English
Arabic
Chinese (Simplified)
Dutch
Esperanto
French
German
Greek
Italian
Japanese
Korean
Portuguese
Russian
Spanish
UrduThese authors aimed to develop a deep learning model to diagnose influenza infection using pharyngeal images and clinical information. They recruited patients who visited clinics and hospitals because of influenza-like symptoms.
In the training stage, the authors developed a diagnostic prediction artificial intelligence (AI) model based on deep learning to predict polymerase chain reaction (PCR)–confirmed influenza from pharyngeal images and clinical information. In the validation stage, they assessed the diagnostic performance of the AI model. In an additional analysis, the authors compared the diagnostic performance of the AI model with that of 3 physicians and interpreted the AI model using importance heat maps.
This process led to the development of the first AI model that can accurately diagnose influenza.
Dr Sho Okiyama, MD, from Aillis, Inc said, “According to the Global Burden of Disease Study 2016, the global burden of influenza is substantial.” Timely and accurate diagnosis of influenza has the potential to prevent widespread transmission of the virus within the population and during subsequent epidemics and pandemics, as well as to prevent the unnecessary prescription of antibiotics in primary care, which is a cause of emerging antibiotic-resistant bacteria.
The COVID-19 pandemic and surge in the use of telemedicine highlighted the importance of accurately diagnosing influenza infection without increasing the risk of spreading the virus through physical interaction. The gold-standard method for diagnosing influenza infection is the reverse transcription–PCR (RT-PCR) of nasopharyngeal aspirates or swabs; however, RT-PCR is not easily performed in primary care, and the result’s turnaround time could delay prompt diagnosis and preventive or treatment interventions.
Neither of these tests can be performed through telemedicine, and the sensitivity and specificity of diagnosing influenza using clinical information only are suboptimal. Given the recent increase in the number of patients being diagnosed through telemedicine, an alternative influenza test that can be conducted through telemedicine is warranted.
Dr Okiyama and the research team concluded in their JMIR Publications Research Article, “we developed the first AI-assisted diagnostic camera for influenza and prospectively validated its high performance. We found that the AI model often focused on follicles, which confirmed previous case reports and series suggesting that visual inspection of the pharynx would help in the diagnosis of influenza infection.”
About the Journal of Medical Internet Research
The Journal of Medical Internet Research (JMIR) (founded in 1999, now in its 23rd year!), is the pioneer open access eHealth journal and is the flagship journal of JMIR Publications. It is a leading digital health journal globally in terms of quality/visibility (Journal Impact Factor™ 7.08 (Clarivate, 2022)) and is also the largest journal in the field. The journal focuses on emerging technologies, medical devices, apps, engineering, telehealth and informatics applications for patient education, prevention, population health and clinical care.
JMIR is indexed in all major literature indices including MEDLINE, PubMed/PMC, Scopus, Psycinfo, SCIE, JCR, EBSCO/EBSCO Essentials, DOAJ, GoOA and others. As a leading high-impact journal in its disciplines, ranking Q1 in both the ‘Medical Informatics’ and ‘Health Care Sciences and Services’ categories, it is a selective journal complemented by almost 30 specialty JMIR sister journals, which have a broader scope, and which together receive over 6.000 submissions a year.
As an open access journal, we are read by clinicians, allied health professionals, informal caregivers, and patients alike, and have (as with all JMIR journals) a focus on readable and applied science reporting the design and evaluation of health innovations and emerging technologies. We publish original research, viewpoints, and reviews (both literature reviews and medical device/technology/app reviews). Peer-review reports are portable across JMIR journals and papers can be transferred, so authors save time by not having to resubmit a paper to a different journal but can simply transfer it between journals.
We are also a leader in participatory and open science approaches, and offer the option to publish new submissions immediately as preprints, which receive DOIs for immediate citation (eg, in grant proposals), and for open peer-review purposes. We also invite patients to participate (eg, as peer-reviewers) and have patient representatives on editorial boards.
As all JMIR journals, the journal encourages Open Science principles and strongly encourages publication of a protocol before data collection. Authors who have published a protocol in JMIR Research Protocols get a discount of 20% on the Article Processing Fee when publishing a subsequent results paper in any JMIR journal.
Be a widely cited leader in the digital health revolution and submit your paper today!
###
DOI – https://doi.org/10.2196/38751
Full-text – https://www.jmir.org/2022/12/e38751/
Corresponding author – Sho Okiyama, MD, Aillis, Inc, 1-10-1-11F, Yurakucho, Chiyoda-ku, Tokyo, JP
Phone – 81 3-5218-2374
Email – [email protected]
Keywords – influenza, physical examination, pharynx, deep learning, diagnostic prediction
About JMIR Publications
JMIR Publications is a leading, born-digital, open access publisher of 30+ academic journals and other innovative scientific communication products that focus on the intersection of health and technology. Its flagship journal, the Journal of Medical Internet Research, is the leading digital health journal globally in content breadth and visibility, and it is the largest journal in the medical informatics field.
To learn more about JMIR Publications, please visit https://www.JMIRPublications.com or connect with us via:
YouTube – https://www.youtube.com/c/JMIRPublications
Facebook – https://www.facebook.com/JMedInternetRes
Twitter – https://twitter.com/jmirpub
LinkedIn – https://www.linkedin.com/company/jmir-publications
Instagram – https://www.instagram.com/jmirpub/
Head Office – 130 Queens Quay East, Unit 1100 Toronto, ON, M5A 0P6 Canada
Media Contact – [email protected]
The content of this communication is licensed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, published by JMIR Publications, is properly cited.
