Diaphanography and optical mammography

Principle

Diaphanography or transillumination is a method to detect breast tumors by compressing the breast between glass plates and using red light to shine through the tissue (see Fig. 3a in More Info). Nowadays, diaphanography or better, optical mammography is based on ultra-short (10^-12 regime) laser pulses, which are used as an alternative to traditional X-ray.

Fig. 1 Absorption coefficients (see Lambert-Beer law) of hemoglobin (Hb) (see also Pulse oximetrv),

water and fat (lipid) at concentrations typical for female breasts.

Fig. 2 N-images obtained with 788 nm. These mammograms refer to a 58 year old woman affected by

breast cancer (tumor size: 30mm) in the left breast.

Tissue optical properties

The main absorbers in breasts at visible and near-IR light are hemoglobin, water and fat (see for absorption spectra Fig. 1). The 650-950 nm region is most suited for deep-structure detection, and the penetration depth is a few cm. The scattering is caused by small differences in refractive index (see Snell's Law and Light: refraction) at the microscopic level, mainly of fat droplets and structures inside the cell nucleus. The scattering increases for shorter wavelengths (λ), approximately obeying the λ law of Mie scattering theory (see Light: scattering). The scattering, however, is strong even for λ > 1μm.

The main advantage of using short-pulse lasers for deep-structure is that by using time-resolved detection it is possible to select light, which arrives at the detector at different time slots. The light that arrives early has traveled a shorter and straighter path than the late light. The object of the new generation of optical tomography is to reconstruct an image of both the absorption (direct pathway) and the scattering properties inside the tissue, based on measurements at several light-source and detector position on the skin. Actually, the whole shape of the detected pulses can be of use.

Oxvgen as a key

On the basis of the 3 different wavelengths the kind of tissues can be detected. A tumor wants to grow and therefore induces growth of a lot of little blood vessels around itself, (particularly an invasive one). But vessels growth is delayed and therefore the tumor gets more oxygen out of the blood than normally. So the volume of blood increase, but the oxygenation level of the blood goes down. By detecting both at the same time one can distinguish between benign and malignant tumors.

Application

Diaphanography has been developed to overcome the draw back of missing tumors in classical screening mammography due to a too low resolution. Fig. 2 shows an example of a scan.

Optical mammography is not yet full-grown commercial and in many western countries X-ray screening is still the practice, the more since today doses are very low and seldom causes the metastasizing of an existing cancer. Despite this, nowadays, retrorespective detection rate is about 90% of histological validated malign tumors.

More Info

Fig. 3 a Principle of operation. b Principle of the design of the scanner.

In addition to the use of the whole detected light pulse (see Fig. 3a), the lasers may be chosen at λ's that enhance the contrast between the tumor and surrounding tissue, as in classical spectroscopy, where different compositions of the absorbers give rise to different absorption at various wavelengths. The scattering can also vary between tissue types and wavelengths, which can also be used for contrast enhancement. The choice of wavelengths and knowledge for the biological processes that determine the tissue composition is crucial.

Pulsed laser diodes with so called time-correlated single-photon-counting detection are used to get time-resolved measurements with a sub-ns resolution. The small light sources and detector are placed in a conical configuration, more resembling the conical geometry of conventional tomography (Fig. 3b). The breast is compressed, and the source-detector optical fibers are scanned, for a series of point-wise measurements. Acquiring data for a whole breast with about 1000 points takes a few minutes. Because of this relatively long time, the breast is not compressed as hard as for X-ray mammography.

Finding the best geometry is related to the second problem, development of reconstruction algorithms. For example, should the breast be compressed or not, and it is possible to get better results using fluids to match the optical properties between the skin and the detectors, are some of the questions under investigation.

The principle of the reconstruction algorithm is to divide the problem into two parts, the forward and the inverse problem. The forward problem deals with computation of what the detected signal would be, given that the absorption and scattering are known. The inverse problem is the most likely anatomy which matches best the signals measured with the detectors. Since the problem is highly non-linear, the reconstruction is based on iterations. The solution strongly depends on contrast function that discriminates the tumors. The relative compositions of water, fat and hemoglobin vary not only between tumors and healthy tissue, but also with for example age and hormonal cycles. Furthermore, there is not a single type of breast cancer. Tumors vary a lot in composition and structure. All these parameters have to be understood and quantified.

References

Dehghani H. et al. Multiwavelength Three-Dimensional Near-Infrared Tomography of the Breast: Initial Simulation, Phantom, and Clinical Results, Applied Optics, 2003, 42, 135-145.

Paola Taroni P. et al. Time-resolved optical mammography between 637 and 985 nm: clinical study on the detection and identification of breast lesions Phys. Med. Biol. 2005, 50, 2469-2488.

Rinneberg H. et al. Scanning time-domain optical mammography: detection and characterization of breast tumors in vivo. Technol Cancer Res Treat. 2005, 4, 483-96.