Automated skin lesion detection and prevalence estimation in Tamanend's bottlenose dolphins
Anthropogenic global change is occurring at alarming rates, leading to increased urgency in the ability to monitor wildlife health in real time. Monitoring sentinel marine species, such as bottlenose dolphins, is particularly important due to extensive anthropogenic modifications to their habitats....
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Ecological Informatics |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1574954124004849 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832595399777452032 |
|---|---|
| author | Colin J. Murphy Melissa A. Collier Ann-Marie Jacoby Eric M. Patterson Megan M. Wallen Janet Mann Shweta Bansal |
| author_facet | Colin J. Murphy Melissa A. Collier Ann-Marie Jacoby Eric M. Patterson Megan M. Wallen Janet Mann Shweta Bansal |
| author_sort | Colin J. Murphy |
| collection | DOAJ |
| description | Anthropogenic global change is occurring at alarming rates, leading to increased urgency in the ability to monitor wildlife health in real time. Monitoring sentinel marine species, such as bottlenose dolphins, is particularly important due to extensive anthropogenic modifications to their habitats. The most common non-invasive method of monitoring cetacean health is documentation of skin lesions, often associated with poor health or disease, but the current methodology is inefficient and imprecise. Recent advancements in technology, such as machine learning, can provide researchers with more efficient ecological monitoring methods to address health questions at both the population and the individual levels. Our work develops a machine learning model to classify skin lesions on the understudied Tamanend's bottlenose dolphins (Tursiops erebennus) of the Chesapeake Bay, using manual estimates of lesion presence in photographs. We assess the model's performance and find that our best model performs with a high mean average precision (65.6 %–86.8 %), and generally increased accuracy with improved photo quality. We also demonstrate the model's ability to address ecological questions across scales by generating model-based estimates of lesion prevalence and testing the effect of gregariousness on health status. At the population level, our model accurately estimates a prevalence of 72.1 % spot and 27.3 % fringe ring lesions, with a slight underprediction compared to manual estimates (82.2 % and 32.1 %). On the other hand, we find that individual-level analyses from the model predictions may be more sensitive to data quality, and thus, some individual scale questions may not be feasible to address if data quality is inconsistent. Manually, we do find that lesion presence in individuals suggests a positive relationship between lesion presence and gregariousness. This work demonstrates that object detection models on photographic data are reasonably successful, highly efficient, and provide initial estimates on the health status of understudied populations of bottlenose dolphins. |
| format | Article |
| id | doaj-art-6010dd5dfb66468bad8672f3812e0f8d |
| institution | Kabale University |
| issn | 1574-9541 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Ecological Informatics |
| spelling | doaj-art-6010dd5dfb66468bad8672f3812e0f8d2025-01-19T06:24:36ZengElsevierEcological Informatics1574-95412025-03-0185102942Automated skin lesion detection and prevalence estimation in Tamanend's bottlenose dolphinsColin J. Murphy0Melissa A. Collier1Ann-Marie Jacoby2Eric M. Patterson3Megan M. Wallen4Janet Mann5Shweta Bansal6Department of Biology, Georgetown University, Washington, DC, USADepartment of Biology, Georgetown University, Washington, DC, USA; Corresponding authors.Marine Science and Conservation Division, Duke University, Beaufort, NC, USADepartment of Biology, Georgetown University, Washington, DC, USA; Office of Protected Resources, NOAA Fisheries, Silver Spring, MD, USADepartment of Biology, Georgetown University, Washington, DC, USA; Protected Resources Division, West Coast Region, NOAA Fisheries, Seattle, WA, USADepartment of Biology, Georgetown University, Washington, DC, USA; Department of Psychology, Georgetown University, Washington, DC, USADepartment of Biology, Georgetown University, Washington, DC, USA; Corresponding authors.Anthropogenic global change is occurring at alarming rates, leading to increased urgency in the ability to monitor wildlife health in real time. Monitoring sentinel marine species, such as bottlenose dolphins, is particularly important due to extensive anthropogenic modifications to their habitats. The most common non-invasive method of monitoring cetacean health is documentation of skin lesions, often associated with poor health or disease, but the current methodology is inefficient and imprecise. Recent advancements in technology, such as machine learning, can provide researchers with more efficient ecological monitoring methods to address health questions at both the population and the individual levels. Our work develops a machine learning model to classify skin lesions on the understudied Tamanend's bottlenose dolphins (Tursiops erebennus) of the Chesapeake Bay, using manual estimates of lesion presence in photographs. We assess the model's performance and find that our best model performs with a high mean average precision (65.6 %–86.8 %), and generally increased accuracy with improved photo quality. We also demonstrate the model's ability to address ecological questions across scales by generating model-based estimates of lesion prevalence and testing the effect of gregariousness on health status. At the population level, our model accurately estimates a prevalence of 72.1 % spot and 27.3 % fringe ring lesions, with a slight underprediction compared to manual estimates (82.2 % and 32.1 %). On the other hand, we find that individual-level analyses from the model predictions may be more sensitive to data quality, and thus, some individual scale questions may not be feasible to address if data quality is inconsistent. Manually, we do find that lesion presence in individuals suggests a positive relationship between lesion presence and gregariousness. This work demonstrates that object detection models on photographic data are reasonably successful, highly efficient, and provide initial estimates on the health status of understudied populations of bottlenose dolphins.http://www.sciencedirect.com/science/article/pii/S1574954124004849Machine learningObject detectionTamanend's bottlenose dolphinSkin lesionCetacean health |
| spellingShingle | Colin J. Murphy Melissa A. Collier Ann-Marie Jacoby Eric M. Patterson Megan M. Wallen Janet Mann Shweta Bansal Automated skin lesion detection and prevalence estimation in Tamanend's bottlenose dolphins Ecological Informatics Machine learning Object detection Tamanend's bottlenose dolphin Skin lesion Cetacean health |
| title | Automated skin lesion detection and prevalence estimation in Tamanend's bottlenose dolphins |
| title_full | Automated skin lesion detection and prevalence estimation in Tamanend's bottlenose dolphins |
| title_fullStr | Automated skin lesion detection and prevalence estimation in Tamanend's bottlenose dolphins |
| title_full_unstemmed | Automated skin lesion detection and prevalence estimation in Tamanend's bottlenose dolphins |
| title_short | Automated skin lesion detection and prevalence estimation in Tamanend's bottlenose dolphins |
| title_sort | automated skin lesion detection and prevalence estimation in tamanend s bottlenose dolphins |
| topic | Machine learning Object detection Tamanend's bottlenose dolphin Skin lesion Cetacean health |
| url | http://www.sciencedirect.com/science/article/pii/S1574954124004849 |
| work_keys_str_mv | AT colinjmurphy automatedskinlesiondetectionandprevalenceestimationintamanendsbottlenosedolphins AT melissaacollier automatedskinlesiondetectionandprevalenceestimationintamanendsbottlenosedolphins AT annmariejacoby automatedskinlesiondetectionandprevalenceestimationintamanendsbottlenosedolphins AT ericmpatterson automatedskinlesiondetectionandprevalenceestimationintamanendsbottlenosedolphins AT meganmwallen automatedskinlesiondetectionandprevalenceestimationintamanendsbottlenosedolphins AT janetmann automatedskinlesiondetectionandprevalenceestimationintamanendsbottlenosedolphins AT shwetabansal automatedskinlesiondetectionandprevalenceestimationintamanendsbottlenosedolphins |