RESUMO
The implementation of clinical-decision support algorithms for medical imaging faces challenges with reliability and interpretability. Here, we establish a diagnostic tool based on a deep-learning framework for the screening of patients with common treatable blinding retinal diseases. Our framework utilizes transfer learning, which trains a neural network with a fraction of the data of conventional approaches. Applying this approach to a dataset of optical coherence tomography images, we demonstrate performance comparable to that of human experts in classifying age-related macular degeneration and diabetic macular edema. We also provide a more transparent and interpretable diagnosis by highlighting the regions recognized by the neural network. We further demonstrate the general applicability of our AI system for diagnosis of pediatric pneumonia using chest X-ray images. This tool may ultimately aid in expediting the diagnosis and referral of these treatable conditions, thereby facilitating earlier treatment, resulting in improved clinical outcomes. VIDEO ABSTRACT.
Assuntos
Aprendizado Profundo , Diagnóstico por Imagem , Pneumonia/diagnóstico , Criança , Humanos , Redes Neurais de Computação , Pneumonia/diagnóstico por imagem , Curva ROC , Reprodutibilidade dos Testes , Tomografia de Coerência ÓpticaRESUMO
In this paper, we describe the development of a bipedal robot that models the neuromuscular architecture of human walking. The body is based on principles derived from human muscular architecture, using muscles on straps to mimic agonist/antagonist muscle action as well as bifunctional muscles. Load sensors in the straps model Golgi tendon organs. The neural architecture is a central pattern generator (CPG) composed of a half-center oscillator combined with phase-modulated reflexes that is simulated using a spiking neural network. We show that the interaction between the reflex system, body dynamics and CPG results in a walking cycle that is entrained to the dynamics of the system. We also show that the CPG helped stabilize the gait against perturbations relative to a purely reflexive system, and compared the joint trajectories to human walking data. This robot represents a complete physical, or 'neurorobotic', model of the system, demonstrating the usefulness of this type of robotics research for investigating the neurophysiological processes underlying walking in humans and animals.
Assuntos
Geradores de Padrão Central/fisiologia , Simulação por Computador , Locomoção/fisiologia , Desempenho Psicomotor/fisiologia , Robótica/métodos , Potenciais de Ação/fisiologia , Humanos , Rede Nervosa/fisiologia , Robótica/instrumentação , Caminhada/fisiologiaRESUMO
Most conventional robots rely on controlling the location of the center of pressure to maintain balance, relying mainly on foot pressure sensors for information. By contrast,humans rely on sensory data from multiple sources, including proprioceptive, visual, and vestibular sources. Several models have been developed to explain how humans reconcile information from disparate sources to form a stable sense of balance. These models may be useful for developing robots that are able to maintain dynamic balance more readily using multiple sensory sources. Since these information sources may conflict, reliance by the nervous system on any one channel can lead to ambiguity in the system state. In humans, experiments that create conflicts between different sensory channels by moving the visual field or the support surface indicate that sensory information is adaptively reweighted. Unreliable information is rapidly down-weighted,then gradually up-weighted when it becomes valid again.Human balance can also be studied by building robots that model features of human bodies and testing them under similar experimental conditions. We implement a sensory reweighting model based on an adaptive Kalman filter in abipedal robot, and subject it to sensory tests similar to those used on human subjects. Unlike other implementations of sensory reweighting in robots, our implementation includes vision, by using optic flow to calculate forward rotation using a camera (visual modality), as well as a three-axis gyro to represent the vestibular system (non-visual modality), and foot pressure sensors (proprioceptive modality). Our model estimates measurement noise in real time, which is then used to recompute the Kalman gain on each iteration, improving the ability of the robot to dynamically balance. We observe that we can duplicate many important features of postural sw ay in humans, including automatic sensory reweighting,effects, constant phase with respect to amplitude, and a temporal asymmetry in the reweighting gains.
Assuntos
Adaptação Fisiológica/fisiologia , Meio Ambiente , Modelos Biológicos , Dinâmica não Linear , Equilíbrio Postural/fisiologia , Sensação/fisiologia , Simulação por Computador , Humanos , Fluxo Óptico , Estimulação LuminosaRESUMO
We show that an ongoing locomotor pattern can be dynamically controlled by applying discrete pulses of electrical stimulation to the central pattern generator (CPG) for locomotion. Data are presented from a pair of experiments on biological (wetware) and electrical (hardware) models of the CPG demonstrating that stimulation causes brief deviations from the CPG's limit cycle activity. The exact characteristics of the deviation depend strongly on the phase of stimulation. Applications of this work are illustrated by examples showing how locomotion can be controlled by using a feedback loop to monitor CPG activity and applying stimuli at the appropriate times to modulate motor output. Eventually, this approach could lead to development of a neuroprosthetic device for restoring locomotion after paralysis.
Assuntos
Locomoção/fisiologia , Modelos Neurológicos , Neurônios Motores/fisiologia , Rede Nervosa/fisiologia , Periodicidade , Potenciais de Ação/fisiologia , Animais , Estimulação Elétrica , Marcha , Humanos , Lampreias , Matemática , Natação/fisiologiaRESUMO
During locomotion in a cluttered terrain, certain terrain surfaces such as an icy one are not appropriate for foot placement; an alternate choice is required. In a previous study we showed that the selection of foot placement is not random but systematic; the dominant choices made are not uniquely defined by the available or predicted sensory inputs. We argued that selection is guided by specific rules and involves minimal displacement of the foot from its normal landing spot. The experimental protocol involved implicit spatial constraint by requiring individuals to step on the force plate that could trigger a lighted area to be avoided, thereby requiring individuals to respond within one step-cycle. Alternate foot placement was visually identified, but not measured. The purpose of this study was to directly measure foot placement, validate and/or refine the rules used to guide selection, and identify whether the alternate foot placement choices are influenced by spatial and temporal constraints on response selection. The area to be avoided was visible from the start and therefore individuals could plan and implement appropriate avoidance strategies without any temporal constraint. Spatial constraint introduced in this experiment included requirement both to step on a specific location and to avoid stepping on a specific location on the next step. The results provide support for the rules previously identified in guiding foot placement to an alternate location. Minimal displacement of the foot from its normal landing spot was validated as an important factor for selecting alternate foot placement. When several choices satisfied this factor, additional factors guide alternate foot placement. Modifications in the plane of progression are preferred while stepping wide is avoided. When no temporal constraints are imposed on the response selection, enhancing forward progression of the body becomes the dominant determinant followed by stability and lastly by energy costs associated with the modifications. A decision algorithm for selecting foot placement is proposed based on these findings. It is clear that while visual input plays a critical role in guiding foot placement, it is not entirely based on reactive control. This has implications for implementing visually guided adaptive locomotion in legged robots.
Assuntos
Algoritmos , Tomada de Decisões/fisiologia , Locomoção/fisiologia , Comportamento Espacial/fisiologia , Adolescente , Adulto , Análise de Variância , Distribuição de Qui-Quadrado , Feminino , Pé/fisiologia , Humanos , Análise dos Mínimos Quadrados , Masculino , Análise de Regressão , Fatores de Tempo , Caminhada/fisiologiaRESUMO
In biological systems, the task of computing a gait trajectory is shared between the biomechanical and nervous systems. We take the perspective that both of these seemingly different computations are examples of physical computation. Here we describe the progress that has been made toward building a minimal biped system that illustrates this idea. We embed a significant portion of the computation in physical devices, such as capacitors and transistors, to underline the potential power of emphasizing the understanding of physical computation. We describe results in the exploitation of physical computation by (1) using a passive knee to assist in dynamics computation, (2) using an oscillator to drive a monoped mechanism based on the passive knee, (3) using sensory entrainment to coordinate the mechanics with the neural oscillator, (4) coupling two such systems together mechanically at the hip and computationally via the resulting two oscillators to create a biped mechanism, and (5) demonstrating the resulting gait generation in the biped mechanism.