Dolphin Health Classifications from Whistle Features

Bottlenose dolphins often conceal behavioral signs of illness until they reach an advanced stage. Motivated by the efficacy of vocal biomarkers in human health diagnostics, we utilized supervised machine learning methods to assess various model architectures’ effectiveness in classifying dolphin hea...

Full description

Saved in:
Bibliographic Details
Main Authors: Brittany Jones, Jessica Sportelli, Jeremy Karnowski, Abby McClain, David Cardoso, Maximilian Du
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Journal of Marine Science and Engineering
Subjects:
Online Access:https://www.mdpi.com/2077-1312/12/12/2158
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850241838007451648
author Brittany Jones
Jessica Sportelli
Jeremy Karnowski
Abby McClain
David Cardoso
Maximilian Du
author_facet Brittany Jones
Jessica Sportelli
Jeremy Karnowski
Abby McClain
David Cardoso
Maximilian Du
author_sort Brittany Jones
collection DOAJ
description Bottlenose dolphins often conceal behavioral signs of illness until they reach an advanced stage. Motivated by the efficacy of vocal biomarkers in human health diagnostics, we utilized supervised machine learning methods to assess various model architectures’ effectiveness in classifying dolphin health status from the acoustic features of their whistles. A gradient boosting classifier achieved a 72.3% accuracy in distinguishing between normal and abnormal health states—a significant improvement over chance (permutation test; 1000 iterations, <i>p</i> < 0.001). The model was trained on 30,693 whistles from 15 dolphins and the test set (15%) totaled 3612 ‘normal’ and 1775 ‘abnormal’ whistles. The classifier identified the health status of the dolphin from the whistles features with 72.3% accuracy, 73.2% recall, 56.1% precision, and a 63.5% F1 score. These findings suggest the encoding of internal health information within dolphin whistle features, with indications that the severity of illness correlates with classification accuracy, notably in its success for identifying ‘critical’ cases (94.2%). The successful development of this diagnostic tool holds promise for furnishing a passive, non-invasive, and cost-effective means for early disease detection in bottlenose dolphins.
format Article
id doaj-art-de681ff34a32438e9c97e0f29cc5740d
institution OA Journals
issn 2077-1312
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Journal of Marine Science and Engineering
spelling doaj-art-de681ff34a32438e9c97e0f29cc5740d2025-08-20T02:00:28ZengMDPI AGJournal of Marine Science and Engineering2077-13122024-11-011212215810.3390/jmse12122158Dolphin Health Classifications from Whistle FeaturesBrittany Jones0Jessica Sportelli1Jeremy Karnowski2Abby McClain3David Cardoso4Maximilian Du5Naval Information Warfare Center Pacific, 53560 Hull Street, San Diego, CA 92152, USANational Marine Mammal Foundation: 3131, 2240 Shelter Island Dr, San Diego, CA 92106, USAUniversity of California San Diego, 9500 Gilman Dr, La Jolla, CA 92093, USANaval Information Warfare Center Pacific, 53560 Hull Street, San Diego, CA 92152, USASan Diego State University, 5500 Campanile Drive, San Diego, CA 92182, USAStanford University, 450 Jane Stanford Way, Stanford, CA 94305, USABottlenose dolphins often conceal behavioral signs of illness until they reach an advanced stage. Motivated by the efficacy of vocal biomarkers in human health diagnostics, we utilized supervised machine learning methods to assess various model architectures’ effectiveness in classifying dolphin health status from the acoustic features of their whistles. A gradient boosting classifier achieved a 72.3% accuracy in distinguishing between normal and abnormal health states—a significant improvement over chance (permutation test; 1000 iterations, <i>p</i> < 0.001). The model was trained on 30,693 whistles from 15 dolphins and the test set (15%) totaled 3612 ‘normal’ and 1775 ‘abnormal’ whistles. The classifier identified the health status of the dolphin from the whistles features with 72.3% accuracy, 73.2% recall, 56.1% precision, and a 63.5% F1 score. These findings suggest the encoding of internal health information within dolphin whistle features, with indications that the severity of illness correlates with classification accuracy, notably in its success for identifying ‘critical’ cases (94.2%). The successful development of this diagnostic tool holds promise for furnishing a passive, non-invasive, and cost-effective means for early disease detection in bottlenose dolphins.https://www.mdpi.com/2077-1312/12/12/2158bioacousticsmachine learninghealth classifierdolphin whistles
spellingShingle Brittany Jones
Jessica Sportelli
Jeremy Karnowski
Abby McClain
David Cardoso
Maximilian Du
Dolphin Health Classifications from Whistle Features
Journal of Marine Science and Engineering
bioacoustics
machine learning
health classifier
dolphin whistles
title Dolphin Health Classifications from Whistle Features
title_full Dolphin Health Classifications from Whistle Features
title_fullStr Dolphin Health Classifications from Whistle Features
title_full_unstemmed Dolphin Health Classifications from Whistle Features
title_short Dolphin Health Classifications from Whistle Features
title_sort dolphin health classifications from whistle features
topic bioacoustics
machine learning
health classifier
dolphin whistles
url https://www.mdpi.com/2077-1312/12/12/2158
work_keys_str_mv AT brittanyjones dolphinhealthclassificationsfromwhistlefeatures
AT jessicasportelli dolphinhealthclassificationsfromwhistlefeatures
AT jeremykarnowski dolphinhealthclassificationsfromwhistlefeatures
AT abbymcclain dolphinhealthclassificationsfromwhistlefeatures
AT davidcardoso dolphinhealthclassificationsfromwhistlefeatures
AT maximiliandu dolphinhealthclassificationsfromwhistlefeatures