Hyperspectral dark-field microscopy of human breast lumpectomy samples for tumor margin detection in breast-conserving surgery.

Significance: Hyperspectral dark-field microscopy (HSDFM) and data cube analysis algorithms demonstrate successful detection and classification of various tissue types, including carcinoma regions in human post-lumpectomy breast tissues excised during breast-conserving surgeries. Aim: We expand the application of HSDFM to the classification of tissue types and tumor subtypes in pre-histopathology human breast lumpectomy samples. Approach: Breast tissues excised during breast-conserving surgeries were imaged by the HSDFM and analyzed. The performance of the HSDFM is evaluated by comparing the backscattering intensity spectra of polystyrene microbead solutions with the Monte Carlo simulation of the experimental data. For classification algorithms, two analysis approaches, a supervised technique based on the spectral angle mapper (SAM) algorithm and an unsupervised technique based on the K-means algorithm are applied to classify various tissue types including carcinoma subtypes. In the supervised technique, the SAM algorithm with manually extracted endmembers guided by H&E annotations is used as reference spectra, allowing for segmentation maps with classified tissue types including carcinoma subtypes. Results: The manually extracted endmembers of known tissue types and their corresponding threshold spectral correlation angles for classification make a good reference library that validates endmembers computed by the unsupervised K-means algorithm. The unsupervised K-means algorithm, with no a priori information, produces abundance maps with dominant endmembers of various tissue types, including carcinoma subtypes of invasive ductal carcinoma and invasive mucinous carcinoma. The two carcinomas' unique endmembers produced by the two methods agree with each other within <2% residual error margin. Conclusions: Our report demonstrates a robust procedure for the validation of an unsupervised algorithm with the essential set of parameters based on the ground truth, histopathological information. We have demonstrated that a trained library of the histopathology-guided endmembers and associated threshold spectral correlation angles computed against well-defined reference data cubes serve such parameters. Two classification algorithms, supervised and unsupervised algorithms, are employed to identify regions with carcinoma subtypes of invasive ductal carcinoma and invasive mucinous carcinoma present in the tissues. The two carcinomas' unique endmembers used by the two methods agree to <2% residual error margin. This library of high quality and collected under an environment with no ambient background may be instrumental to develop or validate more advanced unsupervised data cube analysis algorithms, such as effective neural networks for efficient subtype classification.

Characterization of an Iodinated Rectal Spacer for Prostate Photon and Proton Radiation Therapy.

PURPOSE: Conventional rectal spacers (nonI-SPs) are low-contrast on computed tomography (CT), often necessitating magnetic resonance imaging for accurate delineation. A new formulation of spacers (I-SPs) incorporates iodine to improve radiopacity and CT visualization. We characterized placement, stability, and plan quality of I-SPs compared to nonI-SPs. METHODS AND MATERIALS: Patients with intact prostate cancer (n = 50) treated with I-SPs and photons were compared to randomly selected patients (n = 50) with nonI-SPs (photon or proton therapy). The I-SP was contoured on the planning CT and cone beam CTs at 3 timepoints: first, middle, and final treatment (n = 200 scans). I-SPs Hounsfield units (HU), volume, surface area (SA), centroid position relative to prostate centroid, and distance between prostate/rectum centroids were compared on the planning CTs between each cohort. I-SP changes were evaluated on cone beam CTs over courses of treatment. Dosimetric evaluations of plan quality and robustness were performed. I-SP was tested in a phantom to characterize its relative linear stopping power for protons. RESULTS: I-SPs yielded a distinct visible contrast on planning CTs compared to nonI-SPs (HU 138 vs 12, P < .001), allowing delineation on CT alone. The delineated volume and SA of I-SPs were smaller than nonI-SPs (volume 8.9 vs 10.6 mL, P < .001; SA 28 vs 35 cm2, P < .001), yet relative spacer position and prostate-rectal separation were similar (P = .79). No significant change in HU, volume, SA, or relative position of the I-SPs hydrogel occurred over courses of treatment (all P > .1). Dosimetric analysis concluded there were no significant changes in plan quality or robustness for I-SPs compared to nonI-SPs. The I-SP relative linear stopping power was 1.018, necessitating HU override for proton planning. CONCLUSIONS: I-SPs provide a manifest CT contrast, allowing for delineation on planning CT alone with no magnetic resonance imaging necessary. I-SPs radiopacity, size, and relative position remained stable over courses of treatment from 28 to 44 fractions. No changes in plan quality or robustness were seen comparing I-SPs and nonI-SPs.

Modeling and Synthesis of Breast Cancer Optical Property Signatures With Generative Models.

Is it possible to find deterministic relationships between optical measurements and pathophysiology in an unsupervised manner and based on data alone? Optical property quantification is a rapidly growing biomedical imaging technique for characterizing biological tissues that shows promise in a range of clinical applications, such as intraoperative breast-conserving surgery margin assessment. However, translating tissue optical properties to clinical pathology information is still a cumbersome problem due to, amongst other things, inter- and intrapatient variability, calibration, and ultimately the nonlinear behavior of light in turbid media. These challenges limit the ability of standard statistical methods to generate a simple model of pathology, requiring more advanced algorithms. We present a data-driven, nonlinear model of breast cancer pathology for real-time margin assessment of resected samples using optical properties derived from spatial frequency domain imaging data. A series of deep neural network models are employed to obtain sets of latent embeddings that relate optical data signatures to the underlying tissue pathology in a tractable manner. These self-explanatory models can translate absorption and scattering properties measured from pathology, while also being able to synthesize new data. The method was tested on a total of 70 resected breast tissue samples containing 137 regions of interest, achieving rapid optical property modeling with errors only limited by current semi-empirical models, allowing for mass sample synthesis and providing a systematic understanding of dataset properties, paving the way for deep automated margin assessment algorithms using structured light imaging or, in principle, any other optical imaging technique seeking modeling. Code is available.

Modeling Resistance and Recurrence Patterns of Combined Targeted-Chemoradiotherapy Predicts Benefit of Shorter Induction Period.

Optimal integration of molecularly targeted therapies, such as tyrosine kinase inhibitors (TKI), with concurrent chemotherapy and radiation (CRT) to improve outcomes in genotype-defined cancers remains a current challenge in clinical settings. Important questions regarding optimal scheduling and length of induction period for neoadjuvant use of targeted agents remain unsolved and vary among clinical trial protocols. Here, we develop and validate a biomathematical framework encompassing drug resistance and radiobiology to simulate patterns of local versus distant recurrences in a non-small cell lung cancer (NSCLC) population with mutated EGFR receiving TKIs and CRT. Our model predicted that targeted induction before CRT, an approach currently being tested in clinical trials, may render adjuvant targeted therapy less effective due to proliferation of drug-resistant cancer cells when using very long induction periods. Furthermore, simulations not only demonstrated the competing effects of drug-resistant cell expansion versus overall tumor regression as a function of induction length, but also directly estimated the probability of observing an improvement in progression-free survival at a given cohort size. We thus demonstrate that such stochastic biological simulations have the potential to quantitatively inform the design of multimodality clinical trials in genotype-defined cancers. SIGNIFICANCE: A biomathematical framework based on fundamental principles of evolution and radiobiology for in silico clinical trial design allows clinicians to optimize administration of TKIs before chemoradiotherapy in oncogene-driven NSCLC.

Structured light imaging for breast-conserving surgery, part II: texture analysis and classification.

Subdiffuse spatial frequency domain imaging (sd-SFDI) data of 42 freshly excised, bread-loafed tumor resections from breast-conserving surgery (BCS) were evaluated using texture analysis and a machine learning framework for tissue classification. Resections contained 56 regions of interest (RoIs) determined by expert histopathological analysis. RoIs were coregistered with sd-SFDI data and sampled into ∼4 × 4 mm2 subimage samples of confirmed and homogeneous histological categories. Sd-SFDI reflectance textures were analyzed using gray-level co-occurrence matrix pixel statistics, image primitives, and power spectral density curve parameters. Texture metrics exhibited statistical significance (p-value < 0.05) between three benign and three malignant tissue subtypes. Pairs of benign and malignant subtypes underwent texture-based, binary classification with correlation-based feature selection. Classification performance was evaluated using fivefold cross-validation and feature grid searching. Classification using subdiffuse, monochromatic reflectance (illumination spatial frequency of fx = 1.37 mm − 1, optical wavelength of λ = 490 nm) achieved accuracies ranging from 0.55 (95% CI: 0.41 to 0.69) to 0.95 (95% CI: 0.90 to 1.00) depending on the benign­malignant diagnosis pair. Texture analysis of sd-SFDI data maintains the spatial context within images, is free of light transport model assumptions, and may provide an alternative, computationally efficient approach for wide field-of-view (cm2) BCS tumor margin assessment relative to pixel-based optical scatter or color properties alone.

Structured light imaging for breast-conserving surgery, part I: optical scatter and color analysis.

Structured light imaging (SLI) with high spatial frequency (HSF) illumination provides a method to amplify native tissue scatter contrast and better differentiate superficial tissues. This was investigated for margin analysis in breast-conserving surgery (BCS) and imaging gross clinical tissues from 70 BCS patients, and the SLI distinguishability was examined for six malignancy subtypes relative to three benign/normal breast tissue subtypes. Optical scattering images recovered were analyzed with five different color space representations of multispectral demodulated reflectance. Excluding rare combinations of invasive lobular carcinoma and fibrocystic disease, SLI was able to classify all subtypes of breast malignancy from surrounding benign tissues (p-value < 0.05) based on scatter and color parameters. For color analysis, HSF illumination of the sample generated more statistically significant discrimination than regular uniform illumination. Pathological information about lesion subtype from a presurgical biopsy can inform the search for malignancy on the surfaces of specimens during BCS, motivating the focus on pairwise classification analysis. This SLI modality is of particular interest for its potential to differentiate tissue classes across a wide field-of-view (∼100 cm2) and for its ability to acquire images of macroscopic tissues rapidly but with microscopic-level sensitivity to structural and morphological tissue constituents.

Review of methods for intraoperative margin detection for breast conserving surgery.

Breast conserving surgery (BCS) is an effective treatment for early-stage cancers as long as the margins of the resected tissue are free of disease according to consensus guidelines for patient management. However, 15% to 35% of patients undergo a second surgery since malignant cells are found close to or at the margins of the original resection specimen. This review highlights imaging approaches being investigated to reduce the rate of positive margins, and they are reviewed with the assumption that a new system would need high sensitivity near 95% and specificity near 85%. The problem appears to be twofold. The first is for complete, fast surface scanning for cellular, structural, and/or molecular features of cancer, in a lumpectomy volume, which is variable in size, but can be large, irregular, and amorphous. A second is for full, volumetric imaging of the specimen at high spatial resolution, to better guide internal radiologic decision-making about the spiculations and duct tracks, which may inform that surfaces are involved. These two demands are not easily solved by a single tool. Optical methods that scan large surfaces quickly are needed with cellular/molecular sensitivity to solve the first problem, but volumetric imaging with high spatial resolution for soft tissues is largely outside of the optical realm and requires x-ray, micro-CT, or magnetic resonance imaging if they can be achieved efficiently. In summary, it appears that a combination of systems into hybrid platforms may be the optimal solution for these two very different problems. This concept must be cost-effective, image specimens within minutes and be coupled to decision-making tools that help a surgeon without adding to the procedure. The potential for optical systems to be involved in this problem is emerging and clinical trials are underway in several of these technologies to see if they could reduce positive margin rates in BCS.

Micro-computed tomography enables rapid surgical margin assessment during breast conserving surgery (BCS): correlation of whole BCS micro-CT readings to final histopathology.

BACKGROUND: Roughly 23% of breast conserving surgery (BCS) patients undergo a second re-excision procedure due to pathologically positive surgical margins. We investigated the feasibility and potential value of micro-Computed Tomography (micro-CT) as a surgical margin guidance tool during BCS. METHODS: A cohort of 32 BCS specimens was prospectively imaged with a pre-clinical micro-CT system upon arrival in the surgical pathology laboratory. Reconstructed micro-CT scans were evaluated retrospectively by an experienced breast radiologist, who provided binary determinations whether lesions extended to the specimen margin. These readings were then compared to the final pathological diagnosis and to 2D specimen radiography readings. RESULTS: Of the 32 specimens imaged, 28 had malignant and four had benign pathological diagnoses. Overall five (four malignant, one benign) of the 32 specimens had lesion tissue extending to the margin. For all 32 specimens, micro-CT reconstructions were calculated (< 4 min. acquisition + reconstruction time) and each specimen was volumetrically analyzed by a radiologist. Of the 28 malignant specimen readings, 18 matched the final pathological diagnosis [64%, 95 CI (47%-81%)], with a negative predictive value of 89% [95 CI (74%-96%)]. Micro-CT readings revealed changes in the tumor location and margin status as compared to single-projection radiography readings. CONCLUSIONS: Micro-CT scanning of BCS specimens enabled margin status assessment over the entirety of the surgical surface in a clinically relevant time frame, provided additional spatial information over single-projection radiography, and may be a potentially useful BCS guidance tool.

Light scattering measured with spatial frequency domain imaging can predict stromal versus epithelial proportions in surgically resected breast tissue.

This study aims to determine if light scatter parameters measured with spatial frequency domain imaging (SFDI) can accurately predict stromal, epithelial, and adipose fractions in freshly resected, unstained human breast specimens. An explicit model was developed to predict stromal, epithelial, and adipose fractions as a function of light scattering parameters, which was validated against a quantitative analysis of digitized histology slides for N = 31 specimens using leave-one-out cross-fold validation. Specimen mean stromal, epithelial, and adipose volume fractions predicted from light scattering parameters strongly correlated with those calculated from digitized histology slides (r = 0.90, 0.77, and 0.91, respectively, p-value <1 × 10 - 6). Additionally, the ratio of predicted epithelium to stroma classified malignant specimens with a sensitivity and specificity of 90% and 81%, respectively, and also classified all pixels in malignant lesions with 63% and 79%, at a threshold of 1. All specimens and pixels were classified as malignant, benign, or fat with 84% and 75% accuracy, respectively. These findings demonstrate how light scattering parameters acquired with SFDI can be used to accurately predict and spatially map stromal, epithelial, and adipose proportions in fresh unstained, human breast tissue, and suggest that these estimations could provide diagnostic value.

Wide-field color imaging of scatter-based tissue contrast using both high spatial frequency illumination and cross-polarization gating.

This study characterizes the scatter-specific tissue contrast that can be obtained by high spatial frequency (HSF) domain imaging and cross-polarization (CP) imaging, using a standard color imaging system, and how combining them may be beneficial. Both HSF and CP approaches are known to modulate the sensitivity of epi-illumination reflectance images between diffuse multiply scattered and superficially backscattered photons, providing enhanced contrast from microstructure and composition than what is achieved by standard wide-field imaging. Measurements in tissue-simulating optical phantoms show that CP imaging returns localized assessments of both scattering and absorption effects, while HSF has uniquely specific sensitivity to scatter-only contrast, with a strong suppression of visible contrast from blood. The combination of CP and HSF imaging provided an expanded sensitivity to scatter compared with CP imaging, while rejecting specular reflections detected by HSF imaging. ex vivo imaging of an atlas of dissected rodent organs/tissues demonstrated the scatter-based contrast achieved with HSF, CP and HSF-CP imaging, with the white light spectral signal returned by each approach translated to a color image for intuitive encoding of scatter-based contrast within images of tissue. The results suggest that visible CP-HSF imaging could have the potential to aid diagnostic imaging of lesions in skin or mucosal tissues and organs, where just CP is currently the standard practice imaging modality.

Calibration and analysis of a multimodal micro-CT and structured light imaging system for the evaluation of excised breast tissue.

A multimodal micro-computed tomography (CT) and multi-spectral structured light imaging (SLI) system is introduced and systematically analyzed to test its feasibility to aid in margin delineation during breast conserving surgery (BCS). Phantom analysis of the micro-CT yielded a signal-to-noise ratio of 34, a contrast of 1.64, and a minimum detectable resolution of 240 µm for a 1.2 min scan. The SLI system, spanning wavelengths 490 nm to 800 nm and spatial frequencies up to 1.37 [Formula: see text], was evaluated with aqueous tissue simulating phantoms having variations in particle size distribution, scatter density, and blood volume fraction. The reduced scattering coefficient, [Formula: see text] and phase function parameter, γ, were accurately recovered over all wavelengths independent of blood volume fractions from 0% to 4%, assuming a flat sample geometry perpendicular to the imaging plane. The resolution of the optical system was tested with a step phantom, from which the modulation transfer function was calculated yielding a maximum resolution of 3.78 cycles per mm. The three dimensional spatial co-registration between the CT and optical imaging space was tested and shown to be accurate within 0.7 mm. A freshly resected breast specimen, with lobular carcinoma, fibrocystic disease, and adipose, was imaged with the system. The micro-CT provided visualization of the tumor mass and its spiculations, and SLI yielded superficial quantification of light scattering parameters for the malignant and benign tissue types. These results appear to be the first demonstration of SLI combined with standard medical tomography for imaging excised tumor specimens. While further investigations are needed to determine and test the spectral, spatial, and CT features required to classify tissue, this study demonstrates the ability of multimodal CT/SLI to quantify, visualize, and spatially navigate breast tumor specimens, which could potentially aid in the assessment of tumor margin status during BCS.

Monochromatic subdiffusive spatial frequency domain imaging provides in-situ sensitivity to intratumoral morphological heterogeneity in a murine model.

For the first time, spatially resolved quantitative metrics of light scattering recovered with sub-diffusive spatial frequency domain imaging (sd-SFDI) are shown to be sensitive to changes in intratumoral morphology and viability by direct comparison to histopathological analysis. Two freshly excised subcutaneous murine tumor cross-sections were measured with sd-SFDI, and recovered optical scatter parameter maps were co-registered to whole mount histology. Unique clustering of the optical scatter parameters µs' vs. γ (i.e. diffuse scattering vs. relative backscattering) evaluated at a single wavelength showed complete separation between regions of viable tumor, aggresive tumor with stromal growth, varying levels of necrotic tumor, and also peritumor muscle. The results suggest that with further technical development, sd-SFDI may represent a non-destructive screening tool for analysis of excised tissue or a non-invasive approach to investigate suspicious lesions without the need for exogenous labels or spectrally resolved imaging.

Wide-field quantitative imaging of tissue microstructure using sub-diffuse spatial frequency domain imaging.

Localized measurements of scattering in biological tissue provide sensitivity to microstructural morphology but have limited utility to wide-field applications, such as surgical guidance. This study introduces sub-diffusive spatial frequency domain imaging (sd-SFDI), which uses high spatial frequency illumination to achieve wide-field sampling of localized reflectances. Model-based inversion recovers macroscopic variations in the reduced scattering coefficient [Formula: see text] and the phase function backscatter parameter (γ). Measurements in optical phantoms show quantitative imaging of user-tuned phase-function-based contrast with accurate decoupling of parameters that define both the density and the size-scale distribution of scatterers. Measurements of fresh ex vivo breast tissue samples revealed, for the first time, unique clustering of sub-diffusive scattering properties for different tissue types. The results support that sd-SFDI provides maps of microscopic structural biomarkers that cannot be obtained with diffuse wide-field imaging and characterizes spatial variations not resolved by point-based optical sampling.

Molecular dyes used for surgical specimen margin orientation allow for intraoperative optical assessment during breast conserving surgery.

A variety of optical techniques utilizing near-infrared (NIR) light are being proposed for intraoperative breast tumor margin assessment. However, immediately following a lumpectomy excision, the margins are inked, which preserves the orientation of the specimen but prevents optical interrogation of the tissue margins. Here, a workflow is proposed that allows for both NIR optical assessment following full specimen marking using molecular dyes which have negligible absorption and scattering in the NIR. The effect of standard surgical inks in contrast to molecular dyes for an NIR signal is shown. Further, the proposed workflow is demonstrated with full specimen intraoperative imaging on all margins directly after the lumpectomy has been excised and completely marked. This work is an important step in the path to clinical feasibility of intraoperative breast tumor margin assessment using NIR optical methods without having to compromise on the current clinical practice of inking resected specimens for margin orientation.

Sub-diffusive scattering parameter maps recovered using wide-field high-frequency structured light imaging.

This study investigates the hypothesis that structured light reflectance imaging with high spatial frequency patterns [Formula: see text] can be used to quantitatively map the anisotropic scattering phase function distribution [Formula: see text] in turbid media. Monte Carlo simulations were used in part to establish a semi-empirical model of demodulated reflectance ([Formula: see text]) in terms of dimensionless scattering [Formula: see text] and [Formula: see text], a metric of the first two moments of the [Formula: see text] distribution. Experiments completed in tissue-simulating phantoms showed that simultaneous analysis of [Formula: see text] spectra sampled at multiple [Formula: see text] in the frequency range [0.05-0.5] [Formula: see text] allowed accurate estimation of both [Formula: see text] in the relevant tissue range [0.4-1.8] [Formula: see text], and [Formula: see text] in the range [1.4-1.75]. Pilot measurements of a healthy volunteer exhibited [Formula: see text]-based contrast between scar tissue and surrounding normal skin, which was not as apparent in wide field diffuse imaging. These results represent the first wide-field maps to quantify sub-diffuse scattering parameters, which are sensitive to sub-microscopic tissue structures and composition, and therefore, offer potential for fast diagnostic imaging of ultrastructure on a size scale that is relevant to surgical applications.

Structured light scatteroscopy.

A new imaging approach, structured light scatteroscopy (SLS), is demonstrated, which offers rapid wide-field imaging of microscopic morphological variations in bulk tissue surfaces. Elastic scattering of light offers exquisite sensitivity to ultrastructural changes at multiple size scales ranging from nanometers to millimeters, but in bulk tissues the confounding effects of molecular absorption and strong multiple scattering of light often lead to a dramatic reduction in scatter contrast and specificity. It is demonstrated that the SLS using structured high spatial frequency illumination and detection to probe the tissue achieves direct, absorption-independent, high-resolution maps of the scattering response. The scattering response is observed to be dependent on both the wavelength and spatial frequency of choice, indicating a potential for multiscale probing of ultrastructural changes in superficial tissue layers. This methodology can be easily applied in most wide-field imaging systems.

Projection imaging of photon beams by the Cerenkov effect.

PURPOSE: A novel technique for beam profiling of megavoltage photon beams was investigated for the first time by capturing images of the induced Cerenkov emission in water, as a potential surrogate for the imparted dose in irradiated media. METHODS: A high-sensitivity, intensified CCD camera (ICCD) was configured to acquire 2D projection images of Cerenkov emission from a 4 × 4 cm(2) 6 MV linear accelerator (LINAC) x-ray photon beam operating at a dose rate of 400 MU∕min incident on a water tank with transparent walls. The ICCD acquisition was gated to the LINAC sync pulse to reduce background light artifacts, and the measurement quality was investigated by evaluating the signal to noise ratio and measurement repeatability as a function of delivered dose. Monte Carlo simulations were used to derive a calibration factor for differences between the optical images and deposited dose arising from the anisotropic angular dependence of Cerenkov emission. Finally, Cerenkov-based beam profiles were compared to a percent depth dose (PDD) and lateral dose profile at a depth of d(max) from a reference dose distribution generated from the clinical Varian ECLIPSE treatment planning system (TPS). RESULTS: The signal to noise ratio was found to be 20 at a delivered dose of 66.6 cGy, and proportional to the square root of the delivered dose as expected from Poisson photon counting statistics. A 2.1% mean standard deviation and 5.6% maximum variation in successive measurements were observed, and the Monte Carlo derived calibration factor resulted in Cerenkov emission images which were directly correlated to deposited dose, with some spatial issues. The dose difference between the TPS and PDD predicted by Cerenkov measurements was within 20% in the buildup region with a distance to agreement (DTA) of 1.5-2 mm and ±3% at depths beyond d(max). In the lateral profile, the dose difference at the beam penumbra was within ±13% with a DTA of 0-2 mm, ±5% in the central beam region, and 2%-3% in the beam umbra. CONCLUSIONS: The results from this initial study demonstrate the first documented use of Cerenkov emission imaging to profile x-ray photon LINAC beams in water. The proposed modality has several potential advantages over alternative methods, and upon future refinement may prove to be a robust and novel dosimetry method.