Deep learning based phenotyping of medical images improves power for gene discovery of complex disease
Abstract Electronic health records are often incomplete, reducing the power of genetic association studies. For some diseases, such as knee osteoarthritis where the routine course of diagnosis involves an X-ray, image-based phenotyping offers an alternate and unbiased way to ascertain disease cases....
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-08-01
|
| Series: | npj Digital Medicine |
| Online Access: | https://doi.org/10.1038/s41746-023-00903-x |
| _version_ | 1797451405874692096 |
|---|---|
| author | Brianna I. Flynn Emily M. Javan Eugenia Lin Zoe Trutner Karl Koenig Kenoma O. Anighoro Eucharist Kun Alaukik Gupta Tarjinder Singh Prakash Jayakumar Vagheesh M. Narasimhan |
| author_facet | Brianna I. Flynn Emily M. Javan Eugenia Lin Zoe Trutner Karl Koenig Kenoma O. Anighoro Eucharist Kun Alaukik Gupta Tarjinder Singh Prakash Jayakumar Vagheesh M. Narasimhan |
| author_sort | Brianna I. Flynn |
| collection | DOAJ |
| description | Abstract Electronic health records are often incomplete, reducing the power of genetic association studies. For some diseases, such as knee osteoarthritis where the routine course of diagnosis involves an X-ray, image-based phenotyping offers an alternate and unbiased way to ascertain disease cases. We investigated this by training a deep-learning model to ascertain knee osteoarthritis cases from knee DXA scans that achieved clinician-level performance. Using our model, we identified 1931 (178%) more cases than currently diagnosed in the health record. Individuals diagnosed as cases by our model had higher rates of self-reported knee pain, for longer durations and with increased severity compared to control individuals. We trained another deep-learning model to measure the knee joint space width, a quantitative phenotype linked to knee osteoarthritis severity. In performing genetic association analysis, we found that use of a quantitative measure improved the number of genome-wide significant loci we discovered by an order of magnitude compared with our binary model of cases and controls despite the two phenotypes being highly genetically correlated. In addition we discovered associations between our quantitative measure of knee osteoarthritis and increased risk of adult fractures- a leading cause of injury-related death in older individuals-, illustrating the capability of image-based phenotyping to reveal epidemiological associations not captured in the electronic health record. For diseases with radiographic diagnosis, our results demonstrate the potential for using deep learning to phenotype at biobank scale, improving power for both genetic and epidemiological association analysis. |
| first_indexed | 2024-03-09T14:53:14Z |
| format | Article |
| id | doaj.art-647e7ff6418b41ecb8014834aa3790ff |
| institution | Directory Open Access Journal |
| issn | 2398-6352 |
| language | English |
| last_indexed | 2024-03-09T14:53:14Z |
| publishDate | 2023-08-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | npj Digital Medicine |
| spelling | doaj.art-647e7ff6418b41ecb8014834aa3790ff2023-11-26T14:19:31ZengNature Portfolionpj Digital Medicine2398-63522023-08-016111210.1038/s41746-023-00903-xDeep learning based phenotyping of medical images improves power for gene discovery of complex diseaseBrianna I. Flynn0Emily M. Javan1Eugenia Lin2Zoe Trutner3Karl Koenig4Kenoma O. Anighoro5Eucharist Kun6Alaukik Gupta7Tarjinder Singh8Prakash Jayakumar9Vagheesh M. Narasimhan10Department of Integrative Biology, The University of Texas at AustinDepartment of Integrative Biology, The University of Texas at AustinDepartment of Surgery and Perioperative Care, Dell Medical SchoolDepartment of Surgery and Perioperative Care, Dell Medical SchoolDepartment of Surgery and Perioperative Care, Dell Medical SchoolDepartment of Surgery and Perioperative Care, Dell Medical SchoolDepartment of Integrative Biology, The University of Texas at AustinDepartment of Integrative Biology, The University of Texas at AustinThe Department of Psychiatry at Columbia University Irving Medical CenterDepartment of Surgery and Perioperative Care, Dell Medical SchoolDepartment of Integrative Biology, The University of Texas at AustinAbstract Electronic health records are often incomplete, reducing the power of genetic association studies. For some diseases, such as knee osteoarthritis where the routine course of diagnosis involves an X-ray, image-based phenotyping offers an alternate and unbiased way to ascertain disease cases. We investigated this by training a deep-learning model to ascertain knee osteoarthritis cases from knee DXA scans that achieved clinician-level performance. Using our model, we identified 1931 (178%) more cases than currently diagnosed in the health record. Individuals diagnosed as cases by our model had higher rates of self-reported knee pain, for longer durations and with increased severity compared to control individuals. We trained another deep-learning model to measure the knee joint space width, a quantitative phenotype linked to knee osteoarthritis severity. In performing genetic association analysis, we found that use of a quantitative measure improved the number of genome-wide significant loci we discovered by an order of magnitude compared with our binary model of cases and controls despite the two phenotypes being highly genetically correlated. In addition we discovered associations between our quantitative measure of knee osteoarthritis and increased risk of adult fractures- a leading cause of injury-related death in older individuals-, illustrating the capability of image-based phenotyping to reveal epidemiological associations not captured in the electronic health record. For diseases with radiographic diagnosis, our results demonstrate the potential for using deep learning to phenotype at biobank scale, improving power for both genetic and epidemiological association analysis.https://doi.org/10.1038/s41746-023-00903-x |
| spellingShingle | Brianna I. Flynn Emily M. Javan Eugenia Lin Zoe Trutner Karl Koenig Kenoma O. Anighoro Eucharist Kun Alaukik Gupta Tarjinder Singh Prakash Jayakumar Vagheesh M. Narasimhan Deep learning based phenotyping of medical images improves power for gene discovery of complex disease npj Digital Medicine |
| title | Deep learning based phenotyping of medical images improves power for gene discovery of complex disease |
| title_full | Deep learning based phenotyping of medical images improves power for gene discovery of complex disease |
| title_fullStr | Deep learning based phenotyping of medical images improves power for gene discovery of complex disease |
| title_full_unstemmed | Deep learning based phenotyping of medical images improves power for gene discovery of complex disease |
| title_short | Deep learning based phenotyping of medical images improves power for gene discovery of complex disease |
| title_sort | deep learning based phenotyping of medical images improves power for gene discovery of complex disease |
| url | https://doi.org/10.1038/s41746-023-00903-x |
| work_keys_str_mv | AT briannaiflynn deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT emilymjavan deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT eugenialin deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT zoetrutner deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT karlkoenig deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT kenomaoanighoro deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT eucharistkun deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT alaukikgupta deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT tarjindersingh deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT prakashjayakumar deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease AT vagheeshmnarasimhan deeplearningbasedphenotypingofmedicalimagesimprovespowerforgenediscoveryofcomplexdisease |