Predicting cell behaviour parameters from glioblastoma on a chip images. A deep learning approach

Pérez-Aliacar, Marina; Doweidar, Mohamed H.; Doblaré, M.; Ayensa-Jiménez, Jacobo

doi:10.1016/j.compbiomed.2021.104547

Cited by 18 publications

(22 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In particular, GBM culture evolution of the cell line U251-MG in microfluidic devices has been well described, even for different experimental configurations using these expressions [ 28 ]. Similar results were obtained now using Machine Learning tools (in particular, using Convolutional Neural Networks), also revealing some limitations of the parametric model [ 29 ].…”

Section: Methodssupporting

confidence: 75%

“…Recently, we have been able to reproduce these histological structures in vitro [ 26 , 27 ]. Also, we developed a mathematical model incorporating the go or grow hypothesis, which allowed us to reproduce the GBM evolution under different experimental configurations also in vitro [ 28 ], and to derive, from cell culture images, information on the cell behavior [ 29 ].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Understanding glioblastoma invasion using physically-guided neural networks with internal variables

et al. 2022

Self Cite

View full text Add to dashboard Cite

Microfluidic capacities for both recreating and monitoring cell cultures have opened the door to the use of Data Science and Machine Learning tools for understanding and simulating tumor evolution under controlled conditions. In this work, we show how these techniques could be applied to study Glioblastoma, the deadliest and most frequent primary brain tumor. In particular, we study Glioblastoma invasion using the recent concept of Physically-Guided Neural Networks with Internal Variables (PGNNIV), able to combine data obtained from microfluidic devices and some physical knowledge governing the tumor evolution. The physics is introduced by means of nonlinear advection-diffusion-reaction partial differential equation that models the Glioblastoma evolution for defining the network structure. On the other hand, multilayer perceptrons combined with a nodal deconvolution technique are used for learning the go or grow metabolic behavior which characterises the Glioblastoma invasion. The PGNNIV is here trained using synthetic data obtained from in silico tests created under different oxygenation conditions, using a previously validated model. The unravelling capacity of PGNNIV enables discovering complex metabolic processes in a non-parametric way, thus giving explanatory capacity to the networks, and, as a consequence, surpassing the predictive power of any parametric approach and for any kind of stimulus. Besides, the possibility of working, for a particular tumor, with different boundary and initial conditions, permits the use of PGNNIV for defining virtual therapies and for drug design, thus making the first steps towards in silico personalised medicine.

show abstract

Section: Methodssupporting

confidence: 75%

Section: Introductionmentioning

confidence: 99%

Understanding glioblastoma invasion using physically-guided neural networks with internal variables

et al. 2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…As an example, one may employ robotics to automate tasks (i.e., chip operation and data collection) [249] and machine learning (ML) to facilitate data analysis. [25,250] The throughput of OoCs is determined by not only the number of tests per run but also the number of readouts per test. Most traditional methods for molecular and functional characterization of BBB-on-a-chip such as reverse transcriptionpolymerase chain reaction and permeability assays via tracer molecules are invasive, not real-time, and require laborious [227] Copyright 2020, eLife.…”

Section: Increasing Throughput and Scalabilitymentioning

confidence: 99%

“…As an example, one may employ robotics to automate tasks (i.e., chip operation and data collection) [ 249 ] and machine learning (ML) to facilitate data analysis. [ 25,250 ]…”

Section: Current Challenges and Future Directionsmentioning

confidence: 99%

Organ‐On‐A‐Chip Models of the Blood–Brain Barrier: Recent Advances and Future Prospects

Kawakita

Mandal

Mou

et al. 2022

Small

View full text Add to dashboard Cite

refers to a multicellular unit in the brain including cells of the cerebral vasculature and brain parenchyma. [2] As part of the NVU, the blood-brain barrier (BBB) acts as a physiological barrier at the interface between peripheral blood circulation and the central nervous system (CNS). The BBB is indispensable for the proper CNS function and regulation of CNS homeostasis; it controls the transport of substances between the CNS and the blood via various transport mechanisms. [3] The BBB also protects the brain from neurotoxic plasma components, certain chemicals, and pathogens. Functional and structural changes of the BBB are implicated in several neurological diseases and disorders, which have severely impacted individuals worldwide. For example, there are currently estimated to be about 5.7 million people living with Alzheimer's disease (AD), and the number is projected to reach 13.8 million by 2050. [4] Stroke is the second leading cause of death and the third most common cause of disability. [5] Glioblastoma multiforme (GBM), which accounts for almost 50% of all brain tumors (i.e., glioma), is relatively rare with a global incidence rate of 3.19 per 100 000 people; however, it has a devastatingly poor prognosis with a typical survival rate of 14-15 months rendering it a critical public health problem. [6] The human brain and central nervous system (CNS) present unique challenges in drug development for neurological diseases. One major obstacle is the blood-brain barrier (BBB), which hampers the effective delivery of therapeutic molecules into the brain while protecting it from blood-born neurotoxic substances and maintaining CNS homeostasis. For BBB research, traditional in vitro models rely upon Petri dishes or Transwell systems. However, these static models lack essential microenvironmental factors such as shear stress and proper cell-cell interactions. To this end, organ-on-a-chip (OoC) technology has emerged as a new in vitro modeling approach to better recapitulate the highly dynamic in vivo human brain microenvironment socalled the neural vascular unit (NVU). Such BBB-on-a-chip models have made substantial progress over the last decade, and concurrently there has been increasing interest in modeling various neurological diseases such as Alzheimer's disease and Parkinson's disease using OoC technology. In addition, with recent advances in other scientific technologies, several new opportunities to improve the BBB-on-a-chip platform via multidisciplinary approaches are available. In this review, an overview of the NVU and OoC technology is provided, recent progress and applications of BBB-on-a-chip for personalized medicine and drug discovery are discussed, and current challenges and future directions are delineated.The ORCID identification number(s) for the author(s) of this article can be found under

show abstract

“…Se ha desarrollado una red neuronal convolucional para predecir, a partir de una imagen del cultivo celular en un dispositivo microfluidico, el valor de los tres parámetros 𝑮 𝑵 , 𝑲 𝑵 , 𝑯, caracterizando así el comportamiento del cultivo [7]. La red se entrenó previamente con datos sintéticos provenientes de un modelo matemático previamente validado con experimentos [6].…”

Section: Metodologíaunclassified

Microfluídica inteligente: combinación del aprendizaje profundo y la microfluídica para predecir la evolución del glioblastoma.

Pérez-Aliacar¹,

Doweidar

Doblaré³

et al. 2021

Jorn. jovenes investig. I3A

Self Cite

View full text Add to dashboard Cite

Las numerosas posibilidades que ofrece la microfluídica a nivel de obtención y monitorización de grandes cantidades de datos abren la puerta a la aplicación de técnicas de aprendizaje profundo al análisis de la evolución cultivos celulares, un campo poco explorado hasta la fecha. En este trabajo, se desarrolla una red neuronal para identificar, a partir de imágenes de cultivos celulares de glioblastoma en dispositivos microfluidicos, los parámetros que gobiernan el comportamiento de este tumor cerebral.

show abstract

Predicting cell behaviour parameters from glioblastoma on a chip images. A deep learning approach

Cited by 18 publications

References 55 publications

Understanding glioblastoma invasion using physically-guided neural networks with internal variables

Understanding glioblastoma invasion using physically-guided neural networks with internal variables

Organ‐On‐A‐Chip Models of the Blood–Brain Barrier: Recent Advances and Future Prospects

Microfluídica inteligente: combinación del aprendizaje profundo y la microfluídica para predecir la evolución del glioblastoma.

Contact Info

Product

Resources

About