An Automated Fluorescence Microscopy-Based Sensing System for Continuous Detection of Airborne Asbestos Fibers on a PM2.5 Monitoring Platform

Our plain-English summary of An Automated Fluorescence Microscopy-Based Sensing System for Continuous Detection of Airborne Asbestos Fibers on a PM2.5 Monitoring Platform, by A. Kuroda, K. Kaga, T. Nishimura, et al (Sensors, 2026).

The study

• Authors: A. Kuroda, K. Kaga, T. Nishimura, K. Ichikawa, S. Yamazaki, H. Funabashi, T. Ikeda, T. Ishida
• Journal: Sensors
• Published: 2026;26(10)
• Read it: PubMed · Publisher · Free full text

Abstract

Despite regulations restricting asbestos use in many developed countries, asbestos-containing materials (ACMs) persist in aging buildings and can release airborne fibers during demolition and renovation. Therefore, continuous monitoring of airborne asbestos fibers is essential for environmental safety and exposure assessment. Fluorescence microscopy (FM) with fluorescently labeled asbestos-binding proteins offers greater sensitivity and selectivity in detection compared with conventional phase contrast microscopy (PCM). However, its practical application is limited by manual sample preparation and microscopic observations. This study introduces the conceptual design and initial development of an automated FM-based sensing system for monitoring airborne asbestos fibers. The system was constructed by modifying a commercial PM2.5 continuous air sampling platform and integrating automated fluorescent staining, FM imaging, and AI-assisted image analysis for fiber recognition and counting. The system automatically reports airborne asbestos concentrations with corresponding fluorescence images and advances the membrane filter to enable continuous measurements. Performance evaluation using pulverized ACMs showed an overall agreement within 14.1% with PCM-scanning electron microscopy measurements at the group level. Although variability was observed at low fiber concentrations owing to stochastic sampling effects, the results validate the feasibility of automated FM-based sensing for continuous environmental monitoring of airborne asbestos fibers.

In plain English

This was an engineering study. The authors described the design and first laboratory testing of a prototype that automatically collects air, stains asbestos fibers with a fluorescent probe, photographs them by fluorescence microscopy, and counts them with AI software. The prototype was built on a commercial PM2.5 air monitor; a cycle with 20 minutes of sampling took about 30 minutes. In a 400-liter chamber, it was compared with standard phase contrast microscopy (PCM), or PCM adjusted by electron microscopy (PCM–SEM). Test materials were chrysotile, amosite and crocidolite dusts and crushed asbestos-containing boards. For two crushed calcium silicate boards, with images taken at five positions per collection spot by shifting the position by hand, the mean relative difference was −7.3%, with limits of agreement from −99.3% to +84.6%. In low-concentration tests with longer sampling, variability was larger: mean relative difference −14.1%, limits −181.5% to +153.3%. A crushed asbestos-free board produced no fluorescent fibers. The authors concluded that the system suits continuous monitoring and trend detection better than precise single measurements, and that field validation is still needed. One author sits on the board of the company that supplied the probe.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Development of a prototype automated detection device, with a laboratory chamber comparison against manual PCM and PCM–SEM counting using Bland–Altman analysis
• Who was studied: No people were studied. Test materials were standard chrysotile (Zimbabwe), amosite (Transvaal) and crocidolite (Cape) dusts, plus pulverized asbestos-containing boards, dispersed in a 400-L laboratory chamber in Japan.
• Exposure: Airborne fibers from standard chrysotile, amosite and crocidolite dusts. Also pulverized materials: one calcium silicate board with 6.6% amosite, 4.1% crocidolite and 2.8% chrysotile; another with 24% amosite and 0.9% chrysotile; a slate board with 7.4% chrysotile; and an asbestos-free calcium silicate board as a negative control.
• Main results:
  • A measurement cycle with 20 min of sampling at 10 L/min plus about 10 min of staining and detection lasted about 30 minutes.
  • With 20-minute sampling, one detected fiber corresponded to an airborne concentration of 3.0 fibers/L.
  • On its test images, the retrained AI fiber-recognition software achieved accuracy 0.991 ± 0.012, recall 0.902 ± 0.062, precision 0.902 ± 0.111 and F-score 0.902 ± 0.056.
  • Standard asbestos dusts, one imaging position: the mean relative difference between PCM and the automated system was −15.2%, with 95% limits of agreement from −144.2% to +113.9%.
  • Standard asbestos dusts, averaging images from five positions (shifted manually): the mean relative difference was +0.7%, with narrower limits of agreement of −81.3% to +82.6%.
  • Mean relative differences by fiber type were +21.2% for chrysotile, +4.6% for amosite and −19.6% for crocidolite. The authors said the cause of this mineral-dependent difference is unclear.
  • For comparison, two independent PCM counts of the same sample differed by a mean of −22.1%, with limits of agreement from −104.6% to +60.3%.
  • For the pulverized asbestos-free calcium silicate board, the automated system detected no fluorescent fibers.
  • In the asbestos-containing board tests, images were captured at five positions within each collection spot by manually shifting the imaging position.
  • Two pulverized asbestos-containing calcium silicate boards: the mean relative difference between PCM–SEM and the automated system was −7.3%, with limits of agreement from −99.3% to +84.6%.
  • At lower concentrations with longer sampling (30 and 100 min, including slate board samples), the mean relative difference was −14.1%, with limits of agreement from −181.5% to +153.3%. Variability increased as fewer fibers were detected.
• The authors' conclusion: The authors concluded that the prototype showed the feasibility of automated continuous detection. They judged it more suitable for continuous monitoring and trend detection (for example, screening or leakage monitoring) than for precise quantification from a single measurement.
• Limitations the authors state:
  • A formal limit of quantification has not been established because of variability at low concentrations.
  • False positives can occur: when more than 80 fluorescent particles were present per field, the software sometimes counted clusters of particles as fibers.
  • The system is a prototype that needs further optimization, and field validation under real-world conditions is still required.
  • Reagent stability and performance under high or fluctuating temperatures have not yet been examined.

Topics

Medical terms in this study, explained in our Lexicon: Asbestos-Containing Material (ACM), phase contrast microscopy (PCM), Air Sampling, Scanning Electron Microscopy (SEM).

Related studies

• Rapid on-site detection of airborne asbestos fibers and potentially hazardous nanomaterials using fluorescence microscopy-based biosensing — A. Kuroda, M. Alexandrov, et al. (2016)
• Development of an automated asbestos counting software based on fluorescence microscopy — M. Alexandrov, E. Ichida, et al. (2015)
• Live-cell imaging of macrophage phagocytosis of asbestos fibers under fluorescence microscopy — T. Ishida, N. Fujihara, et al. (2019)
• Fluorescent labeling of asbestos fiber for enhanced asbestos detection under fluorescence microscopy — A. Kuroda (2025)
• Evaluation of sensitivity of fluorescence-based asbestos detection by correlative microscopy — T. Ishida, M. Alexandrov, et al. (2012)
• Automated counting of airborne asbestos fibers by a high-throughput microscopy (HTM) method — M.O. Cho, S. Yoon, et al. (2011)

Browse more research in Sensors, or return to medical research.