SA
Sepinoud Azimi
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
29 records found
1
Journal article
(2026)
-
Laura R. Jansén-Storbacka, Kailas S. Honasoge, Sepinoud Azimi, Kateřina Staňková, Eva Molnárová, Arina Soboleva, Bram C. Agema, Dirk Jan A.R. Moes, G. D.Marijn Veerman, Alethea B.T. Barbaro, Roel Dobbe, Irene Grossmann
Evolutionary therapy (ET) applies principles of evolutionary biology to steer tumour dynamics and forestall or delay treatment resistance, typically guided by data-driven mathematical models. Our aim is to assess whether ET protocols, and specifically Zhang et al.’s protocol proposed for metastatic castrate-resistant prostate cancer, can be theoretically effective for fast-growing metastatic cancers such as stage IV non-small-cell lung cancer (NSCLC). Using longitudinal tumour-burden data from NSCLC patients treated with erlotinib, we systematically evaluate 26 two-population differential-equation models based on classical tumour-growth dynamics, with varying assumptions about density- and frequency-dependent interactions, pharmacokinetics, and treatment-induced death. Previous work by Yin et al. on the same dataset employed an exponential model that omitted density- and frequency-dependent interactions; although it provided a good fit to tumour-burden data, its structure would theoretically lead to poorer outcomes under ET protocols. In contrast, our analysis identifies the minimal model structure required to reproduce the resistance-driven regrowth observed in NSCLC, with the Gompertzian model featuring log-kill dynamics and both density- and frequency-dependent interactions providing the best fit. In this model, Zhang et al.’s protocol prolonged median time-to-progression to 42.3 months compared with 24.8 months under maximum tolerated dose. These results indicate that ET is theoretically a viable treatment strategy for NSCLC. This study offers a practical framework for assessing ET feasibility using clinical data and supports future clinical translation of ET in NSCLC.
...
Evolutionary therapy (ET) applies principles of evolutionary biology to steer tumour dynamics and forestall or delay treatment resistance, typically guided by data-driven mathematical models. Our aim is to assess whether ET protocols, and specifically Zhang et al.’s protocol proposed for metastatic castrate-resistant prostate cancer, can be theoretically effective for fast-growing metastatic cancers such as stage IV non-small-cell lung cancer (NSCLC). Using longitudinal tumour-burden data from NSCLC patients treated with erlotinib, we systematically evaluate 26 two-population differential-equation models based on classical tumour-growth dynamics, with varying assumptions about density- and frequency-dependent interactions, pharmacokinetics, and treatment-induced death. Previous work by Yin et al. on the same dataset employed an exponential model that omitted density- and frequency-dependent interactions; although it provided a good fit to tumour-burden data, its structure would theoretically lead to poorer outcomes under ET protocols. In contrast, our analysis identifies the minimal model structure required to reproduce the resistance-driven regrowth observed in NSCLC, with the Gompertzian model featuring log-kill dynamics and both density- and frequency-dependent interactions providing the best fit. In this model, Zhang et al.’s protocol prolonged median time-to-progression to 42.3 months compared with 24.8 months under maximum tolerated dose. These results indicate that ET is theoretically a viable treatment strategy for NSCLC. This study offers a practical framework for assessing ET feasibility using clinical data and supports future clinical translation of ET in NSCLC.
Journal article
(2026)
-
Saad Abdullah, Md Masum Billah, Victor Armando Canales-Lima, Pragati Manandhar, Lameya Islam, Alexis Gbeckor-Kove, Sarosh Krishan, Hergys Rexha, Sepinoud Azimi, More Authors
The black-box nature of deep learning (DL) models presents a significant challenge for their adoption in clinical settings. The field of explainable artificial intelligence (XAI) has emerged to improve the transparency and interpretability of models. However, current techniques do not adequately describe the reasoning underpinning DL models. This study replicates and extends previous research on the use of texture analysis to improve interpretability in clinically geared segmentation tasks. We evaluate Law's Texture Energy Measures (LTEMs) in the learning and decision-making processes of different DL architectures. We extend the work to include breast cancer, skin lesion, and gastrointestinal polyp datasets, as well as CLAHE-enhanced datasets to identify any divergence in learning. Experimental results reiterate that LTEMs, specifically level-edge convolution masks, are highly influential across multiple DL architectures. Additionally, Gray-Level Co-occurrence Matrix (GLCM) analysis highlights autocorrelation as a key descriptor. The results confirm that texture-based representations, learned primarily in the early layers of the network, are sufficient for robust learning. Through LTEMs, we can characterize the patterns learned in DL and associate these patterns with verbal descriptions and clinically objective measures, thus translating the DL learning into human terms. This psychophysical approach eases the clinical interpretability of DL models. Code availability: https://github.com/xrai-lib/xai-texture.
...
The black-box nature of deep learning (DL) models presents a significant challenge for their adoption in clinical settings. The field of explainable artificial intelligence (XAI) has emerged to improve the transparency and interpretability of models. However, current techniques do not adequately describe the reasoning underpinning DL models. This study replicates and extends previous research on the use of texture analysis to improve interpretability in clinically geared segmentation tasks. We evaluate Law's Texture Energy Measures (LTEMs) in the learning and decision-making processes of different DL architectures. We extend the work to include breast cancer, skin lesion, and gastrointestinal polyp datasets, as well as CLAHE-enhanced datasets to identify any divergence in learning. Experimental results reiterate that LTEMs, specifically level-edge convolution masks, are highly influential across multiple DL architectures. Additionally, Gray-Level Co-occurrence Matrix (GLCM) analysis highlights autocorrelation as a key descriptor. The results confirm that texture-based representations, learned primarily in the early layers of the network, are sufficient for robust learning. Through LTEMs, we can characterize the patterns learned in DL and associate these patterns with verbal descriptions and clinically objective measures, thus translating the DL learning into human terms. This psychophysical approach eases the clinical interpretability of DL models. Code availability: https://github.com/xrai-lib/xai-texture.
Journal article
(2026)
-
Katharina Schneider, Louise Spekking, Sepinoud Azimi, Barbora Peltanová, Daniel Rösel, Joel S. Brown, Robert A. Gatenby, Jan Brábek, Kateřina Staňková
Adaptive therapy, which anticipates and counters the evolution of resistance in cancer cells, has gained significant traction, especially following the success of the Zhang et al.'s protocol in treating metastatic castrate-resistant prostate cancer. While several adaptive therapies have now advanced to clinical trials, none currently incorporates migrastatics, i.e. treatments designed to inhibit cancer cell metastasis. In this study, we propose integrating migrastatics into adaptive therapy protocols and evaluate its potential benefits through a spatial game-theoretic model. Our results demonstrate that combining adaptive therapy with migrastatics effectively delays the onset of metastases and reduces both the number and size of metastases in most cancer scenarios analyzed. Including migrastatics to adaptive therapy not only extends the time to the first metastasis, but also enhances the overall efficacy of adaptive therapies. Our findings suggest a promising new direction for cancer treatment, where adaptive therapy, in combination with migrastatic agents, can target both the evolution of resistance and the metastatic spread of cancer cells.
...
Adaptive therapy, which anticipates and counters the evolution of resistance in cancer cells, has gained significant traction, especially following the success of the Zhang et al.'s protocol in treating metastatic castrate-resistant prostate cancer. While several adaptive therapies have now advanced to clinical trials, none currently incorporates migrastatics, i.e. treatments designed to inhibit cancer cell metastasis. In this study, we propose integrating migrastatics into adaptive therapy protocols and evaluate its potential benefits through a spatial game-theoretic model. Our results demonstrate that combining adaptive therapy with migrastatics effectively delays the onset of metastases and reduces both the number and size of metastases in most cancer scenarios analyzed. Including migrastatics to adaptive therapy not only extends the time to the first metastasis, but also enhances the overall efficacy of adaptive therapies. Our findings suggest a promising new direction for cancer treatment, where adaptive therapy, in combination with migrastatic agents, can target both the evolution of resistance and the metastatic spread of cancer cells.
Intelligent nanoparticle design
Unlocking the potential of AI for transformative drug delivery
Artificial intelligence (AI) is revolutionizing nanoparticle (NP)-based drug delivery by tackling design, synthesis, and optimization challenges. Traditional approaches to NP development often rely on trial-and-error methods, leading to scalability, biocompatibility, and targeted drug release inefficiencies. This review explores how AI-driven models are transforming the landscape of NP formulation, from enhancing drug encapsulation and optimizing release kinetics to improving targeted delivery and overcoming physiological barriers. Additionally, we examine the challenges associated with AI integration, including data limitations and model interpretability, and discuss strategies for bridging these gaps. By leveraging AI, the field of nanomedicine can accelerate the transition from laboratory research to clinical applications, ultimately improving treatment outcomes for complex diseases.
...
Artificial intelligence (AI) is revolutionizing nanoparticle (NP)-based drug delivery by tackling design, synthesis, and optimization challenges. Traditional approaches to NP development often rely on trial-and-error methods, leading to scalability, biocompatibility, and targeted drug release inefficiencies. This review explores how AI-driven models are transforming the landscape of NP formulation, from enhancing drug encapsulation and optimizing release kinetics to improving targeted delivery and overcoming physiological barriers. Additionally, we examine the challenges associated with AI integration, including data limitations and model interpretability, and discuss strategies for bridging these gaps. By leveraging AI, the field of nanomedicine can accelerate the transition from laboratory research to clinical applications, ultimately improving treatment outcomes for complex diseases.
Journal article
(2025)
-
Phornphawit Manasut, Md Saleh Ibtasham, Zeynep Yaradanakul, Sepinoud Azimi, Sebastien Lafond, Bogdan Iancu
In recent years, CNN-based object detectors have been widely adopted in autonomous systems. Although their capabilities are employed across various industries, these detectors are inherently susceptible to adversarial attacks. Despite extensive studies on their effects on image classification, adversarial attacks remain largely unexplored in object detection. In particular, we note the reduced number of studies employing benchmarks for these types of attacks. Object detectors can be easily deceived by adding carefully devised perturbations to their inputs, rendering them unreliable. This study investigates the transferability of one such adversarial attack type, the Targeted Objectness Gradient (TOG), on different variations of the YOLO architecture to formally assess its vulnerability under different scenarios in the maritime domain. To investigate the significance of TOG adversarial attacks across variations of YOLO architectures and combinations of maritime datasets (all publicly available), we conducted a statistical analysis of black-box and white-box attacks. Our research questions were formulated to address a range of concerns that encompass various complexities to be considered in the detection of maritime objects. Our presented results underline the transferable nature of TOG adversarial attacks and the compelling need to benchmark such attacks in the maritime object detection domain.
...
In recent years, CNN-based object detectors have been widely adopted in autonomous systems. Although their capabilities are employed across various industries, these detectors are inherently susceptible to adversarial attacks. Despite extensive studies on their effects on image classification, adversarial attacks remain largely unexplored in object detection. In particular, we note the reduced number of studies employing benchmarks for these types of attacks. Object detectors can be easily deceived by adding carefully devised perturbations to their inputs, rendering them unreliable. This study investigates the transferability of one such adversarial attack type, the Targeted Objectness Gradient (TOG), on different variations of the YOLO architecture to formally assess its vulnerability under different scenarios in the maritime domain. To investigate the significance of TOG adversarial attacks across variations of YOLO architectures and combinations of maritime datasets (all publicly available), we conducted a statistical analysis of black-box and white-box attacks. Our research questions were formulated to address a range of concerns that encompass various complexities to be considered in the detection of maritime objects. Our presented results underline the transferable nature of TOG adversarial attacks and the compelling need to benchmark such attacks in the maritime object detection domain.
Conference paper
(2024)
-
Md Masum Billah, Pragati Manandhar, Sarosh Krishan, Alejandro Cedillo, Hergys Rexha, Sébastien Lafond, Kurt K Benke, Sepinoud Azimi, Janan Arslan
Despite their predictive capabilities and rapid advancement, the black-box nature of Artificial Intelligence (AI) models, particularly in healthcare, has sparked debate regarding their trustworthiness and accountability. In response, the field of Explainable AI (XAI) has emerged, aiming to create transparent AI technologies. We present a novel approach to enhance AI interpretability by leveraging texture analysis, with a focus on cancer datasets. By focusing on specific texture features and their correlations with a prediction outcome extracted from medical images, our proposed methodology aims to elucidate the underlying mechanics of AI, improve AI trustworthiness, and facilitate human understanding. The code is available at https://github.com/xrai-lib/xai-texture.
...
Despite their predictive capabilities and rapid advancement, the black-box nature of Artificial Intelligence (AI) models, particularly in healthcare, has sparked debate regarding their trustworthiness and accountability. In response, the field of Explainable AI (XAI) has emerged, aiming to create transparent AI technologies. We present a novel approach to enhance AI interpretability by leveraging texture analysis, with a focus on cancer datasets. By focusing on specific texture features and their correlations with a prediction outcome extracted from medical images, our proposed methodology aims to elucidate the underlying mechanics of AI, improve AI trustworthiness, and facilitate human understanding. The code is available at https://github.com/xrai-lib/xai-texture.
Predicting efficacy of drug-carrier nanoparticle designs for cancer treatment
A machine learning-based solution
Journal article
(2023)
-
Md Raisul Kibria, Refo Ilmiya Akbar, Poonam Nidadavolu, Oksana Havryliuk, Sébastien Lafond, Sepinoud Azimi
Molecular Dynamic (MD) simulations are very effective in the discovery of nanomedicines for treating cancer, but these are computationally expensive and time-consuming. Existing studies integrating machine learning (ML) into MD simulation to enhance the process and enable efficient analysis cannot provide direct insights without the complete simulation. In this study, we present an ML-based approach for predicting the solvent accessible surface area (SASA) of a nanoparticle (NP), denoting its efficacy, from a fraction of the MD simulations data. The proposed framework uses a time series model for simulating the MD, resulting in an intermediate state, and a second model to calculate the SASA in that state. Empirically, the solution can predict the SASA value 260 timesteps ahead 7.5 times faster with a very low average error of 1956.93. We also introduce the use of an explainability technique to validate the predictions. This work can reduce the computational expense of both processing and data size greatly while providing reliable solutions for the nanomedicine design process.
...
Molecular Dynamic (MD) simulations are very effective in the discovery of nanomedicines for treating cancer, but these are computationally expensive and time-consuming. Existing studies integrating machine learning (ML) into MD simulation to enhance the process and enable efficient analysis cannot provide direct insights without the complete simulation. In this study, we present an ML-based approach for predicting the solvent accessible surface area (SASA) of a nanoparticle (NP), denoting its efficacy, from a fraction of the MD simulations data. The proposed framework uses a time series model for simulating the MD, resulting in an intermediate state, and a second model to calculate the SASA in that state. Empirically, the solution can predict the SASA value 260 timesteps ahead 7.5 times faster with a very low average error of 1956.93. We also introduce the use of an explainability technique to validate the predictions. This work can reduce the computational expense of both processing and data size greatly while providing reliable solutions for the nanomedicine design process.
Journal article
(2023)
-
Muhammad Junaid Haris, Aanchal Upreti, Melih Kurtaran, Filip Ginter, Sebastien Lafond, Sepinoud Azimi
The problem of gender bias is highly prevalent and well known. In this paper, we have analysed the portrayal of gender roles in English movies, a medium that effectively influences society in shaping people’s beliefs and opinions. First, we gathered scripts of films from different genres and derived sentiments and emotions using natural language processing techniques. Afterwards, we converted the scripts into embeddings, i.e., a way of representing text in the form of vectors. With a thorough investigation, we found specific patterns in male and female characters’ personality traits in movies that align with societal stereotypes. Furthermore, we used mathematical and machine learning techniques and found some biases wherein men are shown to be more dominant and envious than women, whereas women have more joyful roles in movies. In our work, we introduce, to the best of our knowledge, a novel technique to convert dialogues into an array of emotions by combining it with Plutchik’s wheel of emotions. Our study aims to encourage reflections on gender equality in the domain of film and facilitate other researchers in analysing movies automatically instead of using manual approaches.
...
The problem of gender bias is highly prevalent and well known. In this paper, we have analysed the portrayal of gender roles in English movies, a medium that effectively influences society in shaping people’s beliefs and opinions. First, we gathered scripts of films from different genres and derived sentiments and emotions using natural language processing techniques. Afterwards, we converted the scripts into embeddings, i.e., a way of representing text in the form of vectors. With a thorough investigation, we found specific patterns in male and female characters’ personality traits in movies that align with societal stereotypes. Furthermore, we used mathematical and machine learning techniques and found some biases wherein men are shown to be more dominant and envious than women, whereas women have more joyful roles in movies. In our work, we introduce, to the best of our knowledge, a novel technique to convert dialogues into an array of emotions by combining it with Plutchik’s wheel of emotions. Our study aims to encourage reflections on gender equality in the domain of film and facilitate other researchers in analysing movies automatically instead of using manual approaches.
Journal article
(2023)
-
Janine Grolleman, Nicole C. A. van Engeland, Minahil Raza, Sepinoud Azimi , Vito Conte, Cecilia M. Sahlgren, Carlijn V. C. Bouten
Recent experimental evidence indicates a role for the intermediate filament vimentin in regulating cellular mechanical homeostasis, but its precise contribution remains to be discovered. Mechanical homeostasis requires a balanced bi-directional interplay between the cell’s microenvironment and the cellular morphological and mechanical state—this balance being regulated via processes of mechanotransduction and mechanoresponse, commonly referred to as mechanoreciprocity. Here, we systematically analyze vimentin-expressing and vimentin-depleted cells in a swatch of in vitro cellular microenvironments varying in stiffness and/or ECM density. We find that vimentin-expressing cells maintain mechanical homeostasis by adapting cellular morphology and mechanics to micromechanical changes in the microenvironment. However, vimentin-depleted cells lose this mechanoresponse ability on short timescales, only to reacquire it on longer time scales. Indeed, we find that the morphology and mechanics of vimentin-depleted cell in stiffened microenvironmental conditions can get restored to the homeostatic levels of vimentin-expressing cells. Additionally, we observed vimentin-depleted cells increasing collagen matrix synthesis and its crosslinking, a phenomenon which is known to increase matrix stiffness, and which we now hypothesize to be a cellular compensation mechanism for the loss of vimentin. Taken together, our findings provide further insight in the regulating role of intermediate filament vimentin in mediating mechanoreciprocity and mechanical homeostasis.
...
Recent experimental evidence indicates a role for the intermediate filament vimentin in regulating cellular mechanical homeostasis, but its precise contribution remains to be discovered. Mechanical homeostasis requires a balanced bi-directional interplay between the cell’s microenvironment and the cellular morphological and mechanical state—this balance being regulated via processes of mechanotransduction and mechanoresponse, commonly referred to as mechanoreciprocity. Here, we systematically analyze vimentin-expressing and vimentin-depleted cells in a swatch of in vitro cellular microenvironments varying in stiffness and/or ECM density. We find that vimentin-expressing cells maintain mechanical homeostasis by adapting cellular morphology and mechanics to micromechanical changes in the microenvironment. However, vimentin-depleted cells lose this mechanoresponse ability on short timescales, only to reacquire it on longer time scales. Indeed, we find that the morphology and mechanics of vimentin-depleted cell in stiffened microenvironmental conditions can get restored to the homeostatic levels of vimentin-expressing cells. Additionally, we observed vimentin-depleted cells increasing collagen matrix synthesis and its crosslinking, a phenomenon which is known to increase matrix stiffness, and which we now hypothesize to be a cellular compensation mechanism for the loss of vimentin. Taken together, our findings provide further insight in the regulating role of intermediate filament vimentin in mediating mechanoreciprocity and mechanical homeostasis.
Towards Integrated Digital-Twins
An Application Framework for Autonomous Maritime Surface Vessel Development
Journal article
(2022)
-
Minahil Raza, Hanna Prokopova, Samir Huseynzade, Sepinoud Azimi, Sebastien Lafond
The use of digital twins for the development of Autonomous Maritime Surface Vessels (AMSVs) has enormous potential to resolve the increasing need for water-based navigation and safety at the sea. Aiming at the problem of lack of broad and integrated digital twin implementations with live data along with the absence of a digital twin-driven framework for AMSV design and development, an application framework for the development of a fully autonomous vessel using an integrated digital twin in a 3D simulation environment has been presented. Our framework has 4 layers which ensure that simulation and real-world vessel and the environment are as close as possible. Åboat, an in-house, experimental research platform for maritime automation and autonomous surface vessel applications, equipped with two trolling electric motors, cameras, LiDARs, IMU and GPS has been used as the case study to provide a proof of concept. Åboat, its sensors, and the environment have been replicated in a commercial, 3D simulation environment, AILiveSim. Using the proposed application framework, we develop obstacle detection and path planning systems based on machine learning which leverage live data from a 3D simulation environment to mirror the complex dynamics of the real world. Exploiting the proposed application framework, the rewards across training episodes of a Deep Reinforcement Learning model are evaluated for live simulated data in AILiveSim.
...
The use of digital twins for the development of Autonomous Maritime Surface Vessels (AMSVs) has enormous potential to resolve the increasing need for water-based navigation and safety at the sea. Aiming at the problem of lack of broad and integrated digital twin implementations with live data along with the absence of a digital twin-driven framework for AMSV design and development, an application framework for the development of a fully autonomous vessel using an integrated digital twin in a 3D simulation environment has been presented. Our framework has 4 layers which ensure that simulation and real-world vessel and the environment are as close as possible. Åboat, an in-house, experimental research platform for maritime automation and autonomous surface vessel applications, equipped with two trolling electric motors, cameras, LiDARs, IMU and GPS has been used as the case study to provide a proof of concept. Åboat, its sensors, and the environment have been replicated in a commercial, 3D simulation environment, AILiveSim. Using the proposed application framework, we develop obstacle detection and path planning systems based on machine learning which leverage live data from a 3D simulation environment to mirror the complex dynamics of the real world. Exploiting the proposed application framework, the rewards across training episodes of a Deep Reinforcement Learning model are evaluated for live simulated data in AILiveSim.
Conference paper
(2022)
-
Minahil Raza, Hanna Prokopova, Samir Huseynzade, Sepinoud Azimi, Sebastien Lafond
Obstacle detection is a fundamental capability of an autonomous maritime surface vessel (AMSV). State-of-the-art obstacle detection algorithms are based on convolutional neural networks (CNNs). While CNNs provide higher detection accuracy and fast detection speed, they require enormous amounts of data for their training. In particular, the availability of domain-specific datasets is a challenge for obstacle detection. The difficulty in conducting onsite experiments limits the collection of maritime datasets. Owing to the logistic cost of conducting on-site operations, simulation tools provide a safe and cost-efficient alternative for data collection. In this work, we introduce SimuShips, a publicly available simulation-based dataset for maritime environments. Our dataset consists of 9471 high-resolution (1920x1080) images which include a wide range of obstacle types, atmospheric and illumination conditions along with occlusion, scale and visible proportion variations. We provide annotations in the form of bounding boxes. In addition, we conduct experiments with YOLOv5 to test the viability of simulation data. Our experiments indicate that the combination of real and simulated images improves the recall for all classes by 2.9%.
...
Obstacle detection is a fundamental capability of an autonomous maritime surface vessel (AMSV). State-of-the-art obstacle detection algorithms are based on convolutional neural networks (CNNs). While CNNs provide higher detection accuracy and fast detection speed, they require enormous amounts of data for their training. In particular, the availability of domain-specific datasets is a challenge for obstacle detection. The difficulty in conducting onsite experiments limits the collection of maritime datasets. Owing to the logistic cost of conducting on-site operations, simulation tools provide a safe and cost-efficient alternative for data collection. In this work, we introduce SimuShips, a publicly available simulation-based dataset for maritime environments. Our dataset consists of 9471 high-resolution (1920x1080) images which include a wide range of obstacle types, atmospheric and illumination conditions along with occlusion, scale and visible proportion variations. We provide annotations in the form of bounding boxes. In addition, we conduct experiments with YOLOv5 to test the viability of simulation data. Our experiments indicate that the combination of real and simulated images improves the recall for all classes by 2.9%.
Neurally adjusted ventilatory assist in ventilated very preterm infants
A crossover study
Journal article
(2021)
-
Arata Oda, Vilhelmiina Parikka, Liisa Lehtonen, Sepinoud Azimi, Ivan Porres, Hanna Soukka
Objective: To assess the effects of neurally adjusted ventilatory assist (NAVA) ventilation on oxygenation and respiratory parameters in preterm infants. Study Design: An observational crossover study with a convenience sample of 19 infants born before 30 gestational weeks. Study parameters were recorded during 3-h periods of both NAVA and conventional ventilation. The proportion of time peripheral oxygen saturation (SpO2) and cerebral regional oxygen saturation (cRSO2) were within their target ranges, plus the number and severity of desaturation episodes were analyzed. In addition, electrical activity of the diaphragm (Edi), neural respiratory rates, and peak inspiratory pressures (PIPs) were recorded. Results: Infants were born at a median age of 264/7 gestational weeks (range: 230/7–293/7); the study was performed at a median age of 20 days (range: 1–82). The proportion of time SpO2 was within the target range, the number of peripheral desaturations or cRSO2 did not differ between the modes. However, the desaturation severity index was lower (131 vs. 152; p =.03) and fewer manual supplemental oxygen adjustments (1.3 vs. 2.2/h; p =.006) were needed during the period of NAVA ventilation following conventional ventilation. The mean Edi (8.1 vs. 11.4 µV; p <.006) and PIP values (14.9 vs. 19.1; p <.001) were lower during the NAVA mode. Conclusions: Although NAVA ventilation did not increase the proportion of time with optimal saturation, it was associated with decreased diaphragmatic activity, lower PIPs, less severe hypoxemic events, and fewer manual oxygen adjustments in very preterm infants.
...
Objective: To assess the effects of neurally adjusted ventilatory assist (NAVA) ventilation on oxygenation and respiratory parameters in preterm infants. Study Design: An observational crossover study with a convenience sample of 19 infants born before 30 gestational weeks. Study parameters were recorded during 3-h periods of both NAVA and conventional ventilation. The proportion of time peripheral oxygen saturation (SpO2) and cerebral regional oxygen saturation (cRSO2) were within their target ranges, plus the number and severity of desaturation episodes were analyzed. In addition, electrical activity of the diaphragm (Edi), neural respiratory rates, and peak inspiratory pressures (PIPs) were recorded. Results: Infants were born at a median age of 264/7 gestational weeks (range: 230/7–293/7); the study was performed at a median age of 20 days (range: 1–82). The proportion of time SpO2 was within the target range, the number of peripheral desaturations or cRSO2 did not differ between the modes. However, the desaturation severity index was lower (131 vs. 152; p =.03) and fewer manual supplemental oxygen adjustments (1.3 vs. 2.2/h; p =.006) were needed during the period of NAVA ventilation following conventional ventilation. The mean Edi (8.1 vs. 11.4 µV; p <.006) and PIP values (14.9 vs. 19.1; p <.001) were lower during the NAVA mode. Conclusions: Although NAVA ventilation did not increase the proportion of time with optimal saturation, it was associated with decreased diaphragmatic activity, lower PIPs, less severe hypoxemic events, and fewer manual oxygen adjustments in very preterm infants.
Journal article
(2021)
-
Namid R. Stillman, Igor Balaz, Michail Antisthenis Tsompanas, Marina Kovacevic, Sepinoud Azimi, Sébastien Lafond, Andrew Adamatzky, Sabine Hauert
We present the EVONANO platform for the evolution of nanomedicines with application to anti-cancer treatments. Our work aims to decrease both the time and cost required to develop nanoparticle designs. EVONANO includes a simulator to grow tumours, extract representative scenarios, and simulate nanoparticle transport through these scenarios in order to predict nanoparticle distribution. The nanoparticle designs are optimised using machine learning to efficiently find the most effective anti-cancer treatments. We demonstrate EVONANO with two examples optimising the properties of nanoparticles and treatment to selectively kill cancer cells over a range of tumour environments. Our platform shows how in silico models that capture both tumour and tissue-scale dynamics can be combined with machine learning to optimise nanomedicine.
...
We present the EVONANO platform for the evolution of nanomedicines with application to anti-cancer treatments. Our work aims to decrease both the time and cost required to develop nanoparticle designs. EVONANO includes a simulator to grow tumours, extract representative scenarios, and simulate nanoparticle transport through these scenarios in order to predict nanoparticle distribution. The nanoparticle designs are optimised using machine learning to efficiently find the most effective anti-cancer treatments. We demonstrate EVONANO with two examples optimising the properties of nanoparticles and treatment to selectively kill cancer cells over a range of tumour environments. Our platform shows how in silico models that capture both tumour and tissue-scale dynamics can be combined with machine learning to optimise nanomedicine.
Conference paper
(2020)
-
Ivan Porres, Sepinoud Azimi, Sebastien Lafond, Johan Lilius, Johanna Salokannel, Mirva Salokorpi
This paper explores the state of the art on to methods to verify and validate navigation algorithms for autonomous surface ships. We perform a systematic mapping study to find research works published in the last 10 years proposing new algorithms for autonomous navigation and collision avoidance and we have extracted what verification and validation approaches have been applied on these algorithms. We observe that most research works use simulations to validate their algorithms. However, these simulations often involve just a few scenarios designed manually. This raises the question if the algorithms have been validated properly. To remedy this, we propose the use of a systematic scenario-based testing approach to validate navigation algorithms extensively.
...
This paper explores the state of the art on to methods to verify and validate navigation algorithms for autonomous surface ships. We perform a systematic mapping study to find research works published in the last 10 years proposing new algorithms for autonomous navigation and collision avoidance and we have extracted what verification and validation approaches have been applied on these algorithms. We observe that most research works use simulations to validate their algorithms. However, these simulations often involve just a few scenarios designed manually. This raises the question if the algorithms have been validated properly. To remedy this, we propose the use of a systematic scenario-based testing approach to validate navigation algorithms extensively.
Conference paper
(2020)
-
Sepinoud Azimi , Johanna Salokannel, Sebastien Lafond, Johan Lilius, Mirva Salokorpi, Ivan Porres
In this article we present the state of the art in the field of autonomous surface ship navigation using machine learning. We discuss the main challenges towards the development of fully autonomous navigation systems with the International Regulations for Preventing Collisions at Sea (COLREGs). Finally, we propose two alternative approaches that are based on machine learning. Existing COLREGs-based navigation and collision avoidance algorithms are based on traditional search-based planning and optimization algorithms. We consider that these approaches are suitable when the problem space is defined completely and rigorously. However, experts believe that is not the case for COLREGs since it leaves many aspects open to the interpretation of the captain. For example, COLREGs expects that any collision avoidance action shall be taken with due regard to the observance of good seamanship, a concept not defined in the convention. Furthermore, many rules are defined using undefined concepts like safe distance, or keywords like early, or substantial, without giving any definition. COLREGs even allow for the rules to be broken to avoid an accident. Due to this, traditional planning approaches may not be able to handle complex scenarios that are underspecified according to COLREGs. An alternative is the use of machine learning (ML), reinforcement learning (RL) and imitation learning (IL) at the core of autonomous navigation systems. Machine learning is known to succeed and outperform traditional approaches specially in vaguely defined problem domains, where it is difficult, if not impossible, to create a full formal specification of the phenomenon under study. We consider this to be the case for COLREGs-based navigation and we conjecture that a ML-based navigation approach can outperform existing search-based and optimization algorithms.
...
In this article we present the state of the art in the field of autonomous surface ship navigation using machine learning. We discuss the main challenges towards the development of fully autonomous navigation systems with the International Regulations for Preventing Collisions at Sea (COLREGs). Finally, we propose two alternative approaches that are based on machine learning. Existing COLREGs-based navigation and collision avoidance algorithms are based on traditional search-based planning and optimization algorithms. We consider that these approaches are suitable when the problem space is defined completely and rigorously. However, experts believe that is not the case for COLREGs since it leaves many aspects open to the interpretation of the captain. For example, COLREGs expects that any collision avoidance action shall be taken with due regard to the observance of good seamanship, a concept not defined in the convention. Furthermore, many rules are defined using undefined concepts like safe distance, or keywords like early, or substantial, without giving any definition. COLREGs even allow for the rules to be broken to avoid an accident. Due to this, traditional planning approaches may not be able to handle complex scenarios that are underspecified according to COLREGs. An alternative is the use of machine learning (ML), reinforcement learning (RL) and imitation learning (IL) at the core of autonomous navigation systems. Machine learning is known to succeed and outperform traditional approaches specially in vaguely defined problem domains, where it is difficult, if not impossible, to create a full formal specification of the phenomenon under study. We consider this to be the case for COLREGs-based navigation and we conjecture that a ML-based navigation approach can outperform existing search-based and optimization algorithms.
Journal article
(2020)
-
Sepinoud Azimi , Carmen-Gabriela Popa, Tatjana Cucić
The birth of massive open online courses (MOOCs) has had an undeniable effect on how teaching is being delivered. It seems that traditional in class teaching is becoming less popular with the young generation, the generation that wants to choose when, where and at what pace they are learning. As such, many universities are moving towards taking their courses, at least partially, online. However, online courses, although very appealing to the younger generation of learners, come at a cost. For example, the dropout rate of such courses is higher than that of more traditional ones, and the reduced in person interaction with the teachers results in less timely guidance and intervention from the educators. Machine learning (ML) based approaches have shown phenomenal successes in other domains. The existing stigma that applying ML based techniques requires a large amount of data seems to be a bottleneck when dealing with small scale courses with limited amounts of produced data. In this study, we show not only that the data collected from an online learning management system could be well utilized in order to predict students overall performance but also that it could be used to propose timely intervention strategies to boost the students performance level. The results of this study indicate that effective intervention strategies could be suggested as early as the middle of the course to change the course of students progress for the better. We also present an assistive pedagogical tool based on the outcome of this study, to assist in identifying challenging students and in suggesting early intervention strategies.
...
The birth of massive open online courses (MOOCs) has had an undeniable effect on how teaching is being delivered. It seems that traditional in class teaching is becoming less popular with the young generation, the generation that wants to choose when, where and at what pace they are learning. As such, many universities are moving towards taking their courses, at least partially, online. However, online courses, although very appealing to the younger generation of learners, come at a cost. For example, the dropout rate of such courses is higher than that of more traditional ones, and the reduced in person interaction with the teachers results in less timely guidance and intervention from the educators. Machine learning (ML) based approaches have shown phenomenal successes in other domains. The existing stigma that applying ML based techniques requires a large amount of data seems to be a bottleneck when dealing with small scale courses with limited amounts of produced data. In this study, we show not only that the data collected from an online learning management system could be well utilized in order to predict students overall performance but also that it could be used to propose timely intervention strategies to boost the students performance level. The results of this study indicate that effective intervention strategies could be suggested as early as the middle of the course to change the course of students progress for the better. We also present an assistive pedagogical tool based on the outcome of this study, to assist in identifying challenging students and in suggesting early intervention strategies.
Conference paper
(2020)
-
Ivan Porres, Sepinoud Azimi, Johan Lilius
We propose a method for scenario-based testing of maritime collision avoidance systems. The goal is to test an autonomous agent in scenarios that can lead to an unacceptable risk of collision or may clearly not comply with the International Regulations for Preventing Collisions at Sea (COLREGs).Our method is based on the use of a discriminating artificial neural network that is trained online while performing the testing of the agents. Our experimental results show that the proposed algorithm generates test suits composed mostly of challenging scenarios. This allows us to validate quickly if the agent under test can perform the collision avoidance maneuvers safely while abiding the COLREGs.
...
We propose a method for scenario-based testing of maritime collision avoidance systems. The goal is to test an autonomous agent in scenarios that can lead to an unacceptable risk of collision or may clearly not comply with the International Regulations for Preventing Collisions at Sea (COLREGs).Our method is based on the use of a discriminating artificial neural network that is trained online while performing the testing of the agents. Our experimental results show that the proposed algorithm generates test suits composed mostly of challenging scenarios. This allows us to validate quickly if the agent under test can perform the collision avoidance maneuvers safely while abiding the COLREGs.
WEBRSIM
A web-based reaction systems simulator
Book chapter
(2018)
-
Sergiu Ivanov, Vladimir Rogojin, Sepinoud Azimi, Ion Petre
We introduce WEBRSIM, the first web-based simulator for reaction systems. The simulator has an easy-to-use interface where the input is a reaction system and four functionalities: the computation of the interactive process driven by a given context sequence, the behaviour graph of the reaction system, its conservation dependency graph, and all its conserved sets. WEBRSIM comes with a browser-based friendly interface and offers a fast software to support computational modeling with reaction systems.
...
We introduce WEBRSIM, the first web-based simulator for reaction systems. The simulator has an easy-to-use interface where the input is a reaction system and four functionalities: the computation of the interactive process driven by a given context sequence, the behaviour graph of the reaction system, its conservation dependency graph, and all its conserved sets. WEBRSIM comes with a browser-based friendly interface and offers a fast software to support computational modeling with reaction systems.
Journal article
(2017)
-
Sepinoud Azimi
Reaction systems, a mathematical formalism inspired by the mechanisms within a biological cell, focuses on an abstract set-based representation of chemical reactions via facilitation and inhibition. The simple yet elegant nature of reaction systems makes them ideal tools for analysing qualitatively the phenomena which typically are dealt with quantitatively. Steady states are one of the well studied and important subjects across various fields of science ranging from biology, to chemistry, to engineering and economics. Finding all steady states of an arbitrary reaction system has been shown to be an NP-complete problem. We study reaction systems with a small number of reactants and inhibitors and we propose an algorithm to list all steady states of such reaction systems. We also show that the complexity of such an algorithm is polynomial. This reduction in complexity opens a door to transform modelling with reaction systems from an abstract concept to a tool that can be used on real-life case studies.
...
Reaction systems, a mathematical formalism inspired by the mechanisms within a biological cell, focuses on an abstract set-based representation of chemical reactions via facilitation and inhibition. The simple yet elegant nature of reaction systems makes them ideal tools for analysing qualitatively the phenomena which typically are dealt with quantitatively. Steady states are one of the well studied and important subjects across various fields of science ranging from biology, to chemistry, to engineering and economics. Finding all steady states of an arbitrary reaction system has been shown to be an NP-complete problem. We study reaction systems with a small number of reactants and inhibitors and we propose an algorithm to list all steady states of such reaction systems. We also show that the complexity of such an algorithm is polynomial. This reduction in complexity opens a door to transform modelling with reaction systems from an abstract concept to a tool that can be used on real-life case studies.
Journal article
(2017)
-
Sepinoud Azimi, Charmi Panchal, Andrzej Mizera, Ion Petre
Quantitative models may exhibit sophisticated behaviour that includes having multiple steady states, bistability, limit cycles, and period-doubling bifurcation. Such behaviour is typically driven by the numerical dynamics of the model, where the values of various numerical parameters play the crucial role. We introduce in this paper natural correspondents of these concepts to reaction systems modelling, a framework based on elementary set theoretical, forbidding/enforcing-based mechanisms. We construct several reaction systems models exhibiting these properties.
...
Quantitative models may exhibit sophisticated behaviour that includes having multiple steady states, bistability, limit cycles, and period-doubling bifurcation. Such behaviour is typically driven by the numerical dynamics of the model, where the values of various numerical parameters play the crucial role. We introduce in this paper natural correspondents of these concepts to reaction systems modelling, a framework based on elementary set theoretical, forbidding/enforcing-based mechanisms. We construct several reaction systems models exhibiting these properties.