Engineering Conferences International

Berruti, Franco

Dufour, Anthony

Garcia-Pérez, Manuel

Prins, Wolter

English

Conference program - Pyroliq II: Pyrolysis and Liquefaction of Biomass and Wastes

Please click Additional Files below to see the full abstrac

Semple, Melissa

Clapp, Ken

Brocato, Nicholas

Jayakar, Radhika

NEXT GENERATION PROCESSES: SINGLE USE TECHNOLOGY AND KEY ENABLERS 
 
Andreas Castan, Cytiva, Björkgatan 30, 751 84 Uppsala, Sweden 
andreas.castan@cytiva.com 
Ken Clapp, Cytiva, 100 Results Way, Marlborough, MA 01752, USA 
Nicholas Brocato, Cytiva, 100 Results Way, Marlborough, MA 01752, USA 
Radhika Jayakar, Cytiva, 100 Results Way, Marlborough, MA 01752, USA 
 
 
Key Words: process intensification, perfusion, cell retention, next generation bioreactors, product development 
 
The biopharma industry faces significant challenges. Among these are increased cost pressure on medicines 
from society and biosimilar manufacturers, the pressure to reduce the time to approval as well as more targeted 
patient populations resulting in smaller batch sizes and use of multi-product facilities. 
 
This pressure forces companies to develop the next generation processes, i.e. more intensified processes with 
higher productivities, smaller footprint in agile plants and at a lower cost. Single-use technology is one of the key 
enablers for this next generation of intensified processes. 
 
This presentation describes different strategies for next generation processing, where perfusion is an integral 
part of all these intensified upstream processes. These include both the seed train and the production 
bioreactor. We highlight, how recent technology developments support these strategies, especially with respect 
to high performing bioreactors and cell retention devices. Furthermore, we describe product development 
strategies including in-silico design like CFD, extensive verification and characterization. In addition, process 
and performance data will be presented for each strategy together with an estimation of the impact on plant 
productivity and process economy. The results provide useful insights for biomanufacturers who are considering 
the value of process intensification and single-use technology to address some of the challenges of the industry. 
 


Next generation processes: Single use technology and key enablers

Mahajan, Ekta

Strauss, Ulrike

MATERIAL CONSIDERATIONS FOR CELL & GENE THERAPY PRODUCTS 
 
Ekta Mahajan, Material Science and Strategy, Genentech, USA 
EKTAM@GENE.COM 
Ulrike Strauss, Material Science and Strategy, Genentech, USA 
 
 
Cell & Gene Therapies use novel critical materials where we have limited knowledge and face challenges in 
development and manufacturing due to their variability and unknown quality attributes. As the field is fairly new, 
we are facing challenges with research grade materials which are not well understood, and new suppliers who 
are not entirely familiar with requirements for pharmaceutical development and manufacturing. Material and 
supplier selection decisions early in the development of ATMPs are crucial and can impact product quality and 
patient supply. The measures taken to establish appropriate controls at early stages enable use of the same raw 
materials in all phases of the product development reducing comparability and/or regulatory risks. Many of the 
materials used for ATMP production are manufactured by suppliers that have limited experience with GMP 
biopharmaceutical manufacturing.  As a result, there can be misalignment between supplier capabilities and 
ATMP developer expectations regarding technical, supply, quality, and regulatory compliance requirements for 
biomanufacturing.   
 
This talk will address the considerations for a science and risk-based approach to material selection and 
qualification, including considerations for aligning internally and externally on definitions and implications of 
material classification and material grades.  
 
 
 
 


Material considerations for cell & gene therapy products

Rusk, Dave

NEEDS AND PATH FORWARD FOR EAC 
 
Dave Rusk, Siddiq Qidwai, Earl Johns 
 


Welsh

Needs and path forward for EAC

Please click Additional Files below to see the full abstract

Puglia, John

Cobia, Jordan

PROCESS SUSTAINABILITY THROUGH NEW TECHNOLOGIES: SINGLE-USE CENTRIFUGE 
IMPLICATIONS ON PROCESS AND PRODUCT EFFICIENCY  
 
Mark Thomas Smith PhD, Thermo Fisher Scientific 
Mark.smith3@thermofisher.com  
Jordan Cobia, Thermo Fisher Scientific 
 
 
Key Words: Harvest, Sustainability, Centrifugation, Single-use, Closed-systems 
 
Background: 
Single-use technologies (SUTs) have enabled flexible, modular, and often closed-system processes for process 
development and pilot/clinical scale. More recently, SUTs have made their way into commercial production for 
the ability to scale-out at pace with the products demands without substantial sunk-cost risk in capital 
traditionally associated with stainless facilities.  
Each piece of the bioproduction workflow has had SUTs emerge over time and become established to varying 
degrees and technical excellent, from upstream, harvest/separations, downstream, formulation, and eventually 
fill/finish.  
 
Harvest – the process of preliminary clarification and separations of solids and debris from supernatant typically 
containing product – traditionally was accomplished in stainless facilities with continuous-flow stainless steel 
centrifuges and stainless SIP depth filter (DF) housings. For SU-facilities, the SU-centrifuges were not available 
and expanded SU-DF became a standard. Cell culture processes have intensified, with more solids and higher 
volumes in SU formats than ever before, driving expanding count of SU-DF required for harvest. 
 
Harvest Sustainability through SU-Centrifugation: 
As processes intensify in SU-formats and SU-DF cartridge count rises per batch, the solids and liquid wastes 
generated by SU-DF can exceed 400 kg and thousands of liters of buffer and caustics. In an efficient facility, a 
single bioreactor process suite could generate 6,000 to 8,000 kgs of SU-DF solid waste annually.  
New SU-centrifuges are coming available that mimic the benefits of the traditional stainless continuous flow 
centrifuges, reducing required filtration areas and enabling large-scale processing, with additional benefits of 
single-use, closed-system processing – are a promising step to make intensifying harvest processes more 
sustainable. Through empirical testing and bioeconomic modelling, we evaluated the potential impact of SU-
centrifuge on the ability to provide a more sustainable single-use harvest solution, projecting a 50-70% reduction 
in solid’s waste and >45% reduction in liquid’s waste using a SU-centrifuge. In addition, we evaluated the impact 
on of the change on labor and suite time, shipping and warehousing, and comparative cost of operation.  
Overall, the (re)introduction of (single-use)centrifuges into the bioprocess promises to enable intensified 
bioprocesses that are simultaneously more sustainable that traditional single-use processes. 
 
 
 
Figure 1 – Example reduction of solids waste of per batch from use of single-use depth filtration vs single-use 
centrifugation and filtration across different reactor batch volumes.  


Process sustainability through new technologies: Single-use centrifuge implications on process and product efficiency

Else, Thomas R.

Hacker, Lina

Bunce, Ellie V.

Tao, Ran

Bohndiek, Sarah E.

Gröhl, Janek

 
 
ADDRESSING THE EFFECTS OF SKIN TONE ON PHOTOACOUSTIC IMAGING 
 
Thomas R. Else, University of Cambridge, United Kingdom 
thomas.else@cruk.cam.ac.uk 
Lina Hacker, University of Cambridge, United Kingdom 
Janek Gröhl, University of Cambridge, United Kingdom 
Ellie V. Bunce, University of Cambridge, United Kingdom 
Ran Tao, University of Cambridge, United Kingdom 
Sarah E. Bohndiek, University of Cambridge, United Kingdom 
 
Keywords: photoacoustic imaging, skin tone, racial bias, optoacoustic, quantification, simulation, phantom 
 
Optical sensing and imaging technologies are known to be 
sensitive to melanin in the epidermis, limiting their 
effectiveness in people with darker skin tones. Biases have 
been observed in several techniques where light passes 
through the skin. Pulse oximeters, bilirubinometers and 
wearable technologies have all been affected, with worse 
outcomes for people with darker skin (1–3). Early studies in 
photoacoustic imaging (PAI) have identified similar biases 
(4, 5). Tackling this issue is crucial for ensuring 
standardization, reproducibility, and equity in clinical PAI.  
We assessed the effects of skin tone on PAI in a controlled 
setting using computational skin models with varying blood 
oxygenation and melanin concentration (Fitzpatrick scale, 
logarithmically from 1% to 25% melanosome volume 
fraction, Figure 1A). We obtained diffuse reflectance values 
consistent with typical variation in the population (0.20 to 
0.72 at 685 nm, Figure 1B). Light transport was simulated 
using a Monte-C arlo model. We observed a linear 
relationship between melanosome volume fraction and 
linear-unmixed sO2 (R2 = 0.998, p < 0.001, gradient = 0.500 
± 0.014); this trend was reduced using a gradient-boosted 
regressor trained on independently simulated data (R2 = 
0.026, p = 0.76, gradient = 0.017 ± 0.051) (Figure 1C).  
Consistent results were observed in skin-mimicking blood 
oxygenation phantoms. Agarose phantoms with two layers 
were constructed with an outer 1 mm layer containing 
synthetic melanin. Standard spectral unmixing and learned 
spectral unmixing were applied and compared to pO2 probe 
measurements. We observed an increase in linear-unmixed 
sO2 with melanin concentration (p < 0.001, gradient = 0.162 
mL/mg). sO2 quantification was improved using the learned 
unmixing approach, removing the dependence on melanin concentration (p = 0.50, gradient = 0.051 mL/mg) 
(Figure 1D). 
 
Finally, we demonstrated the impact of skin pigmentation on PAI in vivo using mouse models. We compared an 
albino mouse (C57BL/6 albino) strain to pigmented black 6 (C67BL/6, B6) mice. Albino mice were imaged once 
and B6 mice twice to allow pigmentation to develop. We observed a strong correlation between skin 
photoacoustic signal and linear-unmixed sO2 in the body of the mice (p < 0.001) (Figure 1E).  
PAI demonstrates a clear measurement bias due to melanin in the skin. Reproducible and equitable application 
of quantitative PAI will likely require a combination of physics-based, data-driven, and empirical methods to 
account for varying light fluence and tissue properties fairly across the population. 
 
1.  M. W. Sjoding, R. P. Dickson, T. J. Iwashyna, S. E. Gay, T. S. Valley, N. Engl. J. Med. 383, 2477–2478 (2020). 
2.  L. Holland, K. Blick, Am. J. Clin. Pathol. 132, 555–561 (2009). 
3.  P. J. Colvonen, P. N. DeYoung, N. O. A. Bosompra, R. L. Owens, Sleep. 43, 1–3 (2020). 
4.  Y. Mantri, J. V. Jokerst, Biomed. Opt. Express. 13, 875 (2022). 
5.  X. Li et al., J. Biophotonics. 12 (2019), doi:10.1002/jbio.201800442. 
 
  Figure 1: Melanin in the skin substantially affects 
quantitative photoacoustic imaging. RGB rendering of 
diffuse reflectance simulations are assigned to the 
Fitzpatrick scale (A). Diffuse reflectance spectra for 
each model (B). Unmixed sO2 depends strongly on 
melanin concentration and is improved with a learned 
unmixing approach, shown in simulations (C) and 
phantoms (D). Linear sO2 estimates increase with skin 
photoacoustic signal in pigmented (pig.) mice (E). 
Poster 2 
 
Poster 3 
 
 
 


Addressing the effects of skin tone on photoacoustic imaging

Lin, Charles P.

Monday, October 16, 2023  Session 3: Microscopic Imaging in Biomedicine I 
CELL TRACKING IN VIVO: FROM HEMATOPOIETIC STEM CELLS TO MATURE LEUKOCYTES 
 
Charles P. Lin, PhD, Wellman Center for Photomedicine, Massachusetts General Hospital 
charles_lin@hms.harvard.edu 
 
 
Key Words: Intravital microscopy, cell tracking, Image-seq 
 
Hematopoiesis is the process of blood cell production that takes place primarily in the bone marrow. Upstream 
of this process sit the primitive hematopoietic stem cells (HSCs) from which all blood cells are derived, and far 
downstream are the mature blood cells, including leukocytes (white blood cells) that are released into the 
circulation to carry out immune functions throughout the body. We have developed intravital microscopy (IVM) 
techniques for tracking individual HSCs and characterizing their local microenvironment in the bone marrow. We 
have further developed a new method, called Image-seq (image-guided single cell RNA sequencing), that 
enables transcriptomic profiling of the exact same cells whose location and division history have been 
documented by IVM. This method therefore integrates spatial, temporal, and molecular information captured 
with the same optical platform. Downstream, we are developing methods for tracking leukocytes in the 
microcirculation and envision a path toward clinical translation by label-free imaging based on oblique back-
illumination microscopy (OBM). We present a miniaturized OBM instrument for noninvasive assessment of host 
(human) immune response to infection and inflammation.
Monday, October 16, 2023  Session 3: Microscopic Imaging in Biomedicine I 
 


Cell tracking in vivo: From hematopoietic stem cells to mature leukocytes

Niedre, Mark

TOWARDS FLUORESCENCE MOLECULAR IN VIVO LIQUID BIOPSY OF CIRCULATING TUMOR CELLS 
 
Mark Niedre, Ph.D., Department of Bioengineering, Northeastern University, Boston, MA, USA 
m.niedre@northeastern.edu 
 
 
In this presentation I will discuss our recent work in development of ‘Diffuse in vivo Flow Cytometry’ (DiFC) for 
non-invasive fluorescence enumeration of circulating tumor cells (CTCs) without drawing blood samples and 
describe progress towards translation of DiFC to humans. Briefly, DiFC is an emerging technique wherein highly-
scattered light is used to non-invasively sample blood flowing in large deep-seated blood vessels. Specially 
designed optical probes and algorithmic signal processing allows detection rare fluorescently-labeled cells and 
multi-cellular clusters in the bloodstream. We have previously used DiFC to study CTC dissemination during tumor 
development in mouse models of metastasis. 
 
In this presentation I will describe our recent progress in application of near-infrared fluorescence molecular 
contrast agents for sensitive and specific labeling of CTCs in vivo. Candidate contrast agents include a folate 
receptor-targeted probe (OTL38, Cytalux), as well as new, purpose designed pan-epithelial targeted contrast 
agents. I will also discuss the relevant tissue optics and instrumentation considerations for potential future 
human translation. Ultimately, DiFC could represent a new method for continuously enumerating CTCs without 
drawing blood samples that could enable early detection of cancer metastasis or monitoring of response to 
therapies. 
 


Towards fluorescence molecular in vivo liquid biopsy of circulating tumor cells

Boppart, Stephen

REAL-TIME LABEL-FREE AI-ENHANCED DIGITAL HISTOPATHOLOGY USING SLAM MICROSCOPY  
 
Stephen A. Boppart, University of Illinois Urbana-Champaign 
boppart@illinois.edu 
 
 
Key Words: Nonlinear optics, Multiphoton microscopy, AI Machine Learning, Cancer, Alzheimer’s disease. 
 
Innovations in biomedical imaging have historically led to discoveries in the life sciences and new detection and 
diagnostic technologies in medicine and surgery. Label-free intravital optical imaging and imaging of fresh, 
unstained, resected tissue specimens, offers a wealth of new biosignatures for revealing the true colors of 
cancer and diagnosing neurodegenerative changes. Using new optical source technology and nonlinear optics 
to generate new excitation wavelengths and manipulate the light stimulus in new ways, Simultaneous Label-free 
Auto-fluorescence Multi-harmonic (SLAM) microscopy can achieve fast simultaneous visualization of the rich 
intrinsic molecular and metabolic features within tissues. Quantitative AI and machine/deep learning analyses of 
these multi-dimensional datasets can be used to identify selective diagnostic biosignatures for disease.  
Specifically, extracellular vesicles (EVs) in carcinogenesis and amyloid-beta plaque composition in Alzheimer’s 
disease were analyzed via their label-free optical signatures and spatial distributions. Utilizing the high 
dimensionality of the spatially and temporally co-registered multimodal image datasets, cell segmentation, cell 
type identification, and the spatiotemporal biological context of cells and microfeatures within the tissue and 
tumor microenvironments were used for deep-profiling of the tissue for AI-enhanced digital histopathology. 
Further image analysis showed that EVs from the tumor microenvironment and variations in amyloid-beta 
plaques have unique optical signatures, in comparison to those vesicles and tissues from healthy subjects. 
Initial clinical demonstrations using these optical biomedical imaging technologies suggest new paradigms for 
real-time point-of-procedure digital histopathology, AI-informed diagnosis, and image-guidance for medical or 
surgical decision making.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
References: 
Tu H, et.al., Nature Photonics, 10:534-540, 2016; You S, et.al., Nature Comm., 9:2125, 2018; You S, et.al., 
Proc. Natl. Acad. Sci., USA, 116:24012-24018, 2019; You S, et.al., Cancer Res., 81:2534-2544, 2021; Shi J, 
et.al., Biomed. Opt. Express, 14:1339-1354, 2023. 
Figure – SLAM microscopy (left) of in-vivo healthy and cancer tissues. Using the high-dimensional 
multimodal datasets to train AI models, multi-class segmentation masks for multiple cell types and 
biological features can be generated, and subsequently applied for clinical diagnostic assessment. 


Real-time label-free AI-enhanced digital histopathology using SLAM microscopy

Sundararaghavan, Veera

MICROSTRUCTURECRACK PATH PREDICTION USING GRAPH THEORY  
 
Veera Sundararaghavan, University of Michigan Ann Arbor 
veeras@umich.edu 
 
 
Key Words: high cycle fatigue, in-situ loading, cohesive energy, grain boundary energy, Griffith theory. 
 
The crack path in crystalline metals is determined by the relative strength and orientation of their multiple 
cleavage directions with respect to the loading direction. The strength of the cleavage surfaces depend on the 
local chemical environment and crystallographic orientation of the interfaces. Understanding the cohesive 
energy in a polycrystalline metal is fundamental to predicting its susceptibility to failure mechanisms such as 
cracking and fatigue. In a recent work, we studied a geometric method based on Wulff's diagram of inverse 
cohesive energy to predict fracture [1]. We find the existence of forbidden regions where cracks cannot 
propagate, and often the presence of multiple optimal crack propagation angles in which case the crack path's 
local orientation switches rapidly to follow the microscopically preferred path. This behavior results in saw-tooth 
patterns on the fatigue cracks. We introduce a novel approach based on multiscale graph theory to predict such 
paths of microscale cracks in polycrystalline materials. We represent the crack path as the boundary of a 
partition in a geometric graph, obtained by optimizing an Ising-type hamiltonian. The hamiltonian parameters are 
chosen so that the partition cost equals the energy of the corresponding crack. The growth angle of the crack in 
polycrystalline materials is influenced by the loading conditions and the microstructure near the crack tip. To 
account for this, we incorporate two length scales, macro and micro, and define the total energy of the crack as 
the sum of macroscopic energy release and microscopic surface energy. The former guides the crack to 
propagate along the direction of maximum energy release, while the latter directs the crack along the 
macroscopically preferred direction. The crack path accommodates both intergranular and transgranular 
fractures, with the former occurring along grain boundaries and the latter along crystallographic cleavage 
planes. We study mixed-mode fracture in a thin foil specimen and include the dihedral angle of the 2D crack to 
define effective surface energy. We validate our model using analytical results for mixed-mode fracture in an 
isotropic medium and mode-I fracture in a medium with a preferred crack direction. Finally, we compare our 
results against thin foil experiments on a HCP Magnesium alloy. Overall, our approach can aid in designing 
materials with optimal fracture resistance by controlling their microstructure. 
 
 
 
Figure.1 (a) Wulff plots show the inverse of surface energy of two planar crystals of differing orientation (30 and 
10 degrees). The gray areas indicate forbidden regions of fracture. In the first case, the crack is predicted to 
follow a zig zag path at two angles. (b) Experimentally observed crack path in a foil specimen. (c) measured 
growth rate of the crack indicating pauses at the grain boundary (d) Crack path predicted by the graph model 
matches experimental data 
 
 
Reference [1] Srivastava, S., Yaghoobi, M. and Sundararaghavan, V., 2021. A graph-theoretic approach for 
multiscale modeling and prediction of crack propagation in polycrystalline materials. Engineering Fracture 
Mechanics, 241, p.107406. 


Not a member yet