Deep learning-enabled whole slide imaging (DeepWSI): oil-immersion quality using dry objectives, longer depth of field, higher system throughput, and better functionality.

Whole slide imaging (WSI) has moved the traditional manual slide inspection process to the era of digital pathology. A typical WSI system translates the sample to different positions and captures images using a high numerical aperture (NA) objective lens. Performing oil-immersion microscopy is a major obstacle for WSI as it requires careful liquid handling during the scanning process. Switching between dry objective and oil-immersion lens is often impossible as it disrupts the acquisition process. For a high-NA objective lens, the sub-micron depth of field also poses a challenge to acquiring in-focus images of samples with uneven topography. Additionally, it implies a small field of view for each tile, thus limiting the system throughput and resulting in a long acquisition time. Here we report a deep learning-enabled WSI platform, termed DeepWSI, to substantially improve the system performance and imaging throughput. With this platform, we show that images captured with a regular dry objective lens can be transformed into images comparable to that of a 1.4-NA oil immersion lens. Blurred images with defocus distance from -5 µm to +5 µm can be virtually refocused to the in-focus plane post measurement. We demonstrate an equivalent data throughput of >2 gigapixels per second, the highest among existing WSI systems. Using the same deep neural network, we also report a high-resolution virtual staining strategy and demonstrate it for Fourier ptychographic WSI. The DeepWSI platform may provide a turnkey solution for developing high-performance diagnostic tools for digital pathology.

Autofluorescence of blood and its application in biomedical and clinical research.

Autofluorescence of blood has been explored as a label free approach for detection of cell types, as well as for diagnosis and detection of infection, cancer, and other diseases. Although blood autofluorescence is used to indicate the presence of several physiological abnormalities with high sensitivity, it often lacks disease specificity due to use of a limited number of fluorophores in the detection of several abnormal conditions. In addition, the measurement of autofluorescence is sensitive to the type of sample, sample preparation, and spectroscopy method used for the measurement. Therefore, while current blood autofluorescence detection approaches may not be suitable for primary clinical diagnosis, it certainly has tremendous potential in developing methods for large scale screening that can identify high risk groups for further diagnosis using highly specific diagnostic tests. This review discusses the source of blood autofluorescence, the role of spectroscopy methods, and various applications that have used autofluorescence of blood, to explore the potential of blood autofluorescence in biomedical research and clinical applications.

Comparing the signal enhancement of a gadolinium based and an iron-oxide based contrast agent in low-field MRI.

Recently, there has been a renewed interest in low-field MRI. Contrast agents (CA) in MRI have magnetic behavior dependent on magnetic field strength. Therefore, the optimal contrast agent for low-field MRI might be different from what is used at higher fields. Ultra-small superparamagnetic iron-oxides (USPIOs), commonly used as negative CA, might also be used for generating positive contrast in low-field MRI. The purpose of this study was to determine whether an USPIO or a gadolinium based contrast agent is more appropriate at low field strengths. Relaxivity values of ferumoxytol (USPIO) and gadoterate (gadolinium based) were used in this research to simulate normalized signal intensity (SI) curves within a concentration range of 0-15 mM. Simulations were experimentally validated on a 0.25T MRI scanner. Simulations and experiments were performed using spin echo (SE), spoiled gradient echo (SGE), and balanced steady-state free precession (bSSFP) sequences. Maximum achievable SIs were assessed for both CAs in a range of concentrations on all sequences. Simulations at 0.25T showed a peak in SIs at low concentrations ferumoxytol versus a wide top at higher concentrations for gadoterate in SE and SGE. Experiments agreed well with the simulations in SE and SGE, but less in the bSSFP sequence due to overestimated relaxivities in simulations. At low magnetic field strengths, ferumoxytol generates similar signal enhancement at lower concentrations than gadoterate.

Fatal methemoglobinemia: A case series highlighting a new trend in intentional sodium nitrite or sodium nitrate ingestion as a method of suicide.

Unintentional exposure to nitrite- or nitrate-containing toxic salts is a recognized cause of acquired methemoglobinemia (MetHb). This systemic alteration of the blood can be fatal if not recognized and treated promptly. The intentional ingestion of sodium nitrite (NaNO2) or sodium nitrate (NaNO3), causing MetHb, is an uncommon and recently identified method of suicide, with the first reported case in the literature occurring in New Zealand in 2010. In this case series we present 28 cases of sudden death of individuals with evidence of MetHb and/or toxic salt ingestion, occurring in the Province of Ontario, Canada, between the years 1980 and 2020, inclusive. Of the 28 deaths in our case series, 25 showed evidence of intentional ingestion of sodium nitrite or sodium nitrate salts. Our year-over-year data demonstrated this is an increasingly used method of suicide in our provincial population, with the majority of cases occurring in the final two years of our study. Postmortem detection of MetHb is typically established via screening techniques such as scene evidence suggesting fatal consumption of a toxic salt in addition to the characteristic grey-purple lividity observed upon the body. The diagnosis can be established via postmortem blood testing demonstrating elevated methemoglobin saturation. Additionally, we have confirmed that postmortem MRI in cases of MetHb demonstrates a T1-bright (hyperintense) signal of the blood; both within intracardiac blood on chest MRIs and postmortem blood samples in tubes.

Optical mesoscopy, machine learning, and computational microscopy enable high information content diagnostic imaging of blood films.

Automated image-based assessment of blood films has tremendous potential to support clinical haematology within overstretched healthcare systems. To achieve this, efficient and reliable digital capture of the rich diagnostic information contained within a blood film is a critical first step. However, this is often challenging, and in many cases entirely unfeasible, with the microscopes typically used in haematology due to the fundamental trade-off between magnification and spatial resolution. To address this, we investigated three state-of-the-art approaches to microscopic imaging of blood films which leverage recent advances in optical and computational imaging and analysis to increase the information capture capacity of the optical microscope: optical mesoscopy, which uses a giant microscope objective (Mesolens) to enable high-resolution imaging at low magnification; Fourier ptychographic microscopy, a computational imaging method which relies on oblique illumination with a series of LEDs to capture high-resolution information; and deep neural networks which can be trained to increase the quality of low magnification, low resolution images. We compare and contrast the performance of these techniques for blood film imaging for the exemplar case of Giemsa-stained peripheral blood smears. Using computational image analysis and shape-based object classification, we demonstrate their use for automated analysis of red blood cell morphology and visualization and detection of small blood-borne parasites such as the malarial parasite Plasmodium falciparum. Our results demonstrate that these new methods greatly increase the information capturing capacity of the light microscope, with transformative potential for haematology and more generally across digital pathology. © 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.

Exploitation of blood non-Newtonian properties for ultrasonic measurement of hematocrit.

New processing techniques for manipulating blood and its components at a microfluidic scale are currently implemented. As for extracorporeal circulation, the in-line evaluation and monitoring of blood properties during these microfluidic techniques is a challenging task. Here, we show that the blood hematocrit can be measured non-invasively in a sub-millimeter medical tube using the non-Newtonian behavior of blood velocity profile. This hematocrit measurement is demonstrated on human blood with a simple Doppler ultrasound system. Results show a mean measurement error of 4.6 ± 1.3%Hct for hematocrit up to 52% and for 5 s-long ultrasonic signals. The simplicity and the measurement scale of the approach make it highly valuable for measuring hematocrit in new blood separation techniques. The approach may have an impact on in-vitro blood processing in general.

Random laser emission from whole blood as the active medium.

We introduce a proof of concept for multimode random laser (RL) emission with fresh whole blood (WB) as the active medium. The experimental principle is adapted from RL emission using rhodamine 6G (R6G). We achieved conditions for fresh WB to fluoresce with stochastic amplification of stimulated emission of radiation. The random conditions are placed with SiO2 particles, suspended in isotonic solvent. The results we report are for: (1) R6G-RL, with 2 nm bandwidth centered at 567 nm, and (2) RL emission for WB at 969 nm and 437 nm, of sub-3 nm bandwidth. For validation purposes, we show that the pumping energy threshold for RL emission from blood is consistent with R6G-RL.

Comparison of magnetization transfer-preparation and T2-preparation for dark-blood delayed-enhancement imaging.

Recently developed dark-blood techniques such as Flow-Independent Dark-blood DeLayed Enhancement (FIDDLE) allow simultaneous visualization of tissue contrast-enhancement and blood-pool suppression. Critical to FIDDLE is the magnetization preparation, which accentuates differences between myocardium and blood-pool. Here, we compared magnetization transfer (MT)-preparation and T2-preparation for use with FIDDLE. Variants of FIDDLE were developed with MT- or T2-preparation modules and tested in 35 patients (11 at 1.5 T, 24 at 3 T). Images were acquired with each FIDDLE variant in an interleaved fashion 10 minutes after gadolinium administration with otherwise identical acquisition parameters. Images were visually and quantitatively assessed for artifacts and differences in right ventricle to left ventricle (RV-to-LV) blood-pool suppression. Bright artifacts, reflecting incomplete blood-pool suppression, were frequently observed in the left atrium with T2-preparation FIDDLE at 1.5 and 3 T (82% and up to 100% of patients, respectively). MT-preparation FIDDLE resulted in fewer patients with artifacts (0% at 1.5 T, 22% at 3 T; P < .01). Left atrial blood-pool signal was significantly more homogeneous with MT-preparation than with T2-preparation at 1.5 and 3 T (P < .001 for all comparisons). Visibly different RV-to-LV blood-pool suppression was observed with T2-preparation in 36% of patients at 1.5 T and up to 94% at 3 T. In these patients, RV blood-pool signal was elevated, reducing the conspicuity of the myocardial-RV blood-pool border. Conversely, there were no visible differences in RV-to-LV blood-pool suppression with MT-preparation. Quantitative assessment of differences in blood-pool suppression and blood-pool artifacts was consistent with visual analyses. We conclude that for dark blood-blood delayed-enhancement imaging of the heart, MT-preparation results in fewer bright blood-pool artifacts and more uniform blood-pool suppression than T2-preparation.

Leukocyte segmentation in peripheral blood images using a novel edge strength cue-based location detection method.

The classification of leukocytes in peripheral blood images is an important milestone to be achieved because it can greatly assist pathologists to diagnose diseases such as leukemia, anemia, and other blood disorders. To a certain extent, a good segmentation method for identifying leukocytes from their background is the first step to the efficient functioning of the leukocytes classification system. However, the morphological structure of leukocytes, poor contrast, and the variations in their shape and size lead to the degradation of the segmentation accuracy. In this paper, we propose a new leukocyte segmentation framework that first locates and then segments leukocytes from peripheral blood images. Here, the locations of the leukocytes are first identified using a novel edge strength cue (ESc), and later, the Grabcut model is deployed to obtain the segmentation of the leukocytes. The novelty lies in the way the location of the leukocytes is detected, and this improves the leukocyte segmentation accuracy. The experimental evaluation is performed on ALL-IDB1, Cellavision, and LISC datasets for leukocyte segmentation based on the detection of the ESc location. Experimental results are evaluated using precision, recall, and F-score measures. The proposed method outperforms the state-of-the-art techniques. Additionally, the computation time of the proposed method is analyzed and presented in the study. Graphical Abstract Leukocytes Location Detection and Segmentation.

Non-Invasive MRI of Blood-Cerebrospinal Fluid Barrier Function.

The blood-cerebrospinal fluid barrier (BCSFB) is a highly dynamic transport interface that serves brain homeostasis. To date, however, understanding of its role in brain development and pathology has been hindered by the absence of a non-invasive technique for functional assessment. Here we describe a method for non-invasive measurement of BSCFB function by using tracer-free MRI to quantify rates of water delivery from arterial blood to ventricular cerebrospinal fluid. Using this method, we record a 36% decrease in BCSFB function in aged mice, compared to a 13% decrease in parenchymal blood flow, itself a leading candidate biomarker of early neurodegenerative processes. We then apply the method to explore the relationship between BCSFB function and ventricular morphology. Finally, we provide proof of application to the human brain. Our findings position the BCSFB as a promising new diagnostic and therapeutic target, the function of which can now be safely quantified using non-invasive MRI.

Automatic detection of blood content in capsule endoscopy images based on a deep convolutional neural network.

BACKGROUND AND AIM: Detecting blood content in the gastrointestinal tract is one of the crucial applications of capsule endoscopy (CE). The suspected blood indicator (SBI) is a conventional tool used to automatically tag images depicting possible bleeding in the reading system. We aim to develop a deep learning-based system to detect blood content in images and compare its performance with that of the SBI. METHODS: We trained a deep convolutional neural network (CNN) system, using 27 847 CE images (6503 images depicting blood content from 29 patients and 21 344 images of normal mucosa from 12 patients). We assessed its performance by calculating the area under the receiver operating characteristic curve (ROC-AUC) and its sensitivity, specificity, and accuracy, using an independent test set of 10 208 small-bowel images (208 images depicting blood content and 10 000 images of normal mucosa). The performance of the CNN was compared with that of the SBI, in individual image analysis, using the same test set. RESULTS: The AUC for the detection of blood content was 0.9998. The sensitivity, specificity, and accuracy of the CNN were 96.63%, 99.96%, and 99.89%, respectively, at a cut-off value of 0.5 for the probability score, which were significantly higher than those of the SBI (76.92%, 99.82%, and 99.35%, respectively). The trained CNN required 250 s to evaluate 10 208 test images. CONCLUSIONS: We developed and tested the CNN-based detection system for blood content in CE images. This system has the potential to outperform the SBI system, and the patient-level analyses on larger studies are required.

Assessing the Effect of Various Blood Glucose Levels on 18F-FDG Activity in the Brain, Liver, and Blood Pool.

Studies have extensively analyzed the effect of hyperglycemia on 18F-FDG uptake in normal tissues and tumors. In this study, we measured SUV in the brain, liver, and blood pool in normoglycemia, hyperglycemia, and hypoglycemia to understand the effect of blood glucose on 18F-FDG uptake and to develop a formula to correct SUV. Methods: Whole-body 18F-FDG PET/CT images of adults were selected for analysis. Brain SUVmax, blood-pool SUVmean, and liver SUVmean were measured at blood glucose ranges of 61-70, 71-80, 81-90, 91-100, 101-110, 111-120, 121-130, 131-140, 141-150, 151-160, 161-170, 171-180, 181-190, 191-200, and 201 mg/dL and above. At each blood glucose range, 10 PET images were analyzed (total, 150). The mean (±SD) SUV of the brain, liver, and blood pool at each blood glucose range was calculated, and blood glucose and SUV curves were generated. Because brain and tumors show a high expression of glucose transporters 1 and 3, we generated an SUV correction formula based on percentage reduction in brain SUVmax with increasing blood glucose level. Results: Mean brain SUVmax gradually decreased with increasing blood glucose level, starting after a level of 110 mg/dL. The approximate percentage reduction in brain SUVmax was 20%, 35%, 50%, 60%, and 65% at blood glucose ranges of 111-120, 121-140, 141-160, 161-200, and 201 mg/dL and above, respectively. In the formula we generated, measured SUVmax is multiplied by a reduction factor of 1.25, 1.5, 2, 2.5, and 2.8 for the blood glucose ranges of 111-120, 121-140, 141-160, 161-200, and 201 mg/dL and above, respectively, to correct SUV. Brain SUVmax did not differ between hypoglycemic and normoglycemic patients (P > 0.05). SUVmean in the blood pool and liver was lower in hypoglycemic patients (P < 0.05) and did not differ between hyperglycemic (P > 0.05) and normoglycemic patients. Conclusion: Hyperglycemia gradually reduces brain 18F-FDG uptake, starting after a blood glucose level of 110 mg/dL. Hyperglycemia does not affect 18F-FDG activity in the liver or blood pool. Hypoglycemia does not seem to affect brain 18F-FDG uptake but appears to reduce liver and blood-pool activity. The simple formula we generated can be used to correct SUV in hyperglycemic adults in selected cases.

Potential of indocyanine green near-infrared fluorescence imaging in experimental and clinical practice.

BACKGROUND: Visualization of Indocyanine Green (ICG) near-infrared (NIR) fluorescence is a promising technique for biomedical applications because of its deep tissue penetration, low autofluorescence, weak dependence on ambient light, safety for patients and medical staff. METHODS: To apply this technique to animal studies and clinical practice, we developed multispectral NIR fluorescence imaging system FLUM-808, which can be used in experimental studies and clinical practice, in open and endoscopic surgery (fluorescence-guided surgery). The object is illuminated either directly or through an illumination channel depending on what optical instrument is used (camera lens, fiber endoscope, rigid endoscope, surgical microscope, etc.). The system is able to simultaneously display NIR images and images captured under white light. RESULTS: This article provides examples of imaging techniques used during open and endoscopic surgery with ICG as an NIR fluorescent dye and photosensitizer, demonstrates applications of the technique for experimental and clinical purposes: for intraoperative imaging of blood and lymphatic vessels; for detection of tumors and sentinel lymph nodes; for assessment of tissue, organ, or anastomotic blood supply. CONCLUSIONS: This method significantly helps to accurately assign the areas of increased fluorescence to certain anatomical structures during surgery. NIR fluorescence diagnosis can be used in important experimental studies and in clinical practice. The described results of ICG studies can help in developing new fluorescence-based diagnostic and theranostic approaches employing more effective exogenous fluorophores and photosensitizers.

Role of T1 mapping as a complementary tool to T2* for non-invasive cardiac iron overload assessment.

BACKGROUND: Iron overload-related heart failure is the principal cause of death in transfusion dependent patients, including those with Thalassemia Major. Linking cardiac siderosis measured by T2* to therapy improves outcomes. T1 mapping can also measure iron; preliminary data suggests it may have higher sensitivity for iron, particularly for early overload (the conventional cut-point for no iron by T2* is 20ms, but this is believed insensitive). We compared T1 mapping to T2* in cardiac iron overload. METHODS: In a prospectively large single centre study of 138 Thalassemia Major patients and 32 healthy controls, we compared T1 mapping to dark blood and bright blood T2* acquired at 1.5T. Linear regression analysis was used to assess the association of T2* and T1. A "moving window" approach was taken to understand the strength of the association at different levels of iron overload. RESULTS: The relationship between T2* (here dark blood) and T1 is described by a log-log linear regression, which can be split in three different slopes: 1) T2* low, <20ms, r2 = 0.92; 2) T2* = 20-30ms, r2 = 0.48; 3) T2*>30ms, weak relationship. All subjects with T2*<20ms had low T1; among those with T2*>20ms, 38% had low T1 with most of the subjects in the T2* range 20-30ms having a low T1. CONCLUSIONS: In established cardiac iron overload, T1 and T2* are concordant. However, in the 20-30ms T2* range, T1 mapping appears to detect iron. These data support previous suggestions that T1 detects missed iron in 1 out of 3 subjects with normal T2*, and that T1 mapping is complementary to T2*. The clinical significance of a low T1 with normal T2* should be further investigated.

Characterization of the diffusion coefficient of blood.

PURPOSE: To characterize the diffusion coefficient of human blood for accurate results in intravoxel incoherent motion imaging. METHODS: Diffusion-weighted MRI of blood samples from 10 healthy volunteers was acquired with a single-shot echo-planar-imaging sequence at body temperature. Effects of gradient profile (monopolar or flow-compensated), diffusion time (40-100 ms), and echo time (60-200 ms) were investigated. RESULTS: Although measured apparent diffusion coefficients of blood were larger for flow-compensated than for monopolar gradients, no dependence of the apparent diffusion coefficient on the diffusion time was found. Large differences between individual samples were observed, with results ranging from 1.26 to 1.66 µm2 /ms for flow-compensated and 0.94 to 1.52 µm2 /ms for monopolar gradients. Statistical analysis indicates correlations of the flow-compensated apparent diffusion coefficient with hematocrit (P = 0.007) and hemoglobin (P = 0.017), but not with mean corpuscular volume (P = 0.64). Results of Monte-Carlo simulations support the experimental observations. CONCLUSIONS: Measured blood apparent diffusion coefficient values depend on hematocrit/hemoglobin concentration and applied gradient profile due to non-Gaussian diffusion. Because in vivo measurement is delicate, an estimation based on blood count results could be an alternative. For intravoxel incoherent motion modeling, the use of a blood self-diffusion constant Db = 1.54 ± 0.12 µm2 /ms for flow-compensated and Db = 1.30 ± 0.18 µm2 /ms for monopolar encoding is suggested. Magn Reson Med 79:2752-2758, 2018. © 2017 The Authors Magnetic Resonance in Medicine published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

The z-spectrum from human blood at 7T.

Chemical Exchange Saturation Transfer (CEST) has been used to assess healthy and pathological tissue in both animals and humans. However, the CEST signal from blood has not been fully assessed. This paper presents the CEST and nuclear Overhauser enhancement (NOE) signals detected in human blood measured via z-spectrum analysis. We assessed the effects of blood oxygenation levels, haematocrit, cell structure and pH upon the z-spectrum in ex vivo human blood for different saturation powers at 7T. The data were analysed using Lorentzian difference (LD) model fitting and AREX (to compensate for changes in T1), which have been successfully used to study CEST effects in vivo. Full Bloch-McConnell fitting was also performed to provide an initial estimate of exchange rates and transverse relaxation rates of the various pools. CEST and NOE signals were observed at 3.5 ppm, -1.7 ppm and -3.5 ppm and were found to originate primarily from the red blood cells (RBCs), although the amide proton transfer (APT) CEST effect, and NOEs showed no dependence upon oxygenation levels. Upon lysing, the APT and NOE signals fell significantly. Different pH levels in blood resulted in changes in both the APT and NOE (at -3.5 ppm), which suggests that this NOE signal is in part an exchange relayed process. These results will be important for assessing in vivo z-spectra.

Improved dark blood late gadolinium enhancement (DB-LGE) imaging using an optimized joint inversion preparation and T2 magnetization preparation.

PURPOSE: To develop a dark blood-late gadolinium enhancement (DB-LGE) sequence that improves scar-blood contrast and delineation of scar region. METHODS: The DB-LGE sequence uses an inversion pulse followed by T2 magnetization preparation to suppress blood and normal myocardium. Time delays inserted after preparation pulses and T2 -magnetization-prep duration are used to adjust tissue contrast. Selection of these parameters was optimized using numerical simulations and phantom experiments. We evaluated DB-LGE in 9 swine and 42 patients (56 ± 14 years, 33 male). Improvement in scar-blood contrast and overall image quality was subjectively evaluated by two independent readers (1 = poor, 4 = excellent). The signal ratios among scar, blood, and myocardium were compared. RESULTS: Simulations and phantom studies demonstrated that simultaneous nulling of myocardium and blood can be achieved by selecting appropriate timing parameters. The scar-blood contrast score was significantly higher for DB-LGE (P < 0.001) with no significant difference in overall image quality (P > 0.05). Scar-blood signal ratios for DB-LGE versus LGE were 5.0 ± 2.8 versus 1.5 ± 0.5 (P < 0.001) for patients, and 2.2 ± 0.7 versus 1.0 ± 0.4 (P = 0.0023) for animals. Scar-myocardium signal ratios were 5.7 ± 2.9 versus 6.3 ± 2.6 (P = 0.35) for patients, and 3.7 ± 1.1 versus 4.1 ± 2.0 (P = 0.60) for swine. CONCLUSIONS: The DB-LGE sequence simultaneously reduces normal myocardium and blood signal intensity, thereby enhancing scar-blood contrast while preserving scar-myocardium contrast. Magn Reson Med 79:351-360, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Three-dimensional computational model of a blood oxygenator reconstructed from micro-CT scans.

Cardiopulmonary bypass procedures are one of the most common operations and blood oxygenators are the centre piece for the heart-lung machines. Blood oxygenators have been tested as entire devices but intricate details on the flow field inside the oxygenators remain unknown. In this study, a novel method is presented to analyse the flow field inside oxygenators based on micro Computed Tomography (µCT) scans. Two Hollow Fibre Membrane (HFM) oxygenator prototypes were scanned and three-dimensional full scale models that capture the device-specific fibre distributions are set up for computational fluid dynamics analysis. The blood flow through the oxygenator is modelled as a non-Newtonian fluid. The results were compared against the flow solution through an ideal fibre distribution and show the importance of a uniform distribution of fibres and that the oxygenators analysed are not susceptible to flow directionality as mass flow versus area remain the same. However the pressure drop across the oxygenator is dependent on flow rate and direction. By comparing residence time of blood against the time frame to fully saturate blood with oxygen we highlight the potential of this method as design optimisation tool. In conclusion, image-based reconstruction is found to be a feasible route to assess oxygenator performance through flow modelling. It offers the possibility to review a product as manufactured rather than as designed, which is a valuable insight as a precursor to the approval processes. Finally, the flow analysis presented may be extended, at computational cost, to include species transport in further studies.