Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

A surprise announcement by then prime minister Tony Blair in 1998 led to the creation of a new senior clinical role - the nurse consultant.

Work in progress: The evolving nurse consultant role has all-round benefits and should be embraced, finds Frances Pickersgill

The chicken chorioallantoic membrane (CAM) is a simple, highly vascularized extraembryonic membrane, which performs multiple functions during embryonic development, including but not restricted to gas exchange. Over the last two decades, interest in the CAM as a robust experimental platform to study blood vessels has been shared by specialists working in bioengineering, development, morphology, biochemistry, transplant biology, cancer research and drug development. The tissue composition and accessibility of the CAM for experimental manipulation, makes it an attractive preclinical in vivo model for drug screening and/or for studies of vascular growth. In this article we provide a detailed review of the use of the CAM to study vascular biology and response of blood vessels to a variety of agonists. We also present distinct cultivation protocols discussing their advantages and limitations and provide a summarized update on the use of the CAM in vascular imaging, drug delivery, pharmacokinetics and toxicology.

The chicken chorioallantoic membrane model in biology, medicine and bioengineering.

OBJECTIVE—Fractal analysis can quantify the geometric complexity of the retinal vascular branching pattern and may therefore offer a new method to quantify early diabetic microvascular damage. In this study, we examined the relationship between retinal fractal dimension and retinopathy in young individuals with type 1 diabetes.

RESEARCH DESIGN AND METHODS—We conducted a cross-sectional study of 729 patients with type 1 diabetes (aged 12–20 years) who had seven-field stereoscopic retinal photographs taken of both eyes. From these photographs, retinopathy was graded according to the modified Airlie House classification, and fractal dimension was quantified using a computer-based program following a standardized protocol.

RESULTS—In this study, 137 patients (18.8%) had diabetic retinopathy signs; of these, 105 had mild retinopathy. Median (interquartile range) retinal fractal dimension was 1.46214 (1.45023–1.47217). After adjustment for age, sex, diabetes duration, A1C, blood pressure, and total cholesterol, increasing retinal vascular fractal dimension was significantly associated with increasing odds of retinopathy (odds ratio 3.92 [95% CI 2.02–7.61] for fourth versus first quartile of fractal dimension). In multivariate analysis, each 0.01 increase in retinal vascular fractal dimension was associated with a nearly 40% increased odds of retinopathy (1.37 [1.21–1.56]). This association remained after additional adjustment for retinal vascular caliber.

CONCLUSIONS—Greater retinal fractal dimension, representing increased geometric complexity of the retinal vasculature, is independently associated with early diabetic retinopathy signs in type 1 diabetes. Fractal analysis of fundus photographs may allow quantitative measurement of early diabetic microvascular damage.

https://care.diabetesjournals.org/content/diacare/32/1/106.full.pdf

Quantitative Assessment of Early Diabetic Retinopathy Using Fractal Analysis

Ocular fundus imaging plays a key role in monitoring the health status of the human eye. Currently, a large number of imaging modalities allow the assessment and/or quantification of ocular changes from a healthy status. This review focuses on the main digital fundus imaging modality, color fundus photography, with a brief overview of complementary techniques, such as fluorescein angiography. While focusing on two-dimensional color fundus photography, the authors address the evolution from nondigital to digital imaging and its impact on diagnosis. They also compare several studies performed along the transitional path of this technology. Retinal image processing and analysis, automated disease detection and identification of the stage of diabetic retinopathy (DR) are addressed as well. The authors emphasize the problems of image segmentation, focusing on the major landmark structures of the ocular fundus: the vascular network, optic disk and the fovea. Several proposed approaches for the automatic detection of signs of disease onset and progression, such as microaneurysms, are surveyed. A thorough comparison is conducted among different studies with regard to the number of eyes/subjects, imaging modality, fundus camera used, field of view and image resolution to identify the large variation in characteristics from one study to another. Similarly, the main features of the proposed classifications and algorithms for the automatic detection of DR are compared, thereby addressing computer-aided diagnosis and computer-aided detection for use in screening programs.

/pdf/digital-ocular-fundus-imaging-a-review-6uzy4lzfqq.pdf

Digital ocular fundus imaging: a review.

Quantification of microvascular remodeling as a meaningful discovery tool requires mapping and measurement of site-specific changes within vascular trees and networks. Vessel density and other critical vascular parameters are often modulated by molecular regulators as determined by local vascular architecture. For example, enlargement of vessel diameter by vascular endothelial growth factor (VEGF) is restricted to specific generations of vessel branching (Parsons-Wingerter et al., Microvascular Research72: 91, 2006). The averaging of vessel diameter over many successively smaller generations is therefore not particularly useful. The newly automated, user-interactive software VESsel GENeration Analysis (VESGEN) quantifies major vessel parameters within two-dimensional (2D) vascular trees, networks, and tree-network composites. This report reviews application of VESGEN 2D to angiogenic and lymphangiogenic tissues that includes the human and murine retina, embryonic coronary vessels, and avian chorioallantoic membrane. Software output includes colorized image maps with quantification of local vessel diameter, fractal dimension, tortuosity, and avascular spacing. The density of parameters such as vessel area, length, number, and branch point are quantified according to site-specific generational branching within vascular trees. The sole user input requirement is a binary (black/white) vascular image. Future applications of VESGEN will include analysis of 3D vascular architecture and bioinformatic dimensions such as blood flow and receptor localization. Branching analysis by VESGEN has demonstrated that numerous regulators including VEGF(165), basic fibroblast growth factor, transforming growth factor beta-1, angiostatin and the clinical steroid triamcinolone acetonide induce 'fingerprint' or 'signature' changes in vascular patterning that provide unique readouts of dominant molecular signaling.

/pdf/vesgen-2d-automated-user-interactive-software-for-avckuc1dvf.pdf

VESGEN 2D: Automated, User‐Interactive Software for Quantification and Mapping of Angiogenic and Lymphangiogenic Trees and Networks

PURPOSE To quantify the effects of the steroid triamcinolone acetonide (TA) on branching morphology within the angiogenic microvascular tree of the chorioallantoic membrane (CAM) of quail embryos METHODS Increasing concentrations of TA (0-16 ng/mL) were applied topically on embryonic day (E) 7 to the chorioallantoic membrane (CAM) of quail embryos cultured in petri dishes and incubated for an additional 24 or 48 hours until fixation Binary (black/white) microscopic images of arterial end points were quantified by generational analysis of vessel branching (VESGEN) software to obtain major vascular parameters that include vessel diameter (D(v)), fractal dimension (D(f)), tortuosity (T(v)), and densities of vessel area, length, number, and branch point (A(v), L(v), N(v), and Br(v)) For assessment of specific changes in vascular morphology induced by TA, the VESGEN software automatically segmented the vascular tree into branching generations (G(1) G(10)) according to changes in vessel diameter and branching RESULTS Vessel density decreased significantly up to 34% as the function of increasing concentration of TA according to A(v), L(v), Br(v), N(v), and D(f) TA selectively inhibited the growth of new, small vessels because L(v) decreased from 1314 +/- 061 cm/cm(2) for controls to 8012 +/- 082 cm/cm(2) at 16 ng TA/mL in smaller branching generations (G(7)-G(10)) and for N(v) from 47383 +/- 2985 cm(-2) to 30232 +/- 3309 cm(-2) In contrast, vessel diameter (D(v)) decreased throughout the vascular tree (G(1)-G(10)) CONCLUSIONS By VESGEN analysis, TA selectively inhibited the angiogenesis of smaller blood vessels, but decreased the vessel diameter of all vessels within the vascular tree

Selective inhibition of angiogenesis in small blood vessels and decrease in vessel diameter throughout the vascular tree by triamcinolone acetonide.

This paper discusses the concept and architecture of a machine learning based router for delay tolerant space networks. The techniques of reinforcement learning and Bayesian learning are used to supplement the routing decisions of the popular Contact Graph Routing algorithm. An introduction to the concepts of Contact Graph Routing, Q-routing and Naive Bayes classification are given. The development of an architecture for a cross-layer feedback framework for DTN protocols is discussed. Finally, initial simulation setup and results are given.

/pdf/a-machine-learning-concept-for-dtn-routing-41z0sqd8j3.pdf

A machine learning concept for DTN routing

There are hefty constraints levied on all space systems, and some of the most fundamental of them are purely physical. Much of the literature published on optical communications for space data systems focuses on the increased capabilities that optical communication systems have to offer, but ignores the practical infusion strategies of the technology into an operable network. Significant resources have been spent over the better half of the last century across multiple governments to establish robust radio frequency (RF) infrastructure around the world, and any new spacecraft will include provisions to utilize that investment. Therefore, every optical communication system flown will be a functional subset of a larger communication architecture dominated by RF. The key is to optimize the architecture and integrated components into a system that can be used to support hybridized RF and optical communications within the same asset, throughout diverse atmospheric (weather) and in-space conditions. The focus of this paper is on the physical realizability of a hybrid optical and RF communications system, and its ability to increase mission capability without imparting extra burden to the host spacecraft in the form of either increased physical mass or power requirements or additional operational requirements. The space applications may feature a hybrid RF/optical aperture this conjunction of a telescope and an antenna has been dubbed the teletenna. The teletenna is a single co-boresighted aperture that combines the radio and optical beams together into one physical package. The tradeoffs associated with the teletenna optimization are discussed, as well as the pointing requirements and strategies. Initial optical terminals on spacecraft have utilized Earth-based beacons Aerospace Technologist in Telecommunications, Optics and Photonics Branch (LCP), 21000 Brookpark Road/Mail Stop 54-1, AIAA and Technical Committee Member Aerospace Technologist in Telecommunications, High Frequency Branch (LCH), 21000 Brookpark Road/Mail Stop 54-1 Senior Aerospace Technologist in Telecommunications, Architectures, Networks and System Integration Branch, (LCA), 21000 Brookpark Road/Mail Stop 54-1 Aerospace Technologist in Data Systems, Information and Signal Processing Branch (LCI), 21000 Brookpark Road/Mail Stop 54-1 Research Aerospace Technologist in Data Systems, Networking and Architectures Branch (LCA), 21000 Brookpark Road/Mail Stop 54-1 Research Aerospace Technologist in Telecommunications, Information and Signal Processing Branch (LCI), 21000 Brookpark Road/Mail Stop 54-1 Aerospace Technologist in Mechanical Components, (LMT), 21000 Brookpark Road/Mail Stop 86-2 to establish beam tracking, but eventually beacon-less pointing will be developed. Investments in technologies to implement autonomous on-board navigation (and maneuvering) will permit a reduction in dependence on ground-based tracking, ranging, trajectory/orbit/attitude determination and maneuver planning support functions resulting in increased intelligence and autonomy of the complete payload. Innovative approaches include exploiting the optical communications terminal to perform navigational measurements such as star sighting, or a star tracker technology developed to process images at very high data rates. This paper discusses how such a beaconless strategy may be incorporated to work with the proposed hybrid communications system, and highlights the synergistic benefit to both communication and navigation functions for the purpose of decreased size, mass and power burden to users. A common constraint between both deep space and near Earth scenarios is mass. As elegant as a proposed optical communications solution may be, it must be shown that the final product is not so cumbersome as to outweigh the benefits, and that includes the necessary RF technology to complete the full communications payload. We establish a stringent goal of compressing the aforementioned highly capable hybrid RF/optical system into a package which installs within a comparable size, mass, and power envelope as the telecommunications subsystem of Mars Reconnaissance Orbiter (MRO). The goal of this paper is to demonstrate that with modern technology our goal is feasible and therefore has a place in today’s space data systems as well as building a bridge to tomorrow’s.

On the Physical Realizability of Hybrid RF and Optical Communications Platforms for Deep Space Applications

One of the greatest challenges in developing new space technology is in navigating the transition from ground based laboratory demonstration at Technology Readiness Level 6 (TRL-6) to conducting a prototype demonstration in space (TRL-7). This challenge is com- pounded by the relatively low availability of new spacecraft missions when compared with aeronautical craft to bridge this gap, leading to the general adoption of a low-risk stance by mission management to accept new, unproven technologies into the system. Also in consideration of risk, the limited selection and availability of proven space-grade components imparts a severe limitation on achieving high performance systems by current terrestrial technology standards. Finally from a space communications point of view the long duration characteristic of most missions imparts a major constraint on the entire space and ground network architecture, since any new technologies introduced into the system would have to be compliant with the duration of the currently deployed operational technologies, and in some cases may be limited by surrounding legacy capabilities. Beyond ensuring that the new technology is verified to function correctly and validated to meet the needs of the end users the formidable challenge then grows to additionally include: carefully timing the maturity path of the new technology to coincide with a feasible and accepting future mission so it flies before its relevancy has passed, utilizing a limited catalog of available components to their maximum potential to create meaningful and unprecedented new capabilities, designing and ensuring interoperability with aging space and ground infrastructures while simultaneously providing a growth path to the future. The International Space Station (ISS) is approaching 20 years of age. To keep the ISS relevant, technology upgrades are continuously taking place. Regarding communications, the state-of-the-art communication system upgrades underway include high-rate laser terminals. These must interface with the existing, aging data infrastructure. The High Data Rate Architecture (HiDRA) project is designed to provide networked store, carry, and forward capability to optimize data flow through both the existing radio frequency (RF) and new laser communications terminal. The networking capability is realized through the Delay Tolerant Networking (DTN) protocol, and is used for scheduling data movement as well as optimizing the performance of existing RF channels. HiDRA is realized as a distributed FPGA memory and interface controller that is itself controlled by a local computer running DTN software. Thus HiDRA is applicable to other arenas seeking to employ next-generation communications technologies, e.g. deep space. In this paper, we describe HiDRA and its far-reaching research implications.

/pdf/high-data-rate-architecture-hidra-43mokw03v9.pdf

Alan Hylton

Papers

VESGEN 2D: Automated, User‐Interactive Software for Quantification and Mapping of Angiogenic and Lymphangiogenic Trees and Networks

Selective inhibition of angiogenesis in small blood vessels and decrease in vessel diameter throughout the vascular tree by triamcinolone acetonide.

A machine learning concept for DTN routing

On the Physical Realizability of Hybrid RF and Optical Communications Platforms for Deep Space Applications

High Data Rate Architecture (HiDRA)