Hyperspectral Camera FigSpec FS-22 Assists Research Teams in Achieving Rapid and Accurate Detection of Wound Bacteria

A scientific research achievement that has made significant progress in the field of rapid detection of wound infection bacteria has been officially published. The study was jointly completed by Chongqing University of Posts and Telecommunications, Daping Hospital of the Army Medical University, and other institutions. By combining fluorescence hyperspectral imaging technology with deep learning algorithms, the research achieved non-invasive, rapid classification and identification of multiple common wound pathogens. During the data acquisition stage, the research team used the FigSpec FS-22 hyperspectral imaging camera from CHNSpec, which provided critical spectral image data support for the experiments and demonstrated the device’s application potential in precision bio-optical detection.

Timely diagnosis of wound bacterial infections is of great importance for clinical treatment. However, traditional methods such as bacterial culture and PCR are often time-consuming and require invasive sampling. Therefore, developing a technology capable of rapidly and non-invasively identifying bacteria has become an urgent need. Hyperspectral imaging technology can simultaneously obtain spatial information and continuous spectral information of a target, while fluorescence hyperspectral imaging goes a step further by inducing the sample to emit fluorescence through excitation at specific wavelengths, thereby enhancing the ability to detect differences in internal chemical substances of microorganisms. This study leveraged this principle to systematically collect and analyze the spectral characteristics of eight common wound infection bacteria.

In the experiments, the CHNSpec FigSpec FS-22 hyperspectral camera played a crucial role. The system features a spectral detection range of 400–1000 nm and a high spatial resolution of 1920 × 1920, enabling it to finely capture the autofluorescence signals generated by bacteria under 405 nm laser excitation. The research team constructed a large fluorescence hyperspectral dataset covering different bacterial species, concentrations, and growth times, totaling 25,600 samples. Faced with the challenges of high dimensionality, large information volume, and subtle spectral differences among bacteria in hyperspectral data, the researchers independently designed a deep learning model called the “Spatial–Spectral Multi-Scale Attention Network.” This model can effectively focus on bacterial colony regions, suppress background interference such as culture substrates, and deeply extract discriminative features from the spectral dimension, thereby enabling collaborative identification of bacterial species, growth states, and even concentrations.

The research results show that this method achieved a bacterial classification accuracy of 98.52% under different growth conditions, a species-level identification accuracy of 98.71%, and maintained effective detection even at bacterial concentrations as low as 10⁴ CFU/mL. Compared with multiple existing algorithm models evaluated in the study, this network trained on hyperspectral data from CHNSpec demonstrated reliable performance. These results verify the feasibility of combining fluorescence hyperspectral imaging with advanced algorithms in the field of rapid microorganism detection and provide important technical references for the future development of instant diagnostic devices applicable to clinical environments.

Although the study was conducted under controlled laboratory conditions using purified bacterial strains, its technical pathway clearly demonstrates the value of hyperspectral imaging in biomedical detection. The FigSpec FS-22 hyperspectral camera from CHNSpec, with its stable imaging performance and rich spectral information acquisition capabilities, provides a solid hardware foundation for such cutting-edge exploration. Looking ahead, with further optimization of algorithms and deeper research into clinical applicability, this technology solution integrating advanced imaging and intelligent analysis is expected to move closer to the goal of truly realizing real-time, non-invasive, and precise diagnosis of wound infections, offering new tool options for clinical infection prevention and control.

Product Recommendation: FigSpec FS-22 Hyperspectral Imaging Camera

Product Features

1. Visible / Near-Infrared:

• Spectral range: 400–1000 nm, wavelength resolution better than 2.5 nm, up to 1200 spectral channels
• Image resolution up to 2048 × 2048

2. Short-Wave Infrared:

• Spectral range: 900–1700 nm, wavelength resolution better than 6 nm, up to 1024 spectral channels
• Image resolution up to 1280 × 1280