset as the training dataset. However, they deteriorated in datasets from different endoscopes. Cycle-consistent adversarial network can successfully convert image characteristics to ameliorate the AI performance. The improvements were statistically significant and greater in datasets from different endoscope manufacturers [original → AI-trained style, increased area under the receiver operating characteristic (ROC) curve, P-value: CV-260SL → CV-290, 0.0056, P = 0.0106; CV-260SL → EPK-i, 0.0182, P = 0.0158; CV-290 → CV-260SL, 0.0134, P < 0.0001; CV-290 → EPK-i, 0.0299, P = 0.0001; EPK-i → CV-260SL, 0.0215, P = 0.0024; and EPK-i → CV-290, 0.0616, P < 0.0001]. In conclusion, cycleGAN can transform the diverse image characteristics of endoscope models into an AI-trained style to improve the detection performance of AI.

show abstract

Development and Validation of a Deep Neural Network for Accurate Identification of Endoscopic Images From Patients With Ulcerative Colitis and Crohn's Disease

Cited by 18 publications

References 28 publications

Machine learning based gut microbiota pattern and response to fiber as a diagnostic tool for chronic inflammatory diseases

Machine learning based gut microbiota pattern and response to fiber as a diagnostic tool for chronic inflammatory diseases

CB‐HRNet: A Class‐Balanced High‐Resolution Network for the evaluation of endoscopic activity in patients with ulcerative colitis

Reduced detection rate of artificial intelligence in images obtained from untrained endoscope models and improvement using domain adaptation algorithm

Contact Info

Product

Resources

About