Pub Date : 2019-09-30DOI: 10.21037/jmai.2019.08.01
P. Aggarwal
Background: Acute macular degeneration (AMD), central serous retinopathy (CSR), diabetic retinopathy (DR) and macular hole (MH) are common vision impairing pathologies in the field of ophthalmology. Machine learning with deep convolutional neural networks can be used to analyze ophthalmological diseases using fundus and optical coherence tomography (OCT) images, but with limited accuracy. In order to improve the sensitivity and specificity of these models, the objective of this study was to examine the effect of data augmentation on the performance of the neural network. Methods: OCT Images for above pathologies and normal eye were acquired from the Optical Coherence Tomography Image Database. Keras, a neural network framework, was used to retrain Visual Geometry Group 16 (VGG16), a deep neural network, using these images. Retraining was performed with and without data augmentation on two separate models. Data augmentation techniques included rotation, shear, horizontal flip and Gaussian noise. Results: Average Matthews correlation coefficient (MCC) increased from 0.83 in the model without data augmentation to 0.93 in the model with data augmentation. Average statistical measures- sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), MCC and F1 score increased with data augmentation. The average area under the curve (AUC) increased from 0.91 to 0.97 with data augmentation addition. Conclusions: Data augmentation techniques can be used in machine learning to appreciably increase the accuracy of a deep convolutional neural network. In future applications, the model created in this analysis can be retrained with a higher quantity and better quality of images and provided to physicians as an aid when examining OCT images.
{"title":"Machine learning of retinal pathology in optical coherence tomography images","authors":"P. Aggarwal","doi":"10.21037/jmai.2019.08.01","DOIUrl":"https://doi.org/10.21037/jmai.2019.08.01","url":null,"abstract":"Background: Acute macular degeneration (AMD), central serous retinopathy (CSR), diabetic retinopathy (DR) and macular hole (MH) are common vision impairing pathologies in the field of ophthalmology. Machine learning with deep convolutional neural networks can be used to analyze ophthalmological diseases using fundus and optical coherence tomography (OCT) images, but with limited accuracy. In order to improve the sensitivity and specificity of these models, the objective of this study was to examine the effect of data augmentation on the performance of the neural network. \u0000 Methods: OCT Images for above pathologies and normal eye were acquired from the Optical Coherence Tomography Image Database. Keras, a neural network framework, was used to retrain Visual Geometry Group 16 (VGG16), a deep neural network, using these images. Retraining was performed with and without data augmentation on two separate models. Data augmentation techniques included rotation, shear, horizontal flip and Gaussian noise. \u0000 Results: Average Matthews correlation coefficient (MCC) increased from 0.83 in the model without data augmentation to 0.93 in the model with data augmentation. Average statistical measures- sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), MCC and F1 score increased with data augmentation. The average area under the curve (AUC) increased from 0.91 to 0.97 with data augmentation addition. \u0000 Conclusions: Data augmentation techniques can be used in machine learning to appreciably increase the accuracy of a deep convolutional neural network. In future applications, the model created in this analysis can be retrained with a higher quantity and better quality of images and provided to physicians as an aid when examining OCT images.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.21037/jmai.2019.08.01","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49627138","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-21DOI: 10.21037/jmai.2019.09.04
Brandon J. Teng, M. Byrne
Colorectal cancer (CRC) is the second leading cause of cancer death, and is a significant cause of morbidity and mortality. This is a growing topic discussed on public media networks due to the worldwide rise in CRC incidence among people under 50 years of age and recent American Cancer Society recommendations for earlier CRC screening. Colonoscopy remains the most effective method of detection and removal of neoplastic polyps.
{"title":"Artificial intelligence and colorectal polyp detection","authors":"Brandon J. Teng, M. Byrne","doi":"10.21037/jmai.2019.09.04","DOIUrl":"https://doi.org/10.21037/jmai.2019.09.04","url":null,"abstract":"Colorectal cancer (CRC) is the second leading cause of cancer death, and is a significant cause of morbidity and mortality. This is a growing topic discussed on public media networks due to the worldwide rise in CRC incidence among people under 50 years of age and recent American Cancer Society recommendations for earlier CRC screening. Colonoscopy remains the most effective method of detection and removal of neoplastic polyps.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.21037/jmai.2019.09.04","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48307610","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-18DOI: 10.21037/jmai.2019.09.03
J. Mandel, E. Prochownik
Current approaches to cancer diagnosis and management are predicated on several fundamental principles including early detection, accurate and precise diagnosis and staging, and the induction and long-term maintenance of complete remission (1).
{"title":"Liquid biopsies and the promise of what might(o) be","authors":"J. Mandel, E. Prochownik","doi":"10.21037/jmai.2019.09.03","DOIUrl":"https://doi.org/10.21037/jmai.2019.09.03","url":null,"abstract":"Current approaches to cancer diagnosis and management are predicated on several fundamental principles including early detection, accurate and precise diagnosis and staging, and the induction and long-term maintenance of complete remission (1).","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43143139","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-17DOI: 10.21037/jmai.2019.09.02
O. Ahmad, L. Lovat
Colorectal cancer (CRC) is a leading cause of cancer-related mortality worldwide. Colonoscopy is protective against CRC through the detection and removal of neoplastic polyps. Unfortunately, the procedure is highly operator dependent with significant miss rates for polyps. Artificial intelligence (AI) and computer-aided detection software offers a promising solution by providing real-time assistance to highlight lesions that may otherwise be overlooked. Rapid advances have occurred in the field with recent prospective clinical trials demonstrating an improved adenoma detection rate (ADR) with AI assistance. Deployment in routine clinical practice is possible in the near future although further robust clinical trials are necessary and important practical challenges relating to real-world implementation must be addressed.
{"title":"Artificial intelligence for colorectal polyp detection: are we ready for prime time?","authors":"O. Ahmad, L. Lovat","doi":"10.21037/jmai.2019.09.02","DOIUrl":"https://doi.org/10.21037/jmai.2019.09.02","url":null,"abstract":"Colorectal cancer (CRC) is a leading cause of cancer-related mortality worldwide. Colonoscopy is protective against CRC through the detection and removal of neoplastic polyps. Unfortunately, the procedure is highly operator dependent with significant miss rates for polyps. Artificial intelligence (AI) and computer-aided detection software offers a promising solution by providing real-time assistance to highlight lesions that may otherwise be overlooked. Rapid advances have occurred in the field with recent prospective clinical trials demonstrating an improved adenoma detection rate (ADR) with AI assistance. Deployment in routine clinical practice is possible in the near future although further robust clinical trials are necessary and important practical challenges relating to real-world implementation must be addressed.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.21037/jmai.2019.09.02","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44336049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the near future, nearly every type of clinician, from paramedics to certificated medical specialists, will be expected to utilize artificial intelligence (AI) technology, and deep learning (DL) in particular (1). In terms of exceeding human ability, DL has been the backbone of computer science. DL mostly involves automated feature extraction using deep neural networks (DNNs), which can aid in the classification and discrimination of medical images, including mammograms, skin lesions, pathological slides, radiological images, and retinal fundus photographs.
{"title":"Toward the transparency of deep learning in radiological imaging: beyond quantitative to qualitative artificial intelligence","authors":"Y. Hayashi","doi":"10.21037/jmai.2019.09.0","DOIUrl":"https://doi.org/10.21037/jmai.2019.09.0","url":null,"abstract":"In the near future, nearly every type of clinician, from paramedics to certificated medical specialists, will be expected to utilize artificial intelligence (AI) technology, and deep learning (DL) in particular (1). In terms of exceeding human ability, DL has been the backbone of computer science. DL mostly involves automated feature extraction using deep neural networks (DNNs), which can aid in the classification and discrimination of medical images, including mammograms, skin lesions, pathological slides, radiological images, and retinal fundus photographs.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45469289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-08-05DOI: 10.21037/JMAI.2019.04.03
S. Anwar, Ulas Bagci
Breast cancer is the most common cancer in women worldwide and the second most common cancer overall, hence it is a significant public health concern. According to Global Health Estimates (1), over half a million women died in 2011 due to breast cancer.
{"title":"Artificial intelligence as another set of eyes in breast cancer diagnosis","authors":"S. Anwar, Ulas Bagci","doi":"10.21037/JMAI.2019.04.03","DOIUrl":"https://doi.org/10.21037/JMAI.2019.04.03","url":null,"abstract":"Breast cancer is the most common cancer in women worldwide and the second most common cancer overall, hence it is a significant public health concern. According to Global Health Estimates (1), over half a million women died in 2011 due to breast cancer.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.21037/JMAI.2019.04.03","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45848756","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-18DOI: 10.21037/JMAI.2019.06.02
W. Schmidt, M. Regan, M. Fahey, A. Paplinski
The current rate of cerebral palsy (CP) per live births in Australia is between 0.14% and 0.2%, worldwide the rate has been static for 60 years at 0.2%. Typically a CP diagnosis is delayed until around age 2 years; this delay decreases the likelihood of a long-term positive patient outcome. Current early detection is by visual examination of newborns 10 to 20 weeks post gestation. A screening program based on filming babies and processing the video via artificial intelligence (AI) will allow increased early detection and intervention. This paper outlines the practical development, and initial results from, a recurrent deep neural net solution for the classification of newborn videos, specifically targeting CP, using the largest fidgety movements dataset in Australia.
{"title":"General movement assessment by machine learning: why is it so difficult?","authors":"W. Schmidt, M. Regan, M. Fahey, A. Paplinski","doi":"10.21037/JMAI.2019.06.02","DOIUrl":"https://doi.org/10.21037/JMAI.2019.06.02","url":null,"abstract":"The current rate of cerebral palsy (CP) per live births in Australia is between 0.14% and 0.2%, worldwide the rate has been static for 60 years at 0.2%. Typically a CP diagnosis is delayed until around age 2 years; this delay decreases the likelihood of a long-term positive patient outcome. Current early detection is by visual examination of newborns 10 to 20 weeks post gestation. A screening program based on filming babies and processing the video via artificial intelligence (AI) will allow increased early detection and intervention. This paper outlines the practical development, and initial results from, a recurrent deep neural net solution for the classification of newborn videos, specifically targeting CP, using the largest fidgety movements dataset in Australia.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.21037/JMAI.2019.06.02","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47057054","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.21037/JMAI.2019.06.03
Brandon J. Ten, M. Byrne
Gastrointestinal (GI) cancer is a leading cause worldwide of morbidity and mortality. In 2018, GI cancer accounted for 27% of all new cancer diagnoses. The incidence rate of colorectal cancer is rising in many countries, with a recent dramatic increase for people under the age of 50 years.
{"title":"Introduction for the Artificial Intelligence and Gastrointestinal Cancer Column","authors":"Brandon J. Ten, M. Byrne","doi":"10.21037/JMAI.2019.06.03","DOIUrl":"https://doi.org/10.21037/JMAI.2019.06.03","url":null,"abstract":"Gastrointestinal (GI) cancer is a leading cause worldwide of morbidity and mortality. In 2018, GI cancer accounted for 27% of all new cancer diagnoses. The incidence rate of colorectal cancer is rising in many countries, with a recent dramatic increase for people under the age of 50 years.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.21037/JMAI.2019.06.03","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48678961","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-06-11DOI: 10.21037/JMAI.2019.05.02
S. L. Van Es, A. Madabhushi
Dietz and Pantanowitz (1) present a well-explained and informative viewpoint on the history, theory, and science behind, as well as, current and potential future uses and challenges of, artificial intelligence (AI) and machine learning (ML), for pathology. They emphasize the importance of development of a “killer suite” of AI applications whose use is evidence-based, that will accelerate acceptance and integration of digital pathology (DP) into diagnostic practice. This is an invited reflection on their editorial content with reference to findings from other groups.
{"title":"The revolving door for AI and pathologists-docendo discimus?","authors":"S. L. Van Es, A. Madabhushi","doi":"10.21037/JMAI.2019.05.02","DOIUrl":"https://doi.org/10.21037/JMAI.2019.05.02","url":null,"abstract":"Dietz and Pantanowitz (1) present a well-explained and informative viewpoint on the history, theory, and science behind, as well as, current and potential future uses and challenges of, artificial intelligence (AI) and machine learning (ML), for pathology. They emphasize the importance of development of a “killer suite” of AI applications whose use is evidence-based, that will accelerate acceptance and integration of digital pathology (DP) into diagnostic practice. This is an invited reflection on their editorial content with reference to findings from other groups.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.21037/JMAI.2019.05.02","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42242176","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-05-24DOI: 10.21037/JMAI.2019.05.01
Y. Mori, S. Kudo, M. Misawa
Colorectal cancer (CRC) is a major cause of cancer-related mortality in most countries. Colonoscopy during which all neoplastic and pre-malignant polyps (e.g., adenomas) are eradicated is considered beneficial in decreasing the incidence of CRCs and their associated mortality (1,2). This concept has been supported by several large-scale prospective studies (3). The quality of the colonoscopy procedure, however, varies according to the expertise of the endoscopist.
{"title":"Detecting colorectal polyps with use of artificial intelligence","authors":"Y. Mori, S. Kudo, M. Misawa","doi":"10.21037/JMAI.2019.05.01","DOIUrl":"https://doi.org/10.21037/JMAI.2019.05.01","url":null,"abstract":"Colorectal cancer (CRC) is a major cause of cancer-related mortality in most countries. Colonoscopy during which all neoplastic and pre-malignant polyps (e.g., adenomas) are eradicated is considered beneficial in decreasing the incidence of CRCs and their associated mortality (1,2). This concept has been supported by several large-scale prospective studies (3). The quality of the colonoscopy procedure, however, varies according to the expertise of the endoscopist.","PeriodicalId":73815,"journal":{"name":"Journal of medical artificial intelligence","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.21037/JMAI.2019.05.01","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44736507","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：