Development of deep-learning tool to predict appropriate height and weight from chest radiographs in bedridden patients
Summary: Background & Aims: Malnutrition in the elderly, frequently and significantly affects both physical functioning and cognition, as well as incurs direct and indirect costs to society. Guidelines recommend rapid nutritional intervention and initiation of nutritional therapy within 24–...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-10-01
|
| Series: | Clinical Nutrition Open Science |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2667268523000396 |
| _version_ | 1797659938246361088 |
|---|---|
| author | Yasuhiko Nakao Ryo Sasaki Fumihiro Mawatari Kotaro Harakawa Minoru Okita Norisato Mitsutake Kazuhiko Nakao |
| author_facet | Yasuhiko Nakao Ryo Sasaki Fumihiro Mawatari Kotaro Harakawa Minoru Okita Norisato Mitsutake Kazuhiko Nakao |
| author_sort | Yasuhiko Nakao |
| collection | DOAJ |
| description | Summary: Background & Aims: Malnutrition in the elderly, frequently and significantly affects both physical functioning and cognition, as well as incurs direct and indirect costs to society. Guidelines recommend rapid nutritional intervention and initiation of nutritional therapy within 24–48 hours of admission. Height and weight information is essential for proper nutritional assessment; however, it is difficult to obtain individual the height and weight of bedridden elderly patients directly. This study aimed to illustrate the potential of a convolutional neural network model to assess the height and weight based on chest radiographs. Methods: We retrospectively evaluated radiographs obtained over 15 years of follow-up. Overall, 6,453 radiographs from male patients, and 7,879 from female patients were included in the analysis. A convolutional neural network was used to predict the height and weight of the patients (Juzen NST). A ResNet152 classifier was trained using Fastai (V1.0) running on PyTorch to predict the height and weight. Training was performed for four epochs using validation without augmentation. Results: The correlation coefficients between the predicted and measured values using the height prediction model for males and females were R=0.855 and R=0.81, respectively. The correlation coefficients between the values predicted by the weight prediction model and measured values were R=0.793 and R=0.86, respectively. Conclusion: Our chest radiographic prediction model has a high correlation with actual height and weight and can be combined with information regarding clinical nutrition factors for rapid assessment of risk for malnutrition. By training the prediction model using chest radiographs from each hospital, it can be optimized for the most common ethnic groups in the area. Importantly, our model is an illustration of the potential of automated imaging artificial intelligence for proper nutrition prediction models in older adults. |
| first_indexed | 2024-03-11T18:22:33Z |
| format | Article |
| id | doaj.art-caef20dbd20a4abd9fb4d0665c0e9b3c |
| institution | Directory Open Access Journal |
| issn | 2667-2685 |
| language | English |
| last_indexed | 2024-03-11T18:22:33Z |
| publishDate | 2023-10-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Clinical Nutrition Open Science |
| spelling | doaj.art-caef20dbd20a4abd9fb4d0665c0e9b3c2023-10-15T04:38:25ZengElsevierClinical Nutrition Open Science2667-26852023-10-0151109117Development of deep-learning tool to predict appropriate height and weight from chest radiographs in bedridden patientsYasuhiko Nakao0Ryo Sasaki1Fumihiro Mawatari2Kotaro Harakawa3Minoru Okita4Norisato Mitsutake5Kazuhiko Nakao6Department of Gastroenterology and Hepatology, Juzenkai Hospital, Japan; Department of Gastroenterology and Hepatology, Graduate School of Biomedical Sciences, Nagasaki University, Japan; Corresponding author. Department of Gastroenterology and Hepatology, Nagasaki University, 1-7-1 Sakamoto, Nagasaki 852-8501, Japan. Tel.: +81 507 284 0686; fax: +81 507 284 0762.Department of Rehabilitation, Juzenkai Hospital, Japan; Department of Physical Therapy Science, Graduate School of Biomedical Sciences Nagasaki University, JapanDepartment of Gastroenterology and Hepatology, Juzenkai Hospital, Japan; Department of Gastroenterology and Hepatology, Graduate School of Biomedical Sciences, Nagasaki University, JapanDepartment of Radiation Medical Sciences, Atomic Bomb Disease Institute Nagasaki University, Nagasaki, JapanDepartment of Physical Therapy Science, Graduate School of Biomedical Sciences Nagasaki University, JapanDepartment of Radiation Medical Sciences, Atomic Bomb Disease Institute Nagasaki University, Nagasaki, JapanDepartment of Gastroenterology and Hepatology, Graduate School of Biomedical Sciences, Nagasaki University, JapanSummary: Background & Aims: Malnutrition in the elderly, frequently and significantly affects both physical functioning and cognition, as well as incurs direct and indirect costs to society. Guidelines recommend rapid nutritional intervention and initiation of nutritional therapy within 24–48 hours of admission. Height and weight information is essential for proper nutritional assessment; however, it is difficult to obtain individual the height and weight of bedridden elderly patients directly. This study aimed to illustrate the potential of a convolutional neural network model to assess the height and weight based on chest radiographs. Methods: We retrospectively evaluated radiographs obtained over 15 years of follow-up. Overall, 6,453 radiographs from male patients, and 7,879 from female patients were included in the analysis. A convolutional neural network was used to predict the height and weight of the patients (Juzen NST). A ResNet152 classifier was trained using Fastai (V1.0) running on PyTorch to predict the height and weight. Training was performed for four epochs using validation without augmentation. Results: The correlation coefficients between the predicted and measured values using the height prediction model for males and females were R=0.855 and R=0.81, respectively. The correlation coefficients between the values predicted by the weight prediction model and measured values were R=0.793 and R=0.86, respectively. Conclusion: Our chest radiographic prediction model has a high correlation with actual height and weight and can be combined with information regarding clinical nutrition factors for rapid assessment of risk for malnutrition. By training the prediction model using chest radiographs from each hospital, it can be optimized for the most common ethnic groups in the area. Importantly, our model is an illustration of the potential of automated imaging artificial intelligence for proper nutrition prediction models in older adults.http://www.sciencedirect.com/science/article/pii/S2667268523000396HeightWeightBody mass indexBedridden patientsDeep learningChest radiograph |
| spellingShingle | Yasuhiko Nakao Ryo Sasaki Fumihiro Mawatari Kotaro Harakawa Minoru Okita Norisato Mitsutake Kazuhiko Nakao Development of deep-learning tool to predict appropriate height and weight from chest radiographs in bedridden patients Clinical Nutrition Open Science Height Weight Body mass index Bedridden patients Deep learning Chest radiograph |
| title | Development of deep-learning tool to predict appropriate height and weight from chest radiographs in bedridden patients |
| title_full | Development of deep-learning tool to predict appropriate height and weight from chest radiographs in bedridden patients |
| title_fullStr | Development of deep-learning tool to predict appropriate height and weight from chest radiographs in bedridden patients |
| title_full_unstemmed | Development of deep-learning tool to predict appropriate height and weight from chest radiographs in bedridden patients |
| title_short | Development of deep-learning tool to predict appropriate height and weight from chest radiographs in bedridden patients |
| title_sort | development of deep learning tool to predict appropriate height and weight from chest radiographs in bedridden patients |
| topic | Height Weight Body mass index Bedridden patients Deep learning Chest radiograph |
| url | http://www.sciencedirect.com/science/article/pii/S2667268523000396 |
| work_keys_str_mv | AT yasuhikonakao developmentofdeeplearningtooltopredictappropriateheightandweightfromchestradiographsinbedriddenpatients AT ryosasaki developmentofdeeplearningtooltopredictappropriateheightandweightfromchestradiographsinbedriddenpatients AT fumihiromawatari developmentofdeeplearningtooltopredictappropriateheightandweightfromchestradiographsinbedriddenpatients AT kotaroharakawa developmentofdeeplearningtooltopredictappropriateheightandweightfromchestradiographsinbedriddenpatients AT minoruokita developmentofdeeplearningtooltopredictappropriateheightandweightfromchestradiographsinbedriddenpatients AT norisatomitsutake developmentofdeeplearningtooltopredictappropriateheightandweightfromchestradiographsinbedriddenpatients AT kazuhikonakao developmentofdeeplearningtooltopredictappropriateheightandweightfromchestradiographsinbedriddenpatients |