Cascade laser imaging for security and medicine

A handheld mid-IR laser imaging platform detects trace materials quickly and with high sensitivity for explosives detection and medical diagnosis.
20 May 2014
Anadi Mukherjee

The chemical, security, and medical industries increasingly demand sensitive, highly specific, and fast molecular recognition at a distance from the target. Methods such as Raman and near-IR spectroscopy fulfill some of these requirements, but they have yet to quantify trace molecules at videographic speeds. We have now developed a technology that offers higher sensitivity and faster results for use in sectors such as petroleum, defense, healthcare, medical diagnosis, and drug discovery.

Purchase SPIE Field Guide to Interferometric Optical TestingOur approach uses mid-IR laser image-based non-invasive molecular recognition, operating at various distances. We use sensors functioning mainly in the mid-IR wavelength region (typically 3–12 microns in length) because all molecules have strongest vibration absorption bands in this range, enabling us to quantify trace molecules quickly, with very high sensitivity and specificity. We have exploited the recent availability of compact semiconductor mid-IR (cascade) lasers, highly sensitive long-wave uncooled mid-IR cameras, and fast hyperspectral image processing in our first-generation product, the handheld Sensitive Picture of Traces (SPoT) (see Figure 1).

Figure 1. The Sensitive Picture of Traces (SPoT), a cascade laser hyperspectral image-based solid/liquid trace sensor.

For this technology, we developed passive image sensors, which do not need mid-IR light sources such as lasers or LEDs. Instead, commercially available cascade lasers (quantum and interband) produce enough power to flood a scene hundreds of meters away. They use a long-wavelength IR camera to detect reflection from the scene and conduct spectrum analysis to identify the material. An imaging device then displays the target concentration in a false-color map. For example, supposing there are trace elements of TNT on a car door handle, the handheld sensor SPoT can detect the trace just like a spectroscopic video camera. The lasers illuminate and image the scene at the wavelength the hazardous material trace absorbs. On illumination by beams, there are dark patches on the traces. The regions of TNT residue absorb more of this light (which is tuned to its absorption) than surrounding areas with no residue. Our first SPoT model can image four explosive traces and one binder over large areas using a multiplexed set of three quantum cascade lasers. Multiplexing a set of algorithmically chosen lasers enables trace detection from established lists of hazardous materials in known backgrounds.

Using our approach for medical applications, we can tune a laser to the absorption of an early-stage cancer signature in the mid-IR spectrum of a tissue. Current methods of characterizing cancerous tissues include Fourier transform IR (FTIR) microscopy, which is advantageous because it requires no staining of the target tissues with a fluorescent molecule or antibodies for color contrast imaging. However, FTIR is time-consuming and has limited sensitivity. For material analysis it uses an incoherent broadband heat source with emission typically across 700–4000cm−1. Although the method is convenient for analyzing the entire vibrational fingerprint region of a material, the heat source is far less sensitive than a laser equivalent, which has many orders of magnitude and more photons or frequency channels. Furthermore, although widely tunable quantum cascade laser sources are currently available (in the range 5–14mm), they do not cover the full spectral range of FTIR, are prohibitively expensive, and have insufficient output power for demanding image-based detection applications. Our approach, multiplexing Fabry-Pérot and distributed feedback cascade lasers, offers improved domain-specific operation. For medical imaging, our method has the advantage of being quick to execute, requires no staining, and is more sensitive, and it could improve on the slow detection speeds in Raman, near-IR, and FTIR methods. We recently demonstrated our technique in dry biopsy tissues, and achieved quantitative imaging of proteins in blood serum with record high sensitivity and specificity.1

In future, the combination of cost-effective single-wavelength lasers, long-wavelength IR cameras, and our own proprietary optoelectronic packaging could perform the roles of cancer diagnosis and surgery at the same time, saving lives, time, and money. Our next plan is to use a fiber optic probe for in vivo quantitative imaging at the site of a live tumor, to determine the level of malignancy and enable laser surgery immediately at the diagnosed spot. This will allow biopsy-less diagnosis at the live tumor as well as micron-precision laser surgery, removing only the malignant part of the tissue and no more.

Anadi Mukherjee
Fremont, CA

Anadi Mukherjee has more than 30 years of research and development leadership in laser-matter interaction. He was previously a research assistant professor at the Center for High Technology Materials at the University of New Mexico, and at Pranalytica Inc. he developed single-mode tunable external cavity quantum cascade lasers and multiplexed them for multigas detection. He has published more than 50 papers in peer-reviewed journals.

1. A. Mukherjee, Q. Bylund, M. Prasanna, Y. Margalit, T. Tihan, Spectroscopic imaging of serum proteins using quantum cascade lasers, J. Biomed. Opt. 18(3), p. 036011, 2013. doi:10.1117/1.JBO.18.3.036011
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?