Evolutionary design of explainable algorithms for biomedical image segmentation.

An unresolved issue in contemporary biomedicine is the overwhelming number and diversity of complex images that require annotation, analysis and interpretation. Recent advances in Deep Learning have revolutionized the field of computer vision, creating algorithms that compete with human experts in image segmentation tasks. However, these frameworks require large human-annotated datasets for training and the resulting "black box" models are difficult to interpret. In this study, we introduce Kartezio, a modular Cartesian Genetic Programming-based computational strategy that generates fully transparent and easily interpretable image processing pipelines by iteratively assembling and parameterizing computer vision functions. The pipelines thus generated exhibit comparable precision to state-of-the-art Deep Learning approaches on instance segmentation tasks, while requiring drastically smaller training datasets. This Few-Shot Learning method confers tremendous flexibility, speed, and functionality to this approach. We then deploy Kartezio to solve a series of semantic and instance segmentation problems, and demonstrate its utility across diverse images ranging from multiplexed tissue histopathology images to high resolution microscopy images. While the flexibility, robustness and practical utility of Kartezio make this fully explicable evolutionary designer a potential game-changer in the field of biomedical image processing, Kartezio remains complementary and potentially auxiliary to mainstream Deep Learning approaches.

Explainable machine learning framework to predict personalized physiological aging.

Attaining personalized healthy aging requires accurate monitoring of physiological changes and identifying subclinical markers that predict accelerated or delayed aging. Classic biostatistical methods most rely on supervised variables to estimate physiological aging and do not capture the full complexity of inter-parameter interactions. Machine learning (ML) is promising, but its black box nature eludes direct understanding, substantially limiting physician confidence and clinical usage. Using a broad population dataset from the National Health and Nutrition Examination Survey (NHANES) study including routine biological variables and after selection of XGBoost as the most appropriate algorithm, we created an innovative explainable ML framework to determine a Personalized physiological age (PPA). PPA predicted both chronic disease and mortality independently of chronological age. Twenty-six variables were sufficient to predict PPA. Using SHapley Additive exPlanations (SHAP), we implemented a precise quantitative associated metric for each variable explaining physiological (i.e., accelerated or delayed) deviations from age-specific normative data. Among the variables, glycated hemoglobin (HbA1c) displays a major relative weight in the estimation of PPA. Finally, clustering profiles of identical contextualized explanations reveal different aging trajectories opening opportunities to specific clinical follow-up. These data show that PPA is a robust, quantitative and explainable ML-based metric that monitors personalized health status. Our approach also provides a complete framework applicable to different datasets or variables, allowing precision physiological age estimation.

Explaining the Neuroevolution of Fighting Creatures Through Virtual fMRI.

While interest in artificial neural networks (ANNs) has been renewed by the ubiquitous use of deep learning to solve high-dimensional problems, we are still far from general artificial intelligence. In this article, we address the problem of emergent cognitive capabilities and, more crucially, of their detection, by relying on co-evolving creatures with mutable morphology and neural structure. The former is implemented via both static and mobile structures whose shapes are controlled by cubic splines. The latter uses ESHyperNEAT to discover not only appropriate combinations of connections and weights but also to extrapolate hidden neuron distribution. The creatures integrate low-level perceptions (touch/pain proprioceptors, retina-based vision, frequency-based hearing) to inform their actions. By discovering a functional mapping between individual neurons and specific stimuli, we extract a high-level module-based abstraction of a creature's brain. This drastically simplifies the discovery of relationships between naturally occurring events and their neural implementation. Applying this methodology to creatures resulting from solitary and tag-team co-evolution showed remarkable dynamics such as range-finding and structured communication. Such discovery was made possible by the abstraction provided by the modular ANN which allowed groups of neurons to be viewed as functionally enclosed entities.

Finding a Suitable Class Distribution for Building Histological Images Datasets Used in Deep Model Training-The Case of Cancer Detection.

The class distribution of a training dataset is an important factor which influences the performance of a deep learning-based system. Understanding the optimal class distribution is therefore crucial when building a new training set which may be costly to annotate. This is the case for histological images used in cancer diagnosis where image annotation requires domain experts. In this paper, we tackle the problem of finding the optimal class distribution of a training set to be able to train an optimal model that detects cancer in histological images. We formulate several hypotheses which are then tested in scores of experiments with hundreds of trials. The experiments have been designed to account for both segmentation and classification frameworks with various class distributions in the training set, such as natural, balanced, over-represented cancer, and over-represented non-cancer. In the case of cancer detection, the experiments show several important results: (a) the natural class distribution produces more accurate results than the artificially generated balanced distribution; (b) the over-representation of non-cancer/negative classes (healthy tissue and/or background classes) compared to cancer/positive classes reduces the number of samples which are falsely predicted as cancer (false positive); (c) the least expensive to annotate non-ROI (non-region-of-interest) data can be useful in compensating for the performance loss in the system due to a shortage of expensive to annotate ROI data; (d) the multi-label examples are more useful than the single-label ones to train a segmentation model; and (e) when the classification model is tuned with a balanced validation set, it is less affected than the segmentation model by the class distribution of the training set.

Ultrarapid lytic granule release from CTLs activates Ca2+-dependent synaptic resistance pathways in melanoma cells.

Human cytotoxic T lymphocytes (CTLs) exhibit ultrarapid lytic granule secretion, but whether melanoma cells mobilize defense mechanisms with commensurate rapidity remains unknown. We used single-cell time-lapse microscopy to offer high spatiotemporal resolution analyses of subcellular events in melanoma cells upon CTL attack. Target cell perforation initiated an intracellular Ca2+ wave that propagated outward from the synapse within milliseconds and triggered lysosomal mobilization to the synapse, facilitating membrane repair and conferring resistance to CTL induced cytotoxicity. Inhibition of Ca2+ flux and silencing of synaptotagmin VII limited synaptic lysosomal exposure and enhanced cytotoxicity. Multiplexed immunohistochemistry of patient melanoma nodules combined with automated image analysis showed that melanoma cells facing CD8+ CTLs in the tumor periphery or peritumoral area exhibited significant lysosomal enrichment. Our results identified synaptic Ca2+ entry as the definitive trigger for lysosomal deployment to the synapse upon CTL attack and highlighted an unpredicted defensive topology of lysosome distribution in melanoma nodules.

Machine Learning Analysis of the Anatomical Parameters of the Upper Airway Morphology: A Retrospective Study from Cone-Beam CT Examinations in a French Population.

The objective of this study is to assess, using cone-beam CT (CBCT) examinations, the correlation between hard and soft anatomical parameters and their impact on the characteristics of the upper airway using symbolic regression as a machine learning strategy. Methods: On each CBCT, the upper airway was segmented, and 24 anatomical landmarks were positioned to obtain six angles and 19 distances. Some anatomical landmarks were related to soft tissues and others were related to hard tissues. To explore which variables were the most influential to explain the morphology of the upper airway, principal component and symbolic regression analyses were conducted. Results: In total, 60 CBCT were analyzed from subjects with a mean age of 39.5 ± 13.5 years. The intra-observer reproducibility for each variable was between good and excellent. The horizontal soft palate measure mostly contributed to the reduction of the airway volume and minimal section area with a variable importance of around 50%. The tongue and the position of the hyoid bone were also linked to the upper airway morphology. For hard anatomical structures, the anteroposterior position of the mandible and the maxilla had some influence. Conclusions: Although the volume of the airway is not accessible on all CBCT scans performed by dental practitioners, this study demonstrates that a small number of anatomical elements may be markers of the reduction of the upper airway with, potentially, an increased risk of obstructive sleep apnea. This could help the dentist refer the patient to a suitable physician.

Doctors in Medical Data Sciences: A New Curriculum.

Machine Learning (ML), a branch of Artificial Intelligence, which is competing with human experts in many specialized biomedical fields and will play an increasing role in precision medicine. As with any other technological advances in medicine, the keys to understanding must be integrated into practitioner training. To respond to this challenge, this viewpoint discusses some necessary changes in the health studies curriculum that could help practitioners to interpret decisions the made by a machine and question them in relation to the patient's medical context. The complexity of technology and the inherent criticality of its use in medicine also necessitate a new medical profession. To achieve this objective, this viewpoint will propose new medical practitioners with skills in both medicine and data science: the Doctor in Medical Data Sciences.

Hearing Sensitivity of Primates: Recurrent and Episodic Positive Selection in Hair Cells and Stereocilia Protein-Coding Genes.

The large spectrum of hearing sensitivity observed in primates results from the impact of environmental and behavioral pressures to optimize sound perception and localization. Although evidence of positive selection in auditory genes has been detected in mammals including in Hominoids, selection has never been investigated in other primates. We analyzed 123 genes highly expressed in the inner ear of 27 primate species and tested to what extent positive selection may have shaped these genes in the order Primates tree. We combined both site and branch-site tests to obtain a comprehensive picture of the positively selected genes (PSGs) involved in hearing sensitivity, and drew a detailed description of the most affected branches in the tree. We chose a conservative approach, and thus focused on confounding factors potentially affecting PSG signals (alignment, GC-biased gene conversion, duplications, heterogeneous sequencing qualities). Using site tests, we showed that around 12% of these genes are PSGs, an α selection value consistent with average human genome estimates (10-15%). Using branch-site tests, we showed that the primate tree is heterogeneously affected by positive selection, with the black snub-nosed monkey, the bushbaby, and the orangutan, being the most impacted branches. A large proportion of these genes is inclined to shape hair cells and stereocilia, which are involved in the mechanotransduction process, known to influence frequency perception. Adaptive selection, and more specifically recurrent adaptive evolution, could have acted in parallel on a set of genes (ADGRV1, USH2A, PCDH15, PTPRQ, and ATP8A2) involved in stereocilia growth and the whole complex of bundle links connecting them, in species across different habitats, including high altitude and nocturnal environments.

Comparison of different machine learning approaches to predict dental age using Demirjian's staging approach.

CONTEXT: Dental age, one of the indicators of biological age, is inferred by radiological methods. Two of the most commonly used methods are using Demirjian's radiographic stages of permanent teeth excluding the third molar (Demirjian's and Willems' method). The major drawbacks of these methods are that they are based on population-specific conversion tables and may tend to over- or underestimate dental age in other populations. Machine learning (ML) methods make it possible to create complex data schemas more simply while keeping the same annotation system. The objectives of this study are to compare (1) the capacity of ten machine learning algorithms to predict dental age in children using the seven left permanent mandibular teeth compared to reference methods and (2) the capacity of ten machine learning algorithms to predict dental age from childhood to young adulthood using the seven left permanent mandibular teeth and the four third molars. METHODS: Using a large radiological database of 3605 orthopantomograms (1734 females and 1871 males) of healthy French patients aged between 2 and 24 years, seven left permanent mandibular teeth and the 4 third molars were assessed using Demirjian's stages. Dental age estimation was then performed using Demirjian's reference method and various ML regression methods. Two analyses were performed: with the 7 left mandibular teeth without third molars for the under 16 age group and with the third molars for the entire study population. The different methods were compared using mean error, mean absolute error, root mean square error as metrics, and the Bland-Altman graph. RESULTS: All ML methods had a mean absolute error (MAE) under 0.811 years. With Demirjian's and Willems' methods, the MAE was 1.107 and 0.927 years, respectively. Except for the Bayesian ridge regression that gives poorer accuracy, there was no statistical difference between all ML tested. CONCLUSION: Compared to the two reference methods, all the ML methods based on the maturation stages defined by Demirjian were more accurate in estimating dental age. These results support the use of ML algorithms instead of using standard population tables.

Sequential adjustment of cytotoxic T lymphocyte densities improves efficacy in controlling tumor growth.

Understanding the human cytotoxic T lymphocyte (CTL) biology is crucial to develop novel strategies aiming at maximizing their lytic capacity against cancer cells. Here we introduce an agent-based model, calibrated on population-scale experimental data that allows quantifying human CTL per capita killing. Our model highlights higher individual CTL killing capacity at lower CTL densities and fits experimental data of human melanoma cell killing. The model allows extending the analysis over prolonged time frames, difficult to investigate experimentally, and reveals that initial high CTL densities hamper efficacy to control melanoma growth. Computational analysis forecasts that sequential addition of fresh CTL cohorts improves tumor growth control. In vivo experimental data, obtained in a mouse melanoma model, confirm this prediction. Taken together, our results unveil the impact that sequential adjustment of cellular densities has on enhancing CTL efficacy over long-term confrontation with tumor cells. In perspective, they can be instrumental to refine CTL-based therapeutic strategies aiming at controlling tumor growth.

A checkpoint-oriented cell cycle simulation model.

Modeling and in silico simulations are of major conceptual and applicative interest in studying the cell cycle and proliferation in eukaryotic cells. In this paper, we present a cell cycle checkpoint-oriented simulator that uses agent-based simulation modeling to reproduce the dynamics of a cancer cell population in exponential growth. Our in silico simulations were successfully validated by experimental in vitro supporting data obtained with HCT116 colon cancer cells. We demonstrated that this model can simulate cell confluence and the associated elongation of the G1 phase. Using nocodazole to synchronize cancer cells at mitosis, we confirmed the model predictivity and provided evidence of an additional and unexpected effect of nocodazole on the overall cell cycle progression. We anticipate that this cell cycle simulator will be a potential source of new insights and research perspectives.

ASGART: fast and parallel genome scale segmental duplications mapping.

Motivation: Segmental Duplications (SDs) are DNA fragments longer than 1 kbp, distributed within and between chromosomes and sharing more than 90% identity. Although they hold a significant role in genomic fluidity and adaptability, many key questions about their intrinsic characteristics and mutability remain unsolved due to the persistent difficulty of sequencing highly duplicated genomic regions. The recent development of long and linked-read NGS technologies will increase the need to search for SDs in genomes newly sequenced with these technics. The main limitation of SD analysis will soon be the availability of efficient detection software, to retrieve and compare SD genomic component between species or lineages. Results: In this paper, we present the open-source ASGART, 'A Segmental duplications Gathering And Refining Tool', developed to search for segmental duplications (SDs) in any assembled sequence. We have tested and benchmarked ASGART on five models organisms. Our results demonstrate ASGART's ability to extract SDs from any genome-wide sequence, regardless of genomic size or organizational complexity and quicker than any other software available. Availability and implementation: The online version of ASGART is available at http://asgart.irit.fr. The source code of ASGART is available both on the ASGART website and at https://github.com/delehef/asgart. Supplementary information: Supplementary data are available at Bioinformatics online.

Artificial Gene Regulatory Networks-A Review.

In nature, gene regulatory networks are a key mediator between the information stored in the DNA of living organisms (their genotype) and the structural and behavioral expression this finds in their bodies, surviving in the world (their phenotype). They integrate environmental signals, steer development, buffer stochasticity, and allow evolution to proceed. In engineering, modeling and implementations of artificial gene regulatory networks have been an expanding field of research and development over the past few decades. This review discusses the concept of gene regulation, describes the current state of the art in gene regulatory networks, including modeling and simulation, and reviews their use in artificial evolutionary settings. We provide evidence for the benefits of this concept in natural and the engineering domains.

Y-chromosome descent clusters and male differential reproductive success: young lineage expansions dominate Asian pastoral nomadic populations.

High-frequency microsatellite haplotypes of the male-specific Y-chromosome can signal past episodes of high reproductive success of particular men and their patrilineal descendants. Previously, two examples of such successful Y-lineages have been described in Asia, both associated with Altaic-speaking pastoral nomadic societies, and putatively linked to dynasties descending, respectively, from Genghis Khan and Giocangga. Here we surveyed a total of 5321 Y-chromosomes from 127 Asian populations, including novel Y-SNP and microsatellite data on 461 Central Asian males, to ask whether additional lineage expansions could be identified. Based on the most frequent eight-microsatellite haplotypes, we objectively defined 11 descent clusters (DCs), each within a specific haplogroup, that represent likely past instances of high male reproductive success, including the two previously identified cases. Analysis of the geographical patterns and ages of these DCs and their associated cultural characteristics showed that the most successful lineages are found both among sedentary agriculturalists and pastoral nomads, and expanded between 2100 BCE and 1100 CE. However, those with recent origins in the historical period are almost exclusively found in Altaic-speaking pastoral nomadic populations, which may reflect a shift in political organisation in pastoralist economies and a greater ease of transmission of Y-chromosomes through time and space facilitated by the use of horses.

Cracking the egg: virtual embryogenesis of real robots.

All multicellular living beings are created from a single cell. A developmental process, called embryogenesis, takes this first fertilized cell down a complex path of reproduction, migration, and specialization into a complex organism adapted to its environment. In most cases, the first steps of the embryogenesis take place in a protected environment such as in an egg or in utero. Starting from this observation, we propose a new approach to the generation of real robots, strongly inspired by living systems. Our robots are composed of tens of specialized cells, grown from a single cell using a bio-inspired virtual developmental process. Virtual cells, controlled by gene regulatory networks, divide, migrate, and specialize to produce the robot's body plan (morphology), and then the robot is manually built from this plan. Because the robot is as easy to assemble as Lego, the building process could be easily automated.