H2020Individual fellowship2022–2025

FISHEARS · Enigmatic fish ears: solving a sensory biology riddle with bioengineering and Artificial Intelligence

Horizon 2020 — Marie Skłodowska-Curie Actions

Duration
2022-06-01 → 2025-05-31
EU contribution
€272,084
Participants
3
Scheme
MSCA-IF

Lines connect the coordinator with its partners.

Results in brief

Enigmatic fish ears: solving a sensory biology riddle with bioengineering and Artificial Intelligence

The project addresses the enigmatic nature of elasmobranch (sharks and rays) hearing structures. Understanding these structures and their unique diversity is crucial due to the complex sensory biology of these species, which is not well-documented or understood. Understanding elasmobranch hearing structures also has significant implications for biodiversity conservation, ecological research, and environmental monitoring. It contributes to the knowledge base necessary for the conservation of these species, which are often threatened by human activities. The project aimed to: - Generate high-resolution 3D models of elasmobranch hearing structures - Develop the first Finite Element Model (FEM) for elasmobranch hearing. - Use artificial intelligence to unravel the factors that shape the hearing systems of elasmobranch fishes. The project has successfully made significant strides toward these goals. High-resolution imaging methodologies were refined, enabling the creation of detailed 3D models of elasmobranch inner ear structures. Over 40 high-resolution 3D images of elasmobranch heads have been produced from preserved specimens, representing a unique and comprehensive collection of inner ear structures for any fish taxa. The segmentation of these scans provides an impressive representation of the structural diversity across more than 10 species of sharks and rays. The FEM approach was initiated, with essential groundwork laid for computational modeling of elasmobranch hearing mechanisms. Furthermore, the application of artificial intelligence has opened pathways to better understand the evolutionary and functional drivers of hearing diversity in these species. Although the project ended earlier than planned due to a career transition, its outputs have already contributed meaningfully to the field. The project resulted in several high-impact publications, with more in preparation, and generated outreach tools, including 3D-printed inner ear models, enhancing public and academic engagement. Overall, the project has had a transformative impact on the researcher’s career, providing advanced technical skills, fostering international collaborations, and securing a tenure-track position. These accomplishments ensure the continuation of this research and its integration into future initiatives, strengthening the foundation for further exploration of elasmobranch sensory biology and its implications for conservation.

Data: CORDIS, © European Union

Project objective

One of the predominant riddles of sensory biology is the diversity in fish auditory systems. It is widelyaccepted that fishes are well adapted to utilising underwater sounds as sensory cues in key life-historyevents. However, the functional significance and the driving force leading to the differences in fishinner ear sizes and structures are unknown. A complex interplay of physical, evolutionary, functionaland ecological factors may shape the different elements: a multiscale environment too complicatedfor human conceptualisation. I propose to address this question by applying novel bioimaging andcomputational tools to investigate elasmobranch fish ears. Firstly, diffusible iodine-based contrast enhancedcomputed tomography (diceCT) will be used, co-registered with MRI data, to build 3D highresolution models of the inner ears. Secondly, a Finite Element (FE) model will be created to digitallyreplicate a fish ear and understand the biomechanics of its structure. Finally, a statistical frameworkwill be developed to incorporate the factors that may shape the hearing system of elasmobranchfishes, including the collected data, together with the available physiological, ecological andbiogeographical information on each species, as well as species’ acoustic environmental parameters. AMachine Learning algorithm will be applied to infer patterns and relationships between the factors, toperform both cluster and prediction analyses. Thus, a reliable model will be developed, which canpredict the hearing capability of any elasmobranch species based on the ear morphology and the firstevidence of the function of fish ear diversity.

Original text from CORDIS.

Participants

Links

Data: CORDIS, © European Union