Medical THz imaging: an investigation of histo-pathological samples

Universal power-law of terahertz optical properties of borosilicate, tellurite, and chalcogenide glass families

Terahertz (THz) time-domain spectroscopy (TDS, THz-TDS) was used to measure THz optical properties, i.e., refractive indices and absorption coefficients, of borosilicate, tellurite, and chalcogenide glass families. We observe that the THz optical properties depend on glass compositions. THz refractive indices recorded an increasing trend from borosilicate to chalcogenide and to tellurite glass families. Our results demonstrate the ability to select a glass family, system, and composition to target THz optical properties for potential use in THz optical and photonic applications. We report K and β fitting parameters for the power-law model used to describe these properties and show how it can be universally applied to several glass families.

Deep learning for terahertz image denoising in nondestructive historical document analysis

Historical documents contain essential information about the past, including places, people, or events. Many of these valuable cultural artifacts cannot be further examined due to aging or external influences, as they are too fragile to be opened or turned over, so their rich contents remain hidden. Terahertz (THz) imaging is a nondestructive 3D imaging technique that can be used to reveal the hidden contents without damaging the documents. As noise or imaging artifacts are predominantly present in reconstructed images processed by standard THz reconstruction algorithms, this work intends to improve THz image quality with deep learning. To overcome the data scarcity problem in training a supervised deep learning model, an unsupervised deep learning network (CycleGAN) is first applied to generate paired noisy THz images from clean images (clean images are generated by a handwriting generator). With such synthetic noisy-to-clean paired images, a supervised deep learning model using Pix2pixGAN is trained, which is effective to enhance real noisy THz images. After Pix2pixGAN denoising, 99% characters written on one-side of the Xuan paper can be clearly recognized, while 61% characters written on one-side of the standard paper are sufficiently recognized. The average perceptual indices of Pix2pixGAN processed images are 16.83, which is very close to the average perceptual index 16.19 of clean handwriting images. Our work has important value for THz-imaging-based nondestructive historical document analysis.

Linear scanning system for THz imaging

A linear scanning system utilizing constant wave 280 GHz radiation has been developed and characterized. The system comprises a linear array of detectors based on a unique plasma wave approach in terahertz sensing, an impact ionization avalanche transit-time-diode signal generator coupled to a frequency multiplier and an optical system. The performed tests allowed us to estimate the resolution of the system reaching the value of 2.3 mm and to determine the dynamic range of the system to be around 200. The imaging capabilities of the scanner were tested in realistic cases of non-destructive testing and security screening.

Hybrid reflection retrieval method for terahertz dielectric imaging of human bone

Terahertz imaging is becoming a biological imaging modality in its own right, alongside the more mature infrared and X-ray techniques. Nevertheless, extraction of hyperspectral, biometric information of samples is limited by experimental challenges. Terahertz time domain spectroscopy reflection measurements demand highly precise alignment and suffer from limitations of the sample thickness. In this work, a novel hybrid Kramers-Kronig and Fabry-Pérot based algorithm has been developed to overcome these challenges. While its application is demonstrated through dielectric retrieval of glass-backed human bone slices for prospective characterisation of metastatic defects or osteoporosis, the generality of the algorithm offers itself to wider application towards biological materials.

Terahertz image super-resolution based on a complex convolutional neural network

Terahertz (THz) imaging has been applied successfully in numerous applications, from medical imaging to industrial non-destructive detection. However, low resolution has always been a problem due to its long wavelength. A convolution neural network (CNN) is quite effective at improving the resolution of images in optics, in which real numbers are manipulated corresponding to measured intensity. Compared to optics, it is quite feasible to gain both the amplitude and phase information in THz imaging. In this Letter, we have extended the CNN from a real number domain to a complex number domain based on the wave nature of THz light. To the best of our knowledge, this is the first time that such a complex convolution neural network (CCNN) has been shown to be successful in THz imaging. We have proved that resolution can be 0.4 times of the beam size via this approach, and half a wavelength resolution can be obtained easily. Compared to the CNN, the CCNN generates an extra 27.8% increase in terms of contrast, implying a better image. Phase information can be recovered well, which is impossible for the CNN. Although the network is trained by the MNIST dataset, it is quite powerful for image reconstruction. Again, the CCNN outperforms the CNN in terms of generalization capability. We believe such an approach can help to overcome the lower-resolution bottleneck in THz imaging, and it can release the requirement of critical optical components and extensive fine-tuning in systems. THz biomedical imaging, non-destructive testing (NDT), and a lot of imaging applications can benefit from this approach.

Adaptive terahertz image super-resolution with adjustable convolutional neural network

During the real-aperture-scanning imaging process, terahertz (THz) images are often plagued with the problem of low spatial resolution. Therefore, an accommodative super-resolution framework for THz images is proposed. Specifically, the 3D degradation model for the imaging system is firstly proposed by incorporating the focused THz beam distribution, which determines the relationship between the imaging range and the corresponding image restoration level. Secondly, an adjustable CNN is introduced to cope with this range dependent super-resolution problem. By simply tuning an interpolation parameter, the network can be adjusted to produce arbitrary restoration levels between the trained fixed levels without extra training. Finally, by selecting the appropriate interpolation coefficient according to the measured imaging range, each THz image can be coped with its matched network and reach the outstanding super-resolution effect. Both the simulated and real tested data, acquired by a 160 ∼ 220 GHz imager, have been used to demonstrate the superiority of our method.

Two-chip power-scalable THz-generating semiconductor disk laser

We demonstrate a compact two-chip terahertz-emitting vertical-external-cavity surface-emitting laser source, which provides 1 THz output based on intracavity frequency conversion of dual-wavelength emission in a periodically poled lithium niobate crystal. The type-I frequency conversion scheme at room temperature highly benefits from the power-scaling possibilities in a multi-chip cavity with intracavity powers in excess of 500 W.

Line Beam Scanning-Based Ultra-Fast THz Imaging Platform

In order to realize rapid THz detecting and imaging, a line beam scanning-based ultra-fast THz imaging platform is designed combining simple optical components and lightweight mechanical system. The designed THz imaging platform has the resolution of 12 mm, the scanning angle range of ±10.5°, the scanning speed of 0.17 s/frame, and the scanning range of 2 m × 0.8 m; moreover, it can realize rapid human body THz imaging and distinguish metallic objects. Considering its high-quality performance in THz imaging and detecting, it is believed the proposed line beam scanning-based ultra-fast THz imaging platform can be used in the future in various safe screening applications.

Continuous-wave circular polarization terahertz imaging

Biomedical applications of terahertz (THz) radiation are appealing because THz radiation is nonionizing and has the demonstrated ability to detect intrinsic contrasts between cancerous and normal tissue. A linear polarization-sensitive detection technique for tumor margin delineation has already been demonstrated; however, utilization of a circular polarization-sensitive detection technique has yet to be explored at THz frequencies. A reflective, continuous-wave THz imaging system capable of illuminating a target sample at 584 GHz with either linearly or circularly polarized radiation, and capable of collecting both cross- and copolarized signals remitted from the target, is implemented. To demonstrate the system’s utility, a fresh ex vivo human skin tissue specimen containing nonmelanoma skin cancer was imaged. Both polarization-sensitive detection techniques showed contrast between tumor and normal skin tissue, although some differences in images were observed between the two techniques. Our results indicate that further investigation is required to explain the contrast mechanism, as well as to quantify the specificity and sensitivity of the circular polarization-sensitive detection technique.

Development and testing of a single frequency terahertz imaging system for breast cancer detection

The ability to discern malignant from benign tissue in excised human breast specimens in Breast Conservation Surgery (BCS) was evaluated using single frequency terahertz radiation. Terahertz (THz) images of the specimens in reflection mode were obtained by employing a gas laser source and mechanical scanning. The images were correlated with optical histological micrographs of the same specimens, and a mean discrimination of 73% was found for five out of six samples using Receiver Operating Characteristic (ROC) analysis. The system design and characterization is discussed in detail. The initial results are encouraging but further development of the technology and clinical evaluation is needed to evaluate its feasibility in the clinical environment.

Comparing and evaluating the efficacy of the TOR18FG Leeds test X-ray phantom for T-rays

The commercially available X-ray fluoroscopy quality assurance phantom, the Leeds test object TOR18FG, was found to be suitable to assess T-ray image quality in the range (0.1-0.4) THz at a depth of 0.5 cm. Previous to this only custom made phantoms, made especially for the T-ray region, assessed T-ray spatial resolution. However, if sub-wavelength techniques are used, the Leeds test phantom may be implemented to measure the T-ray systems spatial resolution, allowing us to directly compare X-ray and T-ray spatial resolution. The systems compared include a Gulmay Orthovoltage machine (X-ray), the On Board Imager (OBI) of a Varian linear accelerator (X-ray), a two-colour system (T-ray) and Terahertz Time Domain Spectroscopy (THz-TDS) system. X-rays were found to have a spatial resolution of 1.25 lp/mm using the On Board Imager of a Varian Linear Accelerator whilst T-rays imaged using a broadband source imaged through a spatial pinhole had a spatial resolution of 0.56 lp/mm. The TOR18FG background material was found to block, 90% and 99% of the broadband T-rays emitted from a THz-TDS photo-conductive emitter, at 0.4 THz and 0.53 THz respectively. Contrast sensitivity was found to be 3% for 25 cm × 25 cm X-ray field at 65 kV, whilst this value could not be established for T-rays using the TOR18FG. All contrast circles were found to be the same for T-rays i.e. all 40% at 0.1 THz. Images of the same leaf were taken with diagnostic X-rays and both broadband and continuous wave (CW) T-ray systems. T-rays proved superior in providing image contrast, for a hydrated leaf, over X-rays.

Terahertz-wave spectroscopy for precise histopathological imaging of tumor and non-tumor lesions in paraffin sections

Terahertz (THz; 10(12) Hz) waves have a frequency from 0.1 to 10 THz between the visible light and microwave domains. THz waves are expected to be useful for analysis of the histological features, without any staining procedure that is an indispensable prerequisite for optical microscopy. It has been demonstrated that THz transmittances at cancer and normal tissues are different. However, spectroscopy that is currently used is applicable for imaging only small areas at fixed-wavelength. In this study, we have developed a spectrometer employing a gallium phosphide (GaP) THz-generator and applied it to examine large areas of tissue specimens using a wide range of wavelengths. We thus examined the whole areas of two paraffin sections (metastatic liver cancer and acute myocardial infarction) in a frequency range of 1 to 6 THz, and compared the THz images of ordinary paraffin sections with the histological features detected by microscopy. THz imaging showed striking contrasts between cancerous and non-cancerous regions at 3.7 THz. Likewise, the precise imaging was achieved in the infarct myocardium at 3.6 THz. Images of THz transmittances in optimal wavelength were well matched with HE histological features both in cancer and myocardial tissues. Cancer regions showed higher transmittance than non-cancerous regions in liver. Old scar regions showed low transmittance, and necrotic regions showed relatively higher transmittance than normal myocardial areas. Thus, THz imaging precisely reflects tissue conditions such as tumor, non-tumor tissues, tissue degeneration and fibrosis. The newly established THz spectroscopy would be useful for pathological diagnosis of routinely processed specimens.

Sensing Textile Fibers by THz Time-Domain Spectroscopy

Along with the booming development of multi-component blending fabrics, the accurate detection of component of fabrics has become a major goal in textile testing. Terahertz sensing technology provides a new way for detecting the materials. THz time-domain spectroscopy (THz-TDS) is a novel spectroscopic technique which measures the electric field of the radiation through a sample and provides the phase and amplitude changes of the radiation, which can provide information unavailable through conventional methods such as microwave and X-ray techniques. In this investigation, THz-TDS technology was introduced into the textile differentiation. Three kinds of cellulose textile fibers, cotton fiber, bamboo fiber and viscose fiber, were prepared as the sample and detected by THz-TDS at room temperature in the absence of vapor. The temporal and frequency signals of the fibers were obtained. In the THz absorption spectrum, the characteristic absorption peaks of textile fibers in THz wave band were found, which can be used to recognize the fibers. This approach provides a novel non-contact examine method for fiber identification in complicated textiles.

A promising diagnostic method: Terahertz pulsed imaging and spectroscopy

The terahertz band lies between the microwave and infrared regions of the electromagnetic spectrum. This radiation has very low photon energy and thus it does not pose any ionization hazard for biological tissues. It is strongly attenuated by water and very sensitive to water content. Unique absorption spectra due to intermolecular vibrations in this region have been found in different biological materials. These unique features make terahertz imaging very attractive for medical applications in order to provide complimentary information to existing imaging techniques. There has been an increasing interest in terahertz imaging and spectroscopy of biologically related applications within the last few years and more and more terahertz spectra are being reported. This paper introduces terahertz technology and provides a short review of recent advances in terahertz imaging and spectroscopy techniques, and a number of applications such as molecular spectroscopy, tissue characterization and skin imaging are discussed.

Contrast improvement of terahertz images of thin histopathologic sections

We present terahertz images of 10 μm thick histopathologic sections obtained in reflection geometry with a time-domain spectrometer, and demonstrate improved contrast for sections measured in paraffin with water. Automated segmentation is applied to the complex refractive index data to generate clustered terahertz images distinguishing cancer from healthy tissues. The degree of classification of pixels is then evaluated using registered visible microscope images. Principal component analysis and propagation simulations are employed to investigate the origin and the gain of image contrast.

Terahertz pulsed imaging of knee cartilage

Osteoarthritis (OA) is a common form of arthritis caused by cartilage degeneration. In this paper, we investigate the potential use of terahertz (THz) pulsed imaging to quantitatively measure the early symptoms of OA in an animal model. THz images of excised rabbit femoral condyles were taken. We observe THz waves reflected off different layers within samples and demonstrate that the optical delay between reflections can give a quantitative measure of the thicknesses of particular tissues within cartilage.

Technical note: Terahertz imaging of ancient mummies and bone

Ancient mummified soft-tissues are a unique source to study the evolution of disease. Diagnostic imaging of such historic tissues is of foremost interest in paleoanthropology or paleopathology, with conventional x-ray and computed tomography (CT) being the gold-standard. Longer wavelength radiation in the far-infrared or Terahertz region allows diagnostic close-to-surface tissue differentiation of bone morphology while being harmless to human cells. The aim of this study is to show the feasibility and the morpho-diagnostic impact of THz imaging of historic remains. Images of an artificially embalmed ancient Egyptian human mummy hand, an artificially embalmed ancient Egyptian mummified fish and a macerated human lumbar vertebra were obtained by THz-pulse imaging and compared with conventional X-ray and CT images. Although conventional x-ray imaging provides higher spatial resolution, we found that THz-imaging is well-suited for the investigation of ancient mummified soft tissue and embalming-related substances / wrappings. In particular, bone and cartilaginous structures can be well differentiated from surrounding soft-tissues and bandage-wrappings by THz imaging. Furthermore, THz-pulse imaging also measures the time-delay of the pulsed signal when passing through the sample, which provides supplementary information on the optical density of the sample that is not obtained by X-ray and CT. Terahertz radiation provides a completely non-invasive diagnostic imaging modality for historic dry specimens. We anticipate this modality also to be used for detection of hidden objects in historic samples such as funerary amulets still in situ in wrapped mummies, as well as potentially for the identification of spectral signatures from chemical substances, e.g., in embalming essences.

Sensing the hygroscopicity of polymer and copolymer materials using terahertz time-domain spectroscopy

We present the hygroscopicity of polymer and copolymer materials in the low terahertz (THz) frequency range using a linear absorption model. We identify COC 6013 and COC 5013 as optimal THz window materials, possessing both low hygroscopicity and high transmission in the THz regime. The correct choice of window material is of significance for transmission THz spectroscopy and of particular interest for THz liquid spectroscopy.

Medical applications of terahertz imaging: a review of current technology and potential applications in biomedical engineering

Terahertz (THz) imaging is in its early stages of development but already the potential clinical impact of this new imaging modality is clear. From cancer research to DNA analysis THz technology is improving or even making possible imaging of hitherto inaccessible phenomena. In this paper we present a short review of THz imaging from the point of view of biomedical engineering. We discuss the current state of the art in terms of THz imaging systems; describe current applications, future potential and our own approaches to harnessing this novel technology. We draw attention to open problems in the area with respect to the limitations of the technology before concluding with descriptions of our future work in the area.

A comparison of terahertz pulsed imaging with transmission microradiography for depth measurement of enamel demineralisation in vitro

Terahertz pulsed imaging (TPI) is a relatively new, non-ionising and non-destructive imaging technique for studying hard tissues which does not require tooth section preparation, unlike transmission microradiography (TMR). If TPI can measure the depths of caries/demineralisation lesions accurately the same tooth samples could be reused and remeasured during in vitro and in situ studies on de- and/or re-mineralisation. The aim of this study was to compare TPI and TMR for measuring the depths of a range of artificially induced bovine enamel demineralised lesions in vitro. Bovine slabs with artificial caries, induced to different levels of demineralisation by two different but standard demineralisation techniques ('acid gel' and 'carbopol') were measured by TPI and TMR and the readings compared. The set of TPI/TMR measurements obtained on the gel-demineralised slabs showed an extremely high coefficient of determination (r(2) = 0.995). Detailed analysis of the results and theoretical considerations (involving the relationship between refractive index profiling and mineral loss profile) are used to explain the findings and show that for acid gel lesions TPI is measuring demineralisation in the range of 47% of that of TMR depth plus an intercept of 16 microm, with further calculations allowing the TMR depths to be determined to within 5% using TPI.

Terahertz pulsed spectroscopy of human Basal cell carcinoma

Good contrast is seen between normal tissue and regions of tumor in terahertz pulsed imaging of basal cell carcinoma (BCC). To date, the source of contrast at terahertz frequencies is not well understood. In this paper we present results of a spectroscopy study comparing the terahertz properties (absorption coefficient and refractive index) of excised normal human skin and BCC. Both the absorption coefficient and refractive index were higher for skin that contained BCC. The difference was statistically significant over the range 0.2 to 2.0 THz (6.6 cm(-1) to 66.6 cm(-1)) for absorption coefficient and 0.25 to 0.90 THz (8.3 cm(-1) to 30 cm(-1)) for refractive index. The maximum difference for absorption was at 0.5 THz(16.7 cm(-1)). These changes are consistent with higher water content. These results account for the contrast seen in terahertz images of BCC and explain why parameters relating to the reflected terahertz pulse provide information about the lateral spread of the tumor. Knowing the properties of the tissue over the terahertz frequency range will enable the use of mathematical models to improve understanding of the terahertz response of normal and diseased tissue.

Terahertz pulsed imaging of human breast tumors

The feasibility of using terahertz pulsed imaging to map margins of exposed breast tumors was investigated by imaging 22 excised human breast tissue specimens with carcinoma excised from 22 women (mean age, 59 years; range, 39-80 years). The study was approved by the local ethics research committee, and informed consent was obtained from all patients. The size and shape of tumor regions on terahertz images were compared with those identified at histopathologic examination of the imaged section. Two image parameters were investigated: the minimum of the terahertz impulse function and the ratio of the minimum to the maximum of the terahertz impulse function. The correlation coefficient for the tumor area on images compared with that on a photomicrograph of all 22 samples was greater than 0.82 for both parameters. The shape of the tumor regions on terahertz images also correlated well with that on a photomicrograph (median Spearman rank correlation coefficient, 0.69). Findings of this study demonstrate the potential of terahertz pulsed imaging to depict both invasive breast carcinoma and ductal carcinoma in situ under controlled conditions and encourage further studies to determine the sensitivity and specificity of the technique.

Terahertz imaging system based on a backward-wave oscillator

We present an imaging system designed for use in the terahertz range. As the radiation source a backward-wave oscillator was chosen for its special features such as high output power, good wave-front quality, good stability, and wavelength tunability from 520 to 710 GHz. Detection is achieved with a pyroelectric sensor operated at room temperature. The alignment procedure for the optical elements is described, and several methods to reduce the etalon effect that are inherent in monochromatic sources are discussed. The terahertz spot size in the sample plane is 550 microm (nearly the diffraction limit), and the signal-to-noise ratio is 10,000:1; other characteristics were also measured and are presented in detail. A number of preliminary applications are also shown that cover various areas: nondestructive real-time testing for plastic tubes and packaging seals; biological terahertz imaging of fresh, frozen, or freeze-dried samples; paraffin-embedded specimens of cancer tissue; and measurement of the absorption coefficient of water by use of a wedge-shaped cell.

Multispectral classification techniques for terahertz pulsed imaging: an example in histopathology

Terahertz pulsed imaging is a spectroscopic imaging modality using pulses of electromagnetic radiation (100 GHz-10 THz), and there has been recent interest in studying biomedical specimens. It is usual to display parametric images derived from the measured pulses. In this work, classification was achieved by applying multispectral clustering techniques to sets of parametric images. It was hypothesised that adequate information for clustering was carried in a small number of parametric images, providing these were weighted by complementary physical properties. Materials prepared for histopathological examination were chosen because their condition remained stable during long imaging periods and because their dehydrated state led to greater penetration of the radiation. Two specimens were examined in this pilot study, one of basal cell carcinoma and one of melanoma. Unsupervised ISODATA classification using three selected parametric terahertz pulsed images was compared qualitatively with k-means classification using the shape of the whole time series, and with conventional stained microscope slides. There was good qualitative agreement between the classifications. Classifications were consistent with the morphological appearances expected, but further work is required to determine if tumour discrimination is possible. The results have implications for the future development of the technique as the need for only a small number of features could lead to considerably reduced acquisition times.

In vivostudy of human skin using pulsed terahertz radiation

Studies in terahertz (THz) imaging have revealed a significant difference between skin cancer (basal cell carcinoma) and healthy tissue. Since water has strong absorptions at THz frequencies and tumours tend to have different water content from normal tissue, a likely contrast mechanism is variation in water content. Thus, we have previously devised a finite difference time-domain (FDTD) model which is able to closely simulate the interaction of THz radiation with water. In this work we investigate the interaction of THz radiation with normal human skin on the forearm and palm of the hand in vivo. We conduct the first ever systematic in vivo study of the response of THz radiation to normal skin. We take in vivo reflection measurements of normal skin on the forearm and palm of the hand of 20 volunteers. We compare individual examples of THz responses with the mean response for the areas of skin under investigation. Using the in vivo data, we demonstrate that the FDTD model can be applied to biological tissue. In particular, we successfully simulate the interaction of THz radiation with the volar forearm. Understanding the interaction of THz radiation with normal skin will form a step towards developing improved imaging algorithms for diagnostic detection of skin cancer and other tissue disorders using THz radiation.

Optimal discrimination and classification of THz spectra in the wavelet domain

In rapid scan Fourier transform spectrometry, we show that the noise in the wavelet coefficients resulting from the filter bank decomposition of the complex insertion loss function is linearly related to the noise power in the sample interferogram by a noise amplification factor. By maximizing an objective function composed of the power of the wavelet coefficients divided by the noise amplification factor, optimal feature extraction in the wavelet domain is performed. The performance of a classifier based on the output of a filter bank is shown to be considerably better than that of an Euclidean distance classifier in the original spectral domain. An optimization procedure results in a further improvement of the wavelet classifier. The procedure is suitable for enhancing the contrast or classifying spectra acquired by either continuous wave or THz transient spectrometers as well as for increasing the dynamic range of THz imaging systems.