Imaging depth variations in hyperspectral imaging: Development of a method to detect tumor up to the required tumor‐free margin width

Hyperspectral imaging is a promising technique for resection margin assessment during cancer surgery. Thereby, only a specific amount of the tissue below the resection surface, the clinically defined margin width, should be assessed. Since the imaging depth of hyperspectral imaging varies with wavelength and tissue composition, this can have consequences for the clinical use of hyperspectral imaging as margin assessment technique. In this study, a method was developed that allows for hyperspectral analysis of resection margins in breast cancer. This method uses the spectral slope of the diffuse reflectance spectrum at wavelength regions where the imaging depth in tumor and healthy tissue is equal. Thereby, tumor can be discriminated from healthy breast tissue while imaging up to a similar depth as the required tumor‐free margin width of 2 mm. Applying this method to hyperspectral images acquired during surgery would allow for robust margin assessment of resected specimens. In this paper, we focused on breast cancer, but the same approach can be applied to develop a method for other types of cancer.     

Simplifying the assessment of human breast cancer by mapping a micro‐scale heterogeneity index in optical coherence elastography

Surgical treatment of breast cancer aims to identify and remove all malignant tissue. Intraoperative assessment of tumor margins is, however, not exact; thus, re‐excision is frequently needed, or excess normal tissue is removed. Imaging methods applicable intraoperatively could help to reduce re‐excision rates whilst minimizing removal of excess healthy tissue. Optical coherence elastography (OCE) has been proposed for use in breast‐conserving surgery; however, intraoperative interpretation of complex OCE images may prove challenging. Observations of breast cancer on multiple length scales, by OCE, ultrasound elastography, and atomic force microscopy, have shown an increase in the mechanical heterogeneity of malignant breast tumors compared to normal breast tissue. In this study, a micro‐scale mechanical heterogeneity index is introduced and used to form heterogeneity maps from OCE scans of 10 ex vivo human breast tissue samples. Through comparison of OCE, optical coherence tomography images, and corresponding histology, malignant tissue is shown to possess a higher heterogeneity index than benign tissue. The heterogeneity map simplifies the contrast between tumor and normal stroma in breast tissue, facilitating the rapid identification of possible areas of malignancy, which is an important step towards intraoperative margin assessment using OCE.

     

Slide‐free imaging of hematoxylin‐eosin stained whole‐mount tissues using combined third‐harmonic generation and three‐photon fluorescence microscopy

Intraoperative margin assessment of surgical tissues during cancer surgery is clinically important, especially in the case of tissue conserving surgery like Mohs micrographic surgery in which minimization of the surgical area is considered crucial. Frozen pathology is the gold standard of assessing excised tissues for signs of remaining cancerous lesions. The current protocol, however, is time‐consuming and labor‐intensive. Instead of the complex frozen sectioning, staining, and traditional white light microscopy imaging protocol, optically sectioned histopathological imaging of hematoxylin‐eosin stained whole‐mount skin tissues with a subfemtoliter resolution is demonstrated by using nonlinear microscopy in this study. With our proposed method, the reagents of staining and the contrast of imaging are fully consistent with the current clinical standard of frozen pathology, thus facilitating rapid intraoperative assessment of surgical tissues for future applications. Image: Slide‐free nonlinear microscopy imaging of H&E stained whole‐mount skin tissue showing the morphology of sweat glands.      

Portable optical fiber probe-based spectroscopic scanner for rapid cancer diagnosis: a new tool for intraoperative margin assessment

There continues to be a significant clinical need for rapid and reliable intraoperative margin assessment during cancer surgery. Here we describe a portable, quantitative, optical fiber probe-based, spectroscopic tissue scanner designed for intraoperative diagnostic imaging of surgical margins, which we tested in a proof of concept study in human tissue for breast cancer diagnosis. The tissue scanner combines both diffuse reflectance spectroscopy (DRS) and intrinsic fluorescence spectroscopy (IFS), and has hyperspectral imaging capability, acquiring full DRS and IFS spectra for each scanned image pixel. Modeling of the DRS and IFS spectra yields quantitative parameters that reflect the metabolic, biochemical and morphological state of tissue, which are translated into disease diagnosis. The tissue scanner has high spatial resolution (0.25 mm) over a wide field of view (10 cm × 10 cm), and both high spectral resolution (2 nm) and high spectral contrast, readily distinguishing tissues with widely varying optical properties (bone, skeletal muscle, fat and connective tissue). Tissue-simulating phantom experiments confirm that the tissue scanner can quantitatively measure spectral parameters, such as hemoglobin concentration, in a physiologically relevant range with a high degree of accuracy (<5% error). Finally, studies using human breast tissues showed that the tissue scanner can detect small foci of breast cancer in a background of normal breast tissue. This tissue scanner is simpler in design, images a larger field of view at higher resolution and provides a more physically meaningful tissue diagnosis than other spectroscopic imaging systems currently reported in literatures. We believe this spectroscopic tissue scanner can provide real-time, comprehensive diagnostic imaging of surgical margins in excised tissues, overcoming the sampling limitation in current histopathology margin assessment. As such it is a significant step in the development of a platform technology for intraoperative management of cancer, a clinical problem that has been inadequately addressed to date.     

Partial nephrectomy margin imaging using structured illumination microscopy

Partial nephrectomy (PN) is the recommended procedure over radical nephrectomy (RN) for patients with renal masses less than 4 cm in diameter (Stage T1a). Patients with less than 4 cm renal masses can also be treated with PN, but have a higher risk for positive surgical margins (PSM). PSM, when present, are indicative of poor clinical outcomes. The current gold‐standard histopathology method is not well‐suited for the identification of PSM intraoperatively due to processing time and destructive nature. Here, video‐rate structured illumination microscopy (VR‐SIM) was investigated as a potential tool for PSM detection during PN. A clinical image atlas assembled from ex vivo renal biopsies provided diagnostically useful images of benign and malignant kidney, similar to permanent histopathology. VR‐SIM was then used to image entire parenchymal margins of tumor resection covering up to >1800× more margin surface area than standard histology. Aided by the image atlas, the study pathologist correctly classified all parenchymal margins as negative for PSM with VR‐SIM, compared to standard postoperative pathology. The ability to evaluate large surgical margins in a short time frame with VR‐SIM may allow it to be used intraoperatively as a “safety net” for PSM detection, allowing more patients to undergo PN over RN.      

Three‐dimensional static optical coherence elastography based on inverse compositional Gauss‐Newton digital volume correlation

The three‐dimensional (3D) mechanical properties characterization of tissue is essential for physiological and pathological studies, as biological tissue is mostly heterogeneous and anisotropic. A digital volume correlation (DVC)‐based 3D optical coherence elastography (OCE) method is developed to measure the 3D displacement and strain tensors. The DVC algorithm includes a zero‐mean normalized cross‐correlation criterion‐based coarse search regime, an inverse compositional Gauss‐Newton fine search algorithm and a local ternary quadratic polynomial fitting strain calculation method. A 3D optical coherence tomography (OCT) scanning protocol is proposed through theoretical analysis and experimental verification. Measurement errors of the DVC‐based 3D OCE method are evaluated to be less than 2.0 μm for displacements and 0.30% for strains by rigid body motion experiments. The 3D displacements and strains of a phantom and a specimen of chicken breast tissue under compression are measured. Results of the phantom show a good agreement with theoretical analysis and tensile testing. The strains of the chicken breast tissue indicate anisotropic biomechanical properties. This study provides an effective method for 3D biomechanical property studies of soft tissue and improves the development of 3D OCE techniques.     

Concentric photothermal coagulation with basket‐integrated optical device for treatment of tracheal stenosis

A basket‐integrated optical device is developed to consistently treat tubular tissue by centering an optical diffuser in the lumen. Four nitinol arms in conjunction with the optical diffusing applicator are deployed to induce homogeneous circumferential light emission and concentric photothermal coagulation on tracheal tissue. A 1470‐nm laser light is employed for the tissue testing at various irradiation conditions and evaluated in terms of thermal gradient and temperature evolution. Preliminary experiments on liver tissue demonstrate the concentric development of the radial thermal coagulation in the tissue (eccentric ratio = ~5.5%). The interstitial tissue temperature increases with the total amount of energy delivery (around 65°C). Ex vivo trachea testing yields up to 16.5% tissue shrinkage due to dehydration as well as uniform ablation of the cilia and goblet cells in a mucosa layer under 7‐W irradiation for 10 s. The proposed optical device may be a feasible therapeutic method to entail the circumferential coagulation in the tubular tissues in a reliable manner.      

General model for depth‐resolved estimation of the optical attenuation coefficients in optical coherence tomography

We present the proof of concept of a general model that uses the tissue sample transmittance as input to estimate the depth‐resolved attenuation coefficient of tissue samples using optical coherence tomography (OCT). This method allows us to obtain an image of tissue optical properties instead of intensity contrast, guiding diagnosis and tissues differentiation, extending its application from thick to thin samples. The performance of our method was simulated and tested with the assistance of a home built single‐layered and multilayered phantoms (~100 μm each layer) with known attenuation coefficient on the range of 0.9 to 2.32 mm^?1. It is shown that the estimated depth‐resolved attenuation coefficient recovers the reference values, measured by using an integrating sphere followed by the inverse adding doubling processing technique. That was corroborated for all situations when the correct transmittance value is used with an average difference of 7%. Finally, we applied the proposed method to estimate the depth‐resolved attenuation coefficient for a thin biological sample, demonstrating the ability of our method on real OCT images.     

Second and third harmonic generation microscopy visualizes key structural components in fresh unprocessed healthy human breast tissue

Real‐time assessment of excised tissue may help to improve surgical results in breast tumor surgeries. Here, as a step towards this purpose, the potential of second and third harmonic generation (SHG, THG) microscopy is explored. SHG and THG are nonlinear optical microscopic techniques that do not require labeling of tissue to generate 3D images with intrinsic depth‐sectioning at sub‐cellular resolution. Until now, this technique had been applied on fixated breast tissue or to visualize the stroma only, whereas most tumors start in the lobules and ducts. Here, SHG/THG images of freshly excised unprocessed healthy human tissue are shown to reveal key breast components—lobules, ducts, fat tissue, connective tissue and blood vessels, in good agreement with hematoxylin and eosin histology. DNA staining of fresh unprocessed mouse breast tissue was performed to aid in the identification of cell nuclei in label‐free THG images. Furthermore, 2‐ and 3‐photon excited auto‐fluorescence images of mouse and human tissue are collected for comparison. The SHG/THG imaging modalities generate high quality images of freshly excised tissue in less than a minute with an information content comparable to that of the gold standard, histopathology. Therefore, SHG/THG microscopy is a promising tool for real‐time assessment of excised tissue during surgery.      

Label‐free imaging and spectroscopy for early detection of cervical cancer

The label‐free imaging and spectroscopy method was studied on cervical unstained tissue sections obtained from 36 patients. The native fluorescence spectra of tissues are analyzed by the optical redox ratio (ORR), which is defined as fluorescence intensity ratio between NADH and FAD, and indicates the metabolism change with the cancer development. The ORRs of normal tissues are consistently higher than those of precancer or cancerous tissues. A criterion line of ORR at 5.0 can be used to discriminate cervical precancer/cancer from normal tissues. The sensitivity and specificity of the native fluorescence spectroscopy method for cervical cancer diagnosis are determined as 100% and 91%. Moreover, the native fluorescence spectroscopy study is much more sensitive on the healthy region of cervical precancer/cancer patients compared with the traditional clinical staining method. The results suggest label‐free imaging and spectroscopy is a fast, highly sensitive and specific method on the detection of cervical cancer.      

Optical palpation for the visualization of tumor in human breast tissue

Accurate and effective removal of tumor in one operation is an important goal of breast‐conserving surgery. However, it is not always achieved. Surgeons often utilize manual palpation to assess the surgical margin and/or the breast cavity. Manual palpation, however, is subjective and has relatively low resolution. Here, we investigate a tactile imaging technique, optical palpation, for the visualization of tumor. Optical palpation generates maps of the stress at the surface of tissue under static preload compression. Stress is evaluated by measuring the deformation of a contacting thin compliant layer with known mechanical properties using optical coherence tomography. In this study, optical palpation is performed on 34 freshly excised human breast specimens. Wide field‐of‐view (up to ~46 × 46 mm) stress images, optical palpograms, are presented from four representative specimens, demonstrating the capability of optical palpation to visualize tumor. Median stress reported for adipose tissue, 4 kPa, and benign dense tissue, 8 kPa, is significantly lower than for invasive tumor, 60 kPa. In addition, we demonstrate that optical palpation provides contrast consistent with a related optical technique, quantitative micro‐elastography. This study demonstrates that optical palpation holds promise for visualization of tumor in breast‐conserving surgery.      

Combining hyperspectral imaging and chemometrics to assess and interpret the effects of environmental stressors on zebrafish eye images at tissue level

Changes on an organism by the exposure to environmental stressors may be characterized by hyperspectral images (HSI), which preserve the morphology of biological samples, and suitable chemometric tools. The approach proposed allows assessing and interpreting the effect of contaminant exposure on heterogeneous biological samples monitored by HSI at specific tissue levels. In this work, the model example used consists of the study of the effect of the exposure of chlorpyrifos‐oxon on zebrafish tissues. To assess this effect, unmixing of the biological sample images followed by tissue‐specific classification models based on the unmixed spectral signatures is proposed. Unmixing and classification are performed by multivariate curve resolution‐alternating least squares (MCR‐ALS) and partial least squares‐discriminant analysis (PLS‐DA), respectively. Crucial aspects of the approach are: (1) the simultaneous MCR‐ALS analysis of all images from 1 population to take into account biological variability and provide reliable tissue spectral signatures, and (2) the use of resolved spectral signatures from control and exposed populations obtained from resampling of pixel subsets analyzed by MCR‐ALS multiset analysis as information for the tissue‐specific PLS‐DA classification models. Classification results diagnose the presence of a significant effect and identify the spectral regions at a tissue level responsible for the biological change.      

Visualization of Plant Tissues by Optical Coherence Tomography

The internal structure of plant tissues was visualized with optical coherence tomography (OCT). This noninvasive method is suitable for examining intact plants; it produces two-dimensional images of plant tissues at a penetration depth of 1–2 mm from the surface. The potential use of OCT was assessed on Tradescantia blossfeldiana Mild. Plant tissue images measuring 1.5 × 2 mm were obtained in vivo with a spatial resolution of 15 m. The radiation power incident on a sample was 0.5 mW. The acquisition of a two-dimensional image consisting of 200 × 200 pixels required 1–3 s. The OCT method can be used to visualize not only plant tissues and tissue boundaries but also the structure of individual cells.     

In situ tissue classification during laser ablation using acoustic signals

We suggest a novel method to classify the type of tissue that is being ablated, using the recorded acoustic sound waves during pulsed ultraviolet laser ablation. The motivation of the current research is tissue classification during vascular interventions, where the identification of the ablated tissue is vital. We classify the acoustic signatures using Mel‐frequency cepstral coefficients (MFCCs) feature extraction with a Support Vector Machine (SVM) algorithm, and in addition, use a fully connected deep neural network (FC‐DNN) algorithm. First, we classify three different liquids using our method as a preliminary proof of concept. Then, we classify ex vivo porcine aorta and bovine tendon tissues in the presence of saline. Finally, we classify ex vivo porcine aorta and bovine tendon tissues where the acoustic signals are recorded through chicken breast medium. The results for tissue classification in saline and through chicken breast both show high accuracy (>98%), based on tens of thousands of acoustic signals for each experiment. The experiments were conducted in a noisy and challenging setting that tries to imitate practical working conditions. The obtained results could pave the way towards practical tissue classification in various important medical procedures, achieving enhanced efficacy and safety.     

Mass type-specific sparse representation for mass classification in computer-aided detection on mammograms

Background

Breast cancer is the leading cause of both incidence and mortality in women population. For this reason, much research effort has been devoted to develop Computer-Aided Detection (CAD) systems for early detection of the breast cancers on mammograms. In this paper, we propose a new and novel dictionary configuration underpinning sparse representation based classification (SRC). The key idea of the proposed algorithm is to improve the sparsity in terms of mass margins for the purpose of improving classification performance in CAD systems.

Methods

The aim of the proposed SRC framework is to construct separate dictionaries according to the types of mass margins. The underlying idea behind our method is that the separated dictionaries can enhance the sparsity of mass class (true-positive), leading to an improved performance for differentiating mammographic masses from normal tissues (false-positive). When a mass sample is given for classification, the sparse solutions based on corresponding dictionaries are separately solved and combined at score level. Experiments have been performed on both database (DB) named as Digital Database for Screening Mammography (DDSM) and clinical Full Field Digital Mammogram (FFDM) DBs. In our experiments, sparsity concentration in the true class (SCTC) and area under the Receiver operating characteristic (ROC) curve (AUC) were measured for the comparison between the proposed method and a conventional single dictionary based approach. In addition, a support vector machine (SVM) was used for comparing our method with state-of-the-arts classifier extensively used for mass classification.

Results

Comparing with the conventional single dictionary configuration, the proposed approach is able to improve SCTC of up to 13.9% and 23.6% on DDSM and FFDM DBs, respectively. Moreover, the proposed method is able to improve AUC with 8.2% and 22.1% on DDSM and FFDM DBs, respectively. Comparing to SVM classifier, the proposed method improves AUC with 2.9% and 11.6% on DDSM and FFDM DBs, respectively.

Conclusions

The proposed dictionary configuration is found to well improve the sparsity of dictionaries, resulting in an enhanced classification performance. Moreover, the results show that the proposed method is better than conventional SVM classifier for classifying breast masses subject to various margins from normal tissues.

     

Spectral histopathology of the lung: A review of two large studies

This paper summarizes results from two large lung cancer studies comprising over 700 samples that demonstrate the ability of spectral histopathology (SHP) to distinguish cancerous tissue regions from normal tissue, to differentiate benign lesions from normal tissue and cancerous lesions, and to classify lung cancer types. Furthermore, malignancy‐associated changes can be identified in cancer‐adjacent normal tissue. The ability to differentiate a multitude of normal cells and tissue types allow SHP to identify tumor margins and immune cell infiltration. Finally, SHP easily distinguishes small cell lung cancer (SCLC) from non‐SCLC (NSCLC) and provides a further differentiation of NSCLC into adenocarcinomas and squamous cell carcinomas with an accuracy comparable of classical histopathology combined with immunohistochemistry. Case studies are presented that demonstrates that SHP can resolve interobserver discrepancies in standard histopathology.     

Development of an image pre‐processor for operational hyperspectral laryngeal cancer detection

Hyperspectral imaging (HSI) is a technology with high potential in the field of non‐invasive detection of cancer. However, in complex imaging situations like HSI of the larynx with a rigid endoscope, various image interferences can disable a proper classification of cancerous tissue. We identified three main problems: i) misregistration of single images in a HS cube due to patient heartbeat ii) image noise and iii) specular reflections (SR). Consequently, an image pre‐processor is developed in the current paper to overcome these image interferences. It encompasses i) image registration ii) noise removal by minimum noise fraction (MNF) transformation and iii) a novel SR detection method. The results reveal that the pre‐processor improves classification performance, while the newly developed SR detection method outperforms global thresholding technique hitherto used by 46%. The novel pre‐processor will be used for future studies towards the development of an operational scheme for HS‐based larynx cancer detection.

RGB image of the larynx derived from the hyperspectral cube and corresponding specular reflections ( a ) manually segmented and ( b ) detected by a novel specular reflection detection method.     

A novel urine analysis technique combining affinity chromatography with Au nanoparticle based surface enhanced Raman spectroscopy for potential applications in non‐invasive cancer screening

Modified nucleoside in urine samples is one of the most common biomarkers for cancer screening. Therefore, we developed a novel detection method for modified nucleoside detection in human urine. In this work, the modified nucleoside from real cancer patient's urine samples was first separated and purified using the affinity chromatography (AC) technology relying on its specific adsorption capacity. Then, surface‐enhanced Raman spectroscopy (SERS) technology with the capability of single molecular detection was used to sensitively characterize the biomolecular features of modified nucleoside. A total of 141 high‐quality SERS spectra of urinary modified nucleoside can be obtained from 50 gastric cancer patients and 43 breast cancer patients, as well as 48 healthy volunteers. Using principal component analysis combined with linear discriminant analysis (PCA‐LDA), the diagnostic sensitivities for identifying gastric cancer vs normal, breast cancer vs normal, gastric cancer vs breast cancer were 84.0%, 76.7% and 82.0%, respectively, and the corresponding diagnostic specificities for each combination were 95.8%, 87.5% and 90.7%, respectively. These results show that this novel method based on urinary modified nucleoside detection combining AC and SERS technologies holds promising potential for developing a specific, non‐invasive and label‐free tool for cancer screening.      

ImageParser: a tool for finite element generation from three-dimensional medical images

Background

The finite element method (FEM) is a powerful mathematical tool to simulate and visualize the mechanical deformation of tissues and organs during medical examinations or interventions. It is yet a challenge to build up an FEM mesh directly from a volumetric image partially because the regions (or structures) of interest (ROIs) may be irregular and fuzzy.

Methods

A software package, ImageParser, is developed to generate an FEM mesh from 3-D tomographic medical images. This software uses a semi-automatic method to detect ROIs from the context of image including neighboring tissues and organs, completes segmentation of different tissues, and meshes the organ into elements.

Results

The ImageParser is shown to build up an FEM model for simulating the mechanical responses of the breast based on 3-D CT images. The breast is compressed by two plate paddles under an overall displacement as large as 20% of the initial distance between the paddles. The strain and tangential Young's modulus distributions are specified for the biomechanical analysis of breast tissues.

Conclusion

The ImageParser can successfully exact the geometry of ROIs from a complex medical image and generate the FEM mesh with customer-defined segmentation information.

     

Augmented CD133 expression in distal margin correlates with poor prognosis in colorectal cancer

Pathological assessment of excised tumour and surgical margins in colorectal cancer (CRC) play crucial role in prognosis after surgery. Molecular assessment of margins could be more sensitive and informative than conventional histopathological analysis. Considering this view, we evaluated the distal surgical margins for expression of cancer stem cell (CSC) markers. Cellular and molecular assessment of normal, tumour and distal margin tissues were performed by flow cytometry, real‐time q‐PCR and immuno‐histochemical analysis for CRC patients after tumour excision. CRC patients were evaluated for expression of CSC markers in their normal, tumour and distal tissues. Flow cytometry assay revealed CD133 and CD44 enriched cells in distal margin and tumour compared to normal colorectal tissues, which was further confirmed by immunohistochemistry. Most importantly, immunohistochemistry also revealed the enrichment of CSC markers expression in pathologically negative distal margins. Patients with distal margin enriched for CD133 expression showed an increased recurrence rate and decreased disease‐free survival. This study proposes that although distal margin seems to be tumour free in conventional histopathological analysis, it could harbour cells enriched for CSC markers. Further CD133 could be a promising molecule to be used in molecular pathology for disease prognosis after surgery in CRC patients.