We have optimized our prior phantom-based test method for cerebral oximetry performance using a new 3D-printed cerebrovascular module (CVM). In addition, we have and added a surrogate for a thin perfused scalp layer to evaluate perfusion- related confounding factors. The new CVM’s optical properties better represented biological tissue and also incorporated a water-mimicking dye. The modular phantom also included biologically relevant scalp/skull and cerebrospinal fluid (CSF) layers. Performance testing of two commercially available clinical oximeters with the modified CVM over a range of oxygen saturation levels illustrates the utility of our solid phantom-based approach for standardized cerebral oximeter performance assessment.
Near-infrared spectroscopy (NIRS) is emerging as a rapid, low-cost approach for point-of-care triage of hematomas resulting from traumatic brain injury. However, there remains a lack of standardized test methods for benchtop performance assessment of these devices and incomplete understanding of relevant light–tissue interactions. We propose a phantom-based test method for systems operating near the 800-nm oxy-/deoxy-hemoglobin isosbestic point and implement it to evaluate a clinical system. Semi-idealized phantom geometries are designed to represent epidural/subdural, subarachnoid, and intracerebral hemorrhages. Measurements of these phantoms are made with a commercial NIRS-based hematoma detector to quantify the effect of hematoma type, depth, and size, as well as measurement repeatability and detector positioning relative to the hematoma. Results indicated high sensitivity to epidural/subdural and subarachnoid hematomas. Intracerebral hematomas are detectable to a maximum depth of ∼2.5 cm, depending on thickness and diameter. The maximum lateral detection area for the single-emitter/single-collector device studied here appears elliptical and decreases strongly with inclusion depth. Overall, this study provides unique insights into hematoma detector function and indicates the utility of modular polymer tissue phantoms in performance tests for emerging NIRS-based cerebral diagnostic technology.
Cerebral oximetry based on near-infrared spectroscopy (NIRS) has seen increasing clinical use for monitoring of premature infants as well as during neonatal, pediatric and adult cardiac surgery. One key confounding factor and a likely contributor to observed inconsistency amongst commercial NIRS oximeters is skin pigmentation. Clinical studies have shown negative bias in oxygen saturation (StO2) with increasing melanin content. In prior work, we developed a cerebral oximetry phantom comprised of a 3D-printed channel array module representing brain tissue and molded silicone layers simulating extracranial regions. The purpose of the current study was to develop and test epidermis-simulating layers that exhibit realistic human pigmentation properties. Initially, we performed spectroscopic characterization of potential melanin simulating agents – including coffee, India ink, synthetic melanin, and water-soluble nigrosin – in a polydimethylsiloxane (PDMS) substrate. We determined that the NIR absorption spectrum of water-soluble nigrosin most accurately matched human melanin. Layers of 0.1 mm thickness were fabricated with different nigrosin concentrations to simulate epidermis with light, moderate, and dark pigmentation. The brain module channels were filled with bovine blood in the 30-100% oxygenation range and measurements performed with neonatal/pediatric probes from commercially available cerebral oximeters. We found that StO2 reported by the oximeters decreased monotonically with increasing pigmentation level. The magnitude of this impact increased with decreasing StO2, producing a maximum change in saturation of approximately 8%. The consistency of our results with prior clinical findings provides preliminary evidence of the utility of our approach for assessing the impact of epidermal melanin in phantom-based performance testing.
Clinical cerebral oximeters based on near-infrared spectroscopy (NIRS) are a commonly used, non-invasive tool for intraoperative monitoring of hemoglobin saturation. Research to verify performance of cerebral oximeters in human subject trials has shown differences between commercially available devices. Test methods based on tissue-simulating phantoms have been proposed to augment clinical findings. While prior studies have focused on liquid phantoms, this work is aimed at developing methods based on solid polymer phantoms that are stable. Specifically, we have designed and fabricated a neonatal/pediatric head mimicking layered phantoms based on a 3D-printed cerebral matrix incorporating an array of vessel-simulating linear channels. Superficial layers incorporating homogeneous molded polydimethylsiloxane (PDMS) slabs were fabricated to represent CSF, scalp and skull regions. The cerebral matrix was filled with bovine blood desaturated with sodium dithionite to achieve oxygenation levels across the 40-90% range. Measurements were performed with a commercially available cerebral oximeter using two probes with different illumination-collection geometries, as designed for neonatal and pediatric patients. Reference measurements of samples were performed with a CO-oximeter before injection and after extraction. Results from applied cerebral oximeters indicate a strong sensitivity to the thickness of the superficial layer of the phantom. Better correlation with the reference CO-oximeter results were obtained in the superficial layer thickness of 0.8-2.5 mm range. Channel array phantoms with modular superficial layers represent a promising approach for performance testing of NIRS-based cerebral oximeters.
Near-infrared spectroscopy (NIRS) has emerged as a low-cost, portable approach for rapid, point-of-care detection of hematomas caused by traumatic brain injury. As a new technology, there is a need to develop standardized test methods for objective, quantitative performance evaluation of these devices. Towards this goal, we have developed and studied two types of phantom-based testing approaches. The first involves 3D-printed phantoms incorporating hemoglobin-filled inclusions. Phantom layers representing specific cerebral tissues were printed using photopolymers doped with varying levels of titanium oxide and black resin. The accuracy, precision and spectral dependence of printed phantom optical properties were validated using spectrophotometry. The phantom also includes a hematoma inclusion insert which was filled with a hemoglobin solution. Oxygen saturation levels were modified by adding sodium dithionite at calibrated concentrations. The second phantom approach involves molded silicone layers with a superficial region – simulating the scalp and skull – comprised of removable layers to vary hematoma size and depth, and a bottom layer representing brain matter. These phantoms were tested with both a commercial hematoma detector and a custom NIRS system to optimize their designs and validate their utility in performing inter-device comparisons. The effects of hematoma depth, diameter, and height, as well as tissue optical properties and biological variables including hemoglobin saturation level and scalp/skull thickness were studied. Results demonstrate the ability to quantitatively compare NIRS device performance and indicate the promise of using 3D printing to achieve phantoms with realistic variations in tissue optical properties for evaluating biophotonic device performance.
Mobile phone cameras employ sensors with near-infrared (NIR) sensitivity, yet this capability has not been exploited for biomedical purposes. Removing the IR-blocking filter from a phone-based camera opens the door to a wide range of techniques and applications for inexpensive, point-of-care biophotonic imaging and sensing. This study provides proof of principle for one of these modalities – phone-based NIR fluorescence imaging. An imaging system was assembled using a 780 nm light source along with excitation and emission filters with 800 nm and 825 nm cut-off wavelengths, respectively. Indocyanine green (ICG) was used as an NIR fluorescence contrast agent in an ex vivo rodent model, a resolution test target and a 3D-printed, tissue-simulating vascular phantom. Raw and processed images for red, green and blue pixel channels were analyzed for quantitative evaluation of fundamental performance characteristics including spectral sensitivity, detection linearity and spatial resolution. Mobile phone results were compared with a scientific CCD. The spatial resolution of CCD system was consistently superior to the phone, and green phone camera pixels showed better resolution than blue or green channels. The CCD exhibited similar sensitivity as processed red and blue pixels channels, yet a greater degree of detection linearity. Raw phone pixel data showed lower sensitivity but greater linearity than processed data. Overall, both qualitative and quantitative results provided strong evidence of the potential of phone-based NIR imaging, which may lead to a wide range of applications from cancer detection to glucose sensing.
Early detection of neoplastic changes remains a critical challenge in clinical cancer diagnosis and treatment. Many cancers arise from epithelial layers such as those of the gastrointestinal (GI) tract. Current standard endoscopic technology is unable to detect those subsurface lesions. Since cancer development is associated with both morphological and molecular alterations, imaging technologies that can quantitative image tissue’s morphological and molecular biomarkers and assess the depth extent of a lesion in real time, without the need for tissue excision, would be a major advance in GI cancer diagnostics and therapy. In this research, we investigated the feasibility of multi-modal optical imaging including high-resolution optical coherence tomography (OCT) and depth-resolved high-sensitivity fluorescence laminar optical tomography (FLOT) for structural and molecular imaging. APC (adenomatous polyposis coli) mice model were imaged using OCT and FLOT and the correlated histopathological diagnosis was obtained. Quantitative structural (the scattering coefficient) and molecular imaging parameters (fluorescence intensity) from OCT and FLOT images were developed for multi-parametric analysis. This multi-modal imaging method has demonstrated the feasibility for more accurate diagnosis with 87.4% (87.3%) for sensitivity (specificity) which gives the most optimal diagnosis (the largest area under receiver operating characteristic (ROC) curve). This project results in a new non-invasive multi-modal imaging platform for improved GI cancer detection, which is expected to have a major impact on detection, diagnosis, and characterization of GI cancers, as well as a wide range of epithelial cancers.
Near‐Infrared Spectroscopy (NIRS) is an emerging medical countermeasure for rapid, field detection of hematomas
caused by traumatic brain injury (TBI). Bench and animal tests to determine NIRS sensitivity and specificity are needed.
However, current animal models involving non-invasively induced, localized neural damage are limited. We
investigated an in vivo murine hematoma model in which cerebral hemorrhage was induced noninvasively by high-intensity
focused ultrasound (HIFU) with calibrated positioning and parameters. To characterize the morphology of
induced hematomas, we used skull-intact histological evaluation. A multi-wavelength fiber-optic NIRS system with three
source-detector separation distances was used to detect hematoma A 1.1 MHz transducer produced consistent small-to-medium
hematoma localized to a single hemisphere, along with bruising of the scalp, with a low mortality rate. A 220
kHz transducer produced larger, more diffuse hematomas, with higher variability in size and a correspondingly higher
mortality rate. No skin bruising or blood accumulation between the skin and skull was observed following injury
application with the 220 kHz transducer. Histological analysis showed higher sensitivity for larger hematomas (>4x4
mm2). NIRS optical density change after HIFU was able to detect all hematomas, with sensitivity dependent on
wavelength and separation distance. While improvements in methods for validating cerebral blood distribution are
needed, the HIFU hematoma model provided useful insights that will inform development of biologically relevant,
performance test methods for cerebral NIRS systems.
Increasing numbers of devices are emerging which involve biophotonic imaging on a mobile platform. Therefore, effective test methods are needed to ensure that these devices provide a high level of image quality. We have developed novel phantoms for performance assessment of near infrared fluorescence (NIRF) imaging devices. Resin molding and 3D printing techniques were applied for phantom fabrication. Comparisons between two imaging approaches – a CCD-based scientific camera and an NIR-enabled mobile phone – were made based on evaluation of the contrast transfer function and penetration depth. Optical properties of the phantoms were evaluated, including absorption and scattering spectra and fluorescence excitation-emission matrices. The potential viability of contrastenhanced biological NIRF imaging with a mobile phone is demonstrated, and color-channel-specific variations in image quality are documented. Our results provide evidence of the utility of novel phantom-based test methods for quantifying image quality in emerging NIRF devices.
Hyperspectral reflectance imaging (HRI) is an emerging imaging modality being applied for clinical indications such as tissue oximetry, and cancer detection based on endogenous biological constituents including plasmonic nanoparticles. However, there is currently a lack of standardized test methods for objective, quantitative evaluation of HRI system performance. Contrast-detail analysis (CDA) is a phantom-based test method commonly used to evaluate medical imaging devices (e.g., mammography systems) in terms of their lower detection limit. We investigated a modified CDA (mCDA) method to quantify the detectability of gold nanoparticles by HRI systems. Silicone-based turbid phantoms containing micro-fluidic channels were developed for the mCDA tests. Polydimethylsiloxane (PDMS) phantom materials were doped with chromophores and scatterers to achieve biologically relevant optical properties (OPs). Molds were used to produce cylindrical channels of diameters 0.3 to 1.65 mm and depths of 0.2 mm inside the phantoms. Channels were filled with a mixture of hemoglobin and concentrations of gold nanorods (GNR) and measured with our HRI system. The contrast of GNRs was solved with a spectral unmixing algorithm from the reflectance spectra. The lowest detectable concentration was determined as a function of inclusion size and depth and plotted as modified contrast detail curve (mCDC). mCDCs were used to compare the detectabilities of the HRI system with different data processing algorithms. It is demonstrated that our mCDA test method involving turbid microchannel phantoms can help to elucidate the combined performance of imaging devices and plasmonic nanoparticle contrast agents. This approach may be useful for performing clinical trial standardization and device re-calibration, thus ensuring quality control and clinical performance.
The emerging technique of rapid prototyping with three-dimensional (3-D) printers provides a simple yet revolutionary method for fabricating objects with arbitrary geometry. The use of 3-D printing for generating morphologically biomimetic tissue phantoms based on medical images represents a potentially major advance over existing phantom approaches. Toward the goal of image-defined phantoms, we converted a segmented fundus image of the human retina into a matrix format and edited it to achieve a geometry suitable for printing. Phantoms with vessel-simulating channels were then printed using a photoreactive resin providing biologically relevant turbidity, as determined by spectrophotometry. The morphology of printed vessels was validated by x-ray microcomputed tomography. Channels were filled with hemoglobin (Hb) solutions undergoing desaturation, and phantoms were imaged with a near-infrared hyperspectral reflectance imaging system. Additionally, a phantom was printed incorporating two disjoint vascular networks at different depths, each filled with Hb solutions at different saturation levels. Light propagation effects noted during these measurements—including the influence of vessel density and depth on Hb concentration and saturation estimates, and the effect of wavelength on vessel visualization depth—were evaluated. Overall, our findings indicated that 3-D-printed biomimetic phantoms hold significant potential as realistic and practical tools for elucidating light–tissue interactions and characterizing biophotonic system performance.
The emerging technique of three-dimensional (3D) printing provides a revolutionary way to fabricate objects with biologically realistic geometries. Previously we have performed optical and morphological characterization of basic 3D printed tissue-simulating phantoms and found them suitable for use in evaluating biophotonic imaging systems. In this study we assess the potential for printing phantoms with irregular, image-defined vascular networks that can be used to provide clinically-relevant insights into device performance. A previously acquired fundus camera image of the human retina was segmented, embedded into a 3D matrix, edited to incorporate the tubular shape of vessels and converted into a digital format suitable for printing. A polymer with biologically realistic optical properties was identified by spectrophotometer measurements of several commercially available samples. Phantoms were printed with the retinal vascular network reproduced as ~1.0 mm diameter channels at a range of depths up to ~3 mm. The morphology of the printed vessels was verified by volumetric imaging with μ-CT. Channels were filled with hemoglobin solutions at controlled oxygenation levels, and the phantoms were imaged by a near-infrared hyperspectral reflectance imaging system. The effect of vessel depth on hemoglobin saturation estimates was studied. Additionally, a phantom incorporating the vascular network at two depths was printed and filled with hemoglobin solution at two different saturation levels. Overall, results indicated that 3D printed phantoms are useful for assessing biophotonic system performance and have the potential to form the basis of clinically-relevant standardized test methods for assessment of medical imaging modalities.
Optical coherence tomography (OCT) is a high resolution imaging technology that is rapidly being adopted as the
standard of care for medical applications such as ocular and intravascular imaging. However, clinical translation has
been hampered by the lack of standardized test methods for performance evaluation as well as consensus standards
analogous to those that have been developed for established medical imaging modalities (e.g., ultrasound). In this study,
we address low contrast detectability, specifically, the ability of systems to differentiate between regions exhibiting small
differences in scattering coefficient. Based on standard test methods for established medical imaging modalities, we
have developed layered phantoms with well-characterized scattering properties in a biologically relevant range. The
phantoms consisted of polydimethylsiloxane (PDMS) doped with varying concentrations of BaSO4 microparticles.
Microfabrication processes were used to create layered and channel schemes. Two spectral domain OCT systems - a
Fourier domain system at 855 nm and a swept-source device at 1310 nm - were then used to image the phantoms. The
detectability of regions with different scattering levels was evaluated for each system by measuring pixel intensity
differences. Confounding factors such as the inherent attenuation of the phantoms, signal intensity decay due to focusing
and system roll-off were also encountered and addressed. Significant differences between systems were noted. The
minimum differences in scattering coefficient that the Fourier domain and swept source systems could differentiate was
1.50 and 0.46 mm-1 respectively. Overall, this approach to evaluating low contrast detectability represents a key step
towards the development of standard test methods to facilitate clinical translation of novel OCT systems.
In biophotonic imaging, turbid phantoms that are low-cost, biologically-relevant, and durable are desired for
standardized performance assessment. Such phantoms often contain inclusions of varying depths and sizes in order to
quantify key image quality characteristics such as penetration depth, sensitivity and contrast detectability. The emerging
technique of rapid prototyping with three-dimensional (3D) printers provides a potentially revolutionary way to fabricate
these structures. Towards this goal, we have characterized the optical properties and morphology of phantoms fabricated
by two 3D printing approaches: thermosoftening and photopolymerization. Material optical properties were measured by
spectrophotometry while the morphology of phantoms incorporating 0.2-1.0 mm diameter channels was studied by μCT,
optical coherence tomography (OCT) and optical microscopy. A near-infrared absorbing dye and nanorods at several
concentrations were injected into channels to evaluate detectability with a near-infrared hyperspectral reflectance
imaging (HRI) system (650-1100 nm). Phantoms exhibited biologically-relevant scattering and low absorption across
visible and near-infrared wavelengths. Although limitations in resolution were noted, channels with diameters of 0.4
mm or more could be reliably fabricated. The most significant problem noted was the porosity of phantoms generated
with the thermosoftening-based printer. The aforementioned three imaging methods provided a valuable mix of insights
into phantom morphology and may also be useful for detailed structural inspection of medical devices fabricated by
rapid prototyping, such as customized implants. Overall, our findings indicate that 3D printing has significant potential
as a method for fabricating well-characterized, standard phantoms for medical imaging modalities such as HRI.
The emerging technique of three-dimensional (3D) printing provides a simple, fast, and flexible way to fabricate structures with arbitrary spatial features and may prove useful in the development of standardized, phantom-based performance test methods for biophotonic imaging. Acrylonitrile Butadiene Styrene (ABS) is commonly used in the printing process, given its low cost and strength. In this study, we evaluate 3D printing as an approach for fabricating biologically-relevant optical phantoms for hyperspectral reflectance imaging (HRI). The initial phase of this work involved characterization of absorption and scattering coefficients using spectrophotometry. The morphology of phantoms incorporating vessel-like channels with diameters on the order of hundreds of microns was examined by microscopy and OCT. A near-infrared absorbing dye was injected into channels located at a range of depths within the phantom and imaged with a near-infrared HRI system (650-1100 nm). ABS was found to have scattering coefficients comparable to biological tissue and low absorption throughout much of the visible and infrared range. Channels with dimensions on the order of the resolution limit of the 3D printer (~0.2 mm) exhibited pixelation effects as well as a degree of distortion along their edges. Furthermore, phantom porosity sometimes resulted in leakage from channel regions. Contrast-enhanced channel visualization with HRI was possible to a depth of nearly 1 mm – a level similar to that seen previously in biological tissue. Overall, our ABS phantoms demonstrated a high level of optical similarity to biological tissue. While limitations in printer resolution, matrix homogeneity and optical property tunability remain challenging, 3D printed phantoms have significant promise as samples for objective, quantitative evaluation of performance for biophotonic imaging modalities such as HRI.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.