An official website of the United States government
Accurate detection of skin lesions through computer-aided diagnosis has emerged as a critical advancement in dermatology, addressing the inefficiencies and errors inherent in manual visual analysis. Despite the promise of automated diagnostic approaches, challenges such as image size variability, hair artifacts, color inconsistencies, ruler markers, low contrast, lesion dimension differences, and gel bubbles must be overcome. Researchers have made significant strides in binary classification problems, particularly in distinguishing melanocytic lesions from normal skin conditions. Leveraging the “MNIST HAM10000” dataset from the International Skin Image Collaboration, this study integrates Scale-Invariant Feature Transform (SIFT) features with a custom convolutional neural network model called LesionNet. The experimental results reveal the model's robustness, achieving an impressive accuracy of 99.28%. This high accuracy underscores the effectiveness of combining feature extraction techniques with advanced neural network models in enhancing the precision of skin lesion detection.
more »
« less
This content will become publicly available on March 1, 2026
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. 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.
more »
« less
- Award ID(s):
- 2146995
- PAR ID:
- 10562502
- Publisher / Repository:
- Elsevier Science Direct
- Date Published:
- Journal Name:
- Ecological Informatics
- Volume:
- 85
- Issue:
- C
- ISSN:
- 1574-9541
- Page Range / eLocation ID:
- 102942
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Assessing trends in population abundance and demographics is crucial for managing long‐lived and slow‐reproducing species. Obtaining demographic data, and age‐structure information, is challenging, notably for cetaceans. To address this, we combined Unoccupied Aerial System (UAS; drone) photogrammetry data with long‐term (>20 years) photo identification data to assess the age‐structure of the critically endangered sub‐population of common bottlenose dolphins (Tursiops truncatus) of the Gulf of Ambracia, Greece. We compared our findings with two extensively studied non‐endangered bottlenose dolphin populations (T. aduncusin Shark Bay, Australia, andT. truncatusin Sarasota Bay, USA). Using a log‐linear model, we estimated the total body lengths (TL) of 160 known‐aged dolphins between 2021 and 2023 from blowhole‐to‐dorsal‐fin distance (BHDF) measurements collected during surfacing. Subsequently, we tested four growth models to establish an age‐length growth curve. We assessed the sub‐population's age‐structure using three methods: (1) UAS‐derived TL estimates, (2) age‐length growth curve and (3) long‐term monitoring data (i.e. actual age‐structure). UAS‐measured TL (247.6 ± 32.2 cm) and UAS‐estimated TL (246.0 ± 34.7 cm) of the Greek sub‐population showed no differences. The Richards Growth model suggested an asymptotic length of 258.5 cm. In Greece, resulting age‐structure estimates across the three methods revealed no significant differences (P > 0.1). The Gulf of Ambracia and Shark Bay populations shared similar age‐structures, while Sarasota had higher proportions of 2–10 year‐olds and lower proportions of 10+ year‐olds. All populations had a comparable proportion of 0–2 year‐olds (~14%), indicating a similar reproductive rate. Our findings suggest stability in the Greek sub‐population; however, additional monitoring of reproductive parameters is essential before concluding its status. We demonstrated the effectiveness of UAS‐photogrammetry in rapidly quantifying population age‐structure, including scenarios with limited or no demographic data. This technique shows promise for enhancing precision, timeliness, cost‐effectiveness and efficiency in population monitoring and informing timely conservation management decisions.more » « less
-
ABSTRACT Investigating the foraging ecology and trophic interactions of threatened marine predators is critical to assess how community changes due to anthropogenic activities will affect predator–prey relationships. Two species of threatened coastal dolphins, the Indian Ocean humpback dolphin (Sousa plumbea) and the Indo‐Pacific bottlenose dolphin (Tursiops aduncus), occur off Nosy Be, north‐western Madagascar, in a region where artisanal fisheries are ecologically and socioeconomically important. Here, we investigated the feeding ecology of these two coastal dolphins and their trophic interactions with four other odontocetes using bulk stable carbon and nitrogen isotope analysis (δ13C andδ15N). Humpback dolphins had significantly enrichedδ13C values, reflecting a preference for coastal/benthic prey. Bottlenose dolphins had a broader isotopic niche, suggesting a broader range of prey and foraging habitats. The overlap in isotopic niche of all six odontocete species was limited, indicating partitioning of resources and habitats. Bayesian mass‐balance isotopic mixing models revealed that humpback dolphins forage primarily on reef planktivores (38.9%) and inner reef mesopredators (20.5%), while bottlenose dolphins had a broader diet, including reef‐associated (15%–32%) and pelagic prey (12%–23%). Our study reveals that the reliance on inshore prey by humpback dolphins may place them in competition with coastal fisheries.more » « less
-
In 2019, an unusual mortality event (UME) affected bottlenose dolphins (Tursiops truncatus) in the Mississippi Sound (MSS) following an extended dual opening of the Bonnet Carré Spillway (BCS), which prevents flooding in New Orleans. This resulted in low salinity, skin lesions, and electrolyte imbalances in dolphins. Additionally, the influx likely altered the MSS’s environmental chemical composition, including levels of heavy metals and metalloids; thus, we quantified heavy metals, metalloids, and essential elements in the tissues of dolphins that stranded in the MSS before and after the 2019 UME. We hypothesized that levels of heavy metals and metalloids (such as mercury (Hg), arsenic (As), lead (Pb), and cadmium (Cd)) would not show significant changes post-UME. Indeed, we found no major changes associated with the UME in most metals; sodium (Na) and magnesium (Mg) levels were lower in several tissues after 2019, which correlated with the average yearly salinity measured from the MSS. Toxic metals and metalloids were detectable with some changes over time; however, the selenium (Se):Hg molar ratio increased in some tissues post-2019. Additionally, we confirmed that Hg can bioaccumulate, with positive correlations between Hg levels and dolphin size as assessed by straight length. Overall, our findings indicate that freshwater incursions into the MSS can alter dolphin exposure to essential and toxic elements.more » « less
-
Abstract Long‐term studies often rely on natural markings for individual identification across time. The primary method for identification in small cetaceans relies on dorsal fin shape, scars, and other natural markings. However, dorsal fin markings can vary substantially over time and the dorsal fin can become unrecognizable after an encounter with a boat or shark. Although dorsal fins have the advantage in that they always break the water surface when the cetacean breathes, other physical features, such as body scars and pigmentation patterns can supplement. The goal of this study was to explore the use of dorso‐lateral pigment patterns to identify wild bottlenose dolphins. We employed photographic pigment matching tests to determine if pigmentation patterns showed (1) longitudinal consistency and (2) bilateral symmetry using a 30 yr photographic database of bottlenose dolphins (Tursiops aduncus). We compared experienced dolphin researchers and inexperienced undergraduate student subjects in their ability to accurately match images. Both experienced and inexperienced subjects correctly matched dolphin individuals at a rate significantly above chance, even though they only had 10 s to make the match. These results demonstrate that pigment patterns can be used to reliably identify individual wild bottlenose dolphins, and likely other small cetacean species at other sites.more » « less
