Navigation and service

Development of image evaluation methods for the detection and classification of particulate and fibrous hazardous substances using methods of machine learning

Project number: F 2468 Institution: Federal Institute for Occupational Safety and Health (BAuA) Status: Ongoing Project Planned end: 2024-09-30

Description:

The legal requirements for chemical safety and occupational safety and health in Europe require the assessment of the health risks of biodurable fibre dusts from nanomaterials and other fibrous materials. This task requires scientifically sound data on the toxicity, dusting propensity and release probability of fibre dusts as well as on exposure of employees.

For biodurable granular dust particles, health risk can be assessed on the basis of the mass inhaled into the deep airways, whereas risks of biodurable fibre dusts are related to the number of respired fibres that exceed a critical length.

The established detection methods for asbestos fibres and industrially produced mineral fibres do not count fibres with diameters below 200 nanometers, since acquisition and evaluation of microscopic images for thinner fibres have so far been too costly for routine measurements. However, toxicological research on nanomaterials over the last decade has shown that fibres can also pose a health risk even when they are much thinner than this diameter limit. In order to be able to determine their number, software-based procedures for image evaluation are therefore urgently needed. These must be able to cope with a high, but in future unavoidable, amount of image data and reliably detect even very thin fibres of critical length.

Previous work of BAuA on automated characterisation of particles has shown that current algorithmic object recognition methods have reached their limits. The project thus aims at developing a new approach using machine learning methods.

Contact

Unit 4.5 "Particulate Hazardous Substances, Advanced Materials"

Phone: +49 231 9071-1971 Fax: +49 231 9071-2070

Cookies help us to provide our services. By using our website you agree that we can use cookies. Read more about our Privacy Policy and visit the following link: Privacy Policy

OK