MUSES9-HS is a full frame, tunable filter-based hyperspectral imager, offering incomparable light throughput (94%) and tuning range (370-1000nm). Within a few dozen of seconds, MUSES9-HS automatically acquires a hypercube image stack comprising up to 126 (extendable to 315) sharply focused, 6.4-megapixel spectral images, collectively containing 6.4 million spectra.

The full-frame, tunable filter operation is clearly advantageous over line scanning or push broom traditional technologies, especially when examining static, two dimensional scenes. This is because bulky mechanical scanners are no longer required, to the benefit of portability, set up simplicity, versatility and direct adaptability to all kinds of lenses, microscopes and telescopes.

SMMART HTS — Automated Trace Evidence Microscopy System

The SMMART HTS is the world’s first high-content screening microscope system purpose-built for forensic trace analysis. Developed by Spectricon and ranked first in a pan-European pre-commercial procurement (PCP) across leading forensic institutes, it has successfully passed site acceptance tests in major European national forensic labs.

What it does

Traditional trace evidence analysis is slow, subjective, and labor-intensive. The SMMART HTS automates this entire process — identifying, enumerating, and documenting millions of microscopic traces collected from crime scenes, in a fully unsupervised manner and at a speed 178× faster than manual examination. It handles the routine workload of a forensics lab so that experts can focus on high-value decisions.
before scan:                                                       after scan:

How it works

The system combines an advanced robotic microscope with a powerful machine learning platform. In a single automated scanning session, it captures a rich, multidimensional imaging dataset across a large sample stage (640 × 420 cm), using four complementary imaging modes:

• Hyperspectral imaging — scans across 340–1100 nm to capture detailed absorption and reflectance spectra at every pixel
• Multiwavelength fluorescence imaging — illuminates samples with three LED/laser lines (365, 405, and 450 nm) to reveal fluorescent traces
• Stokes polarimetry — measures polarization properties at multiple angles and wavelengths, including degree of linear polarization (DoLP)
• Retardance & birefringence imaging — maps optical anisotropy across samples, calibrated over a 0–2500 nm retardation range

Each scan produces a set of 40 multimodal images per field of view in just 5 seconds, generating up to 8 million spectra per spectral cube. The system can cover 3,600 fields of view in approximately 5 hours, entirely without operator intervention.

Intelligent analysis

The integrated software platform processes all acquired data — spectral, fluorescence, polarimetric, and morphological — to generate trace identification maps and enumeration tables automatically. Key capabilities include:

• Supervised and unsupervised machine learning classifiers
• Morphometry-based trace recognition
• Triage of low vs. high probative value traces (e.g. locating traces likely to contain DNA)
• Side-by-side multimodal image review with zoom tools
• Pixel-level spectroscopy with spectra display on mouse hover
• Remote access to data via an encrypted, role-based database

Why it matters

The SMMART HTS turns a process that was previously subjective and fatigue-prone into one that is objective, standardized, and scalable. It is suitable for both routine casework and research applications, is open to customization, and integrates seamlessly into existing forensic lab workflows as a turnkey solution.