The DOI matches its Crossref record; 1 of the 1 references it cites were found in a registry or an index.
Neural networks for classification and image generation of aging in genetic syndromes
Open accessPreprint — published as 10.3389/fgene.2022.864092 · Paper page
Abstract
In medical genetics, one application of neural networks is the diagnosis of genetic diseases based on images of patient faces. While these applications have been validated in the literature with primarily pediatric subjects, it is not known whether these applications can accurately diagnose patients across a lifespan. We aimed to extend previous works to determine whether age plays a factor in facial diagnosis, as well as to explore other factors that may contribute to the overall diagnosis accuracy. To investigate this, we chose two relatively common conditions, Williams syndrome and 22q11.2 deletion syndrome. We built a neural network classifier trained on images of affected and unaffected individuals of different ages. Our classifier outperformed clinical geneticists at recognizing face images of these two conditions within each of the age groups (the performance varied between the age groups): (1) under 2 years old, (2) 2-9 years old, (3) 10-19 years old, (4) 20-34 years old, and (5) ≥35 years old. The overall accuracy improvement by our classifier over the clinical geneticists was 15.5% and 22.7% for Williams syndrome and 22q11.2 deletion syndrome, respectively. Additionally, comparison of saliency maps revealed that key facial features learned by the neural network differed slightly with respect to age. Finally, joint training real images with multiple different types of fake images created by a generative adversarial network showed up to 3.25% accuracy gain in classification accuracy.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764
Flagged first, then by citations
Research works printed but not linked: 36
- 10.1038/s41591-018-0279-0 · DOI · Paper page
- Duong, D. , Waikel, R.L. , Hu, P. , Tekendo-Ngongang, C. , and Solomon, B.D. (2021). Proof-of-principle neural network models for classification, attribution, creation, style-mixing, and morphing of image data for genetic conditions. medRxiv. · DOI
- Porras, A.R. , Rosenbaum, K. , Tor-Diez, C. , Summar, M. , and Linguraru, M.G. (2021). Development and evaluation of a machine learning-based point-of-care screening tool for genetic syndromes in children: a multinational retrospective study. Lancet Digit Health. · DOI
- Hsieh, T.-C. , Bar-Haim, A. , Moosa, S. , Ehmke, N. , Gripp, K.W. , Pantel, J.T. , Danyel, M. , Mensah, M.A. , Horn, D. , and Fleischer, N. (2021). GestaltMatcher: Overcoming the limits of rare disease matching using facial phenotypic descriptors. medRxiv, 2020.2012.2028.20248193. · DOI
- 10.1038/s41467-019-09025-z · DOI
- Clark, M.M. , Hildreth, A. , Batalov, S. , Ding, Y. , Chowdhury, S. , Watkins, K. , Ellsworth, K. , Camp, B. , Kint, C.I. , Yacoubian, C. , et al. (2019). Diagnosis of genetic diseases in seriously ill children by rapid whole-genome sequencing and automated phenotyping and interpretation. Sci Transl Med 11. · DOI · Paper page
- Frazer, J. , Notin, P. , Dias, M. , Gomez, A. , Min, J.K. , Brock, K. , Gal, Y. , and Marks, D.S. (2021). Disease variant prediction with deep generative models of evolutionary data. Nature. · DOI · Paper page
- 10.1126/science.abj8754 · DOI · Paper page
- 10.1002/ajmg.a.61124 · DOI
- 10.1073/pnas.1302575110 · DOI
- 10.1038/s41436-018-0417-6 · DOI
- Jenkins, B.D. , Fischer, C.G. , Polito, C.A. , Maiese, D.R. , Keehn, A.S. , Lyon, M. , Edick, M.J. , Taylor, M.R.G. , Andersson, H.C. , Bodurtha, J.N. , et al. (2021). The 2019 US medical genetics workforce: a focus on clinical genetics. Genet Med. · DOI
- 10.1038/s41436-018-0398-5 · DOI
- 10.1542/peds.112.1.101 · DOI
- 2002, Prevalence estimation of Williams syndrome, J Child Neurol, 17, 269, 10.1177/088307380201700406 · DOI · Paper page
- 10.1136/adc.2003.026880 · DOI
- Morris, C.A. , Braddock, S.R. , and Council On, G. (2020). Health Care Supervision for Children With Williams Syndrome. Pediatrics 145. · DOI
- 2018, What is new with 22q? An update from the 22q and You Center at the Children’s Hospital of Philadelphia, Am J Med Genet A, 176, 2058, 10.1002/ajmg.a.40637 · DOI
- 2021, Automatic Facial Recognition of Williams-Beuren Syndrome Based on Deep Convolutional Neural Networks, Front Pediatr, 9, 648255, 10.3389/fped.2021.648255 · DOI
- Duong, D. , Waikel, R.L. , Hu, P. , Tekendo-Ngongang, C. , and Solomon, B.D. (2021). Neural network classifiers for images of genetic conditions with cutaneous manifestations. Human Genetics and Genomics Advances, 100053. · DOI
- DeGrave, A.J. , Janizek, J.D. , and Lee, S.I. (2020). AI for radiographic COVID-19 detection selects shortcuts over signal. medRxiv. · DOI
- 2021, Evolving use of ancestry, ethnicity, and race in genetics research-A survey spanning seven decades, Am J Hum Genet, 108, 2215, 10.1016/j.ajhg.2021.10.008 · DOI
- Tan, M. , and Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In International Conference on Machine Learning. (PMLR), pp 6105–6114.
- Meftah, S. , Semmar, N. , Tahiri, M.-A. , Tamaazousti, Y. , Essafi, H. , and Sadat, F. (2020). Multi-Task Supervised Pretraining for Neural Domain Adaptation. In Proceedings of the Eighth International Workshop on Natural Language Processing for Social Media. pp 61–71. · DOI
- 2018, Multi-task Learning for Universal Sentence Embeddings: A Thorough Evaluation using Transfer and Auxiliary Tasks, arXiv preprint
- 2019, Fairface: Face attribute dataset for balanced race, gender, and age, arXiv preprint
- 2018, The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions, Sci Data, 5, 180161, 10.1038/sdata.2018.161 · DOI · Paper page
- 10.1016/S1470-2045(19)30333-X · DOI · Paper page
- 2020, Training generative adversarial networks with limited data, arXiv preprint
- Or-El, R. , Sengupta, S. , Fried, O. , Shechtman, E. , and Kemelmacher-Shlizerman, I. (2020). Lifespan age transformation synthesis. In European Conference on Computer Vision. (Springer), pp 739–755. · DOI
- Zeiler, M.D. , and Fergus, R. (2014). Visualizing and understanding convolutional networks. In European conference on computer vision. (Springer), pp 818–833. · DOI · Paper page
- 10.1002/humu.22048 · DOI
- 2021, An online compendium of treatable genetic disorders, Am J Med Genet C Semin Med Genet, 187, 48, 10.1002/ajmg.c.31874 · DOI
- 2020, A GAN-based image synthesis method for skin lesion classification, Comput Methods Programs Biomed, 195, 105568, 10.1016/j.cmpb.2020.105568 · DOI
- 2018, Towards generative adversarial networks as a new paradigm for radiology education, arXiv preprint
- Solomon, B.D. (2021). Can artificial intelligence save medical genetics? Am J Med Genet A. · DOI
Metadata
- Authors
- Duong, D., Hu, P., Tekendo-Ngongang, C., Hanchard, S. L., Liu, S., Solomon, B. D., & Waikel, R. L.
- Affiliations
- National Human Genome Research Institute (US)
- Corresponding author
- Solomon, Benjamin D.
- Waikel, Rebekah L.
- Venue
- Frontiers in Genetics
- Publisher
- openRxiv
- Published
- 2021-12-11
- Volume
- 13
- Pages
- 864092-864092
- ISSN
- 1664-8021 (ISSN-L)
- PMC
- PMC9035665
- OpenAlex
- W4226350219
- Cited by
- 29
- References
- 1
- Citations by year
- 2026: 5 · 2025: 8 · 2024: 5 · 2023: 8 · 2022: 3
- FWCI
- 1.96
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- Frontiers in Genetics (publishedVersion, cc-by)
- medRxiv (acceptedVersion)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Funders
- National Institutes of Health
- National Human Genome Research Institute
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face recognition and analysis
- Life Sciences › Neuroscience › Cognitive Neuroscience › Face Recognition and Perception
- Life Sciences › Neuroscience › Developmental Neuroscience › Williams Syndrome Research
- Keywords
- Williams syndrome, 22q11.2 deletion syndrome, facial diagnosis, genetic syndromes, age as confounder, deep learning classification, saliency maps, generative adversarial networks, image-to-image translation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
