Search

Identification of lymph nodes (LN) in T2 Magnetic Resonance Imaging (MRI) is an important step performed by radiologists during the assessment of lymphoproliferative diseases. The size of the nodes play a crucial role in their staging, and radiologists sometimes use an additional contrast sequence such as diffusion weighted imaging (DWI) for confirmation. However, lymph nodes have diverse appearances in T2 MRI scans, making it tough to stage for metastasis. Furthermore, radiologists often miss smaller metastatic lymph nodes over the course of a busy day. To deal with these issues, we propose to use the DEtection TRansformer (DETR) network to localize suspicious metastatic lymph nodes for staging in challenging T2 MRI scans acquired by different scanners and exam protocols. False positives (FP) were reduced through a bounding box fusion technique, and a precision of 65.41\% and sensitivity of 91.66\% at 4 FP per image was achieved. To the best of our knowledge, our results improve upon the current state-of-the-art for lymph node detection in T2 MRI scans., Comment: Accepted at SPIE 2022

Francis Bacon popularized the idea that science is based on a process of induction by which repeated observations are, in some unspecified way, generalized to theories based on the assumption that the future resembles the past. This idea was criticized by Hume and others as untenable leading to the famous problem of induction. It wasn't until the work of Karl Popper that this problem was solved, by demonstrating that induction is not the basis for science and that the development of scientific knowledge is instead based on the same principles as biological evolution. Today, machine learning is also taught as being rooted in induction from big data. Solomonoff induction implemented in an idealized Bayesian agent (Hutter's AIXI) is widely discussed and touted as a framework for understanding AI algorithms, even though real-world attempts to implement something like AIXI immediately encounter fatal problems. In this paper, we contrast frameworks based on induction with Donald T. Campbell's universal Darwinism. We show that most AI algorithms in use today can be understood as using an evolutionary trial and error process searching over a solution space. In this work we argue that a universal Darwinian framework provides a better foundation for understanding AI systems. Moreover, at a more meta level the process of development of all AI algorithms can be understood under the framework of universal Darwinism., Comment: accepted to the BICA*AI 2021 conference at the 2021 IS4SI summit

Artificial intelligence has made great strides since the deep learning revolution, but AI systems still struggle to extrapolate outside of their training data and adapt to new situations. For inspiration we look to the domain of science, where scientists have been able to develop theories which show remarkable ability to extrapolate and sometimes predict the existence of phenomena which have never been observed before. According to David Deutsch, this type of extrapolation, which he calls "reach", is due to scientific theories being hard to vary. In this work we investigate Deutsch's hard-to-vary principle and how it relates to more formalized principles in deep learning such as the bias-variance trade-off and Occam's razor. We distinguish internal variability, how much a model/theory can be varied internally while still yielding the same predictions, with external variability, which is how much a model must be varied to accurately predict new, out-of-distribution data. We discuss how to measure internal variability using the size of the Rashomon set and how to measure external variability using Kolmogorov complexity. We explore what role hard-to-vary explanations play in intelligence by looking at the human brain and distinguish two learning systems in the brain. The first system operates similar to deep learning and likely underlies most of perception and motor control while the second is a more creative system capable of generating hard-to-vary explanations of the world. We argue that figuring out how replicate this second system, which is capable of generating hard-to-vary explanations, is a key challenge which needs to be solved in order to realize artificial general intelligence. We make contact with the framework of Popperian epistemology which rejects induction and asserts that knowledge generation is an evolutionary process which proceeds through conjecture and refutation., Comment: Accepted for a talk at BICA 2020 and for publication in Cognitive Systems Research

Pathological science occurs when well-intentioned scientists spend extended time and resources studying a phenomena that isn't real. Researchers who get caught up in pathological science are usually following the scientific method and performing careful experiments, but they get tricked by nature. The study of water has had several protracted episodes of pathological science, a few of which are still ongoing. We discuss four areas of pathological water science -polywater, the Mpemba effect, Pollack's "fourth phase" of water, and the effects of static magnetic fields on water. Some common water-specific issues emerge such as the contamination and confounding of experiments with dissolved solutes and nanobubbles. General issues also emerge such as imprecision in defining what is being studied, bias towards confirmation rather than falsification, and poor standards for reproducibility. We hope this work helps researchers avoid wasting valuable time and resources pursuing pathological science., Comment: to appear in the Springer book "Water in Biomechanical & Related Systems"

We present a novel method for small bowel segmentation where a cylindrical topological constraint based on persistent homology is applied. To address the touching issue which could break the applied constraint, we propose to augment a network with an additional branch to predict an inner cylinder of the small bowel. Since the inner cylinder is free of the touching issue, a cylindrical shape constraint applied on this augmented branch guides the network to generate a topologically correct segmentation. For strict evaluation, we achieved an abdominal computed tomography dataset with dense segmentation ground-truths. The proposed method showed clear improvements in terms of four different metrics compared to the baseline method, and also showed the statistical significance from a paired t-test., Comment: Accepted to MICCAI 2020

Current deep learning based segmentation models often generalize poorly between domains due to insufficient training data. In real-world clinical applications, cross-domain image analysis tools are in high demand since medical images from different domains are often needed to achieve a precise diagnosis. An important example in radiology is generalizing from non-contrast CT to contrast enhanced CTs. Contrast enhanced CT scans at different phases are used to enhance certain pathologies or organs. Many existing cross-domain image-to-image translation models have been shown to improve cross-domain segmentation of large organs. However, such models lack the ability to preserve fine structures during the translation process, which is significant for many clinical applications, such as segmenting small calcified plaques in the aorta and pelvic arteries. In order to preserve fine structures during medical image translation, we propose a patch-based model using shared latent variables from a Gaussian mixture model. We compare our image translation framework to several state-of-the-art methods on cross-domain image translation and show our model does a better job preserving fine structures. The superior performance of our model is verified by performing two tasks with the translated images - detection and segmentation of aortic plaques and pancreas segmentation. We expect the utility of our framework will extend to other problems beyond segmentation due to the improved quality of the generated images and enhanced ability to preserve small structures., Comment: Accepted to Medical Image Computing and Computer Assisted Intervention (MICCAI) 2020

Calcified plaque in the aorta and pelvic arteries is associated with coronary artery calcification and is a strong predictor of heart attack. Current calcified plaque detection models show poor generalizability to different domains (ie. pre-contrast vs. post-contrast CT scans). Many recent works have shown how cross domain object detection can be improved using an image translation model which translates between domains using a single shared latent space. However, while current image translation models do a good job preserving global/intermediate level structures they often have trouble preserving tiny structures. In medical imaging applications, preserving small structures is important since these structures can carry information which is highly relevant for disease diagnosis. Recent works on image reconstruction show that complex real-world images are better reconstructed using a union of subspaces approach. Since small image patches are used to train the image translation model, it makes sense to enforce that each patch be represented by a linear combination of subspaces which may correspond to the different parts of the body present in that patch. Motivated by this, we propose an image translation network using a shared union of subspaces constraint and show our approach preserves subtle structures (plaques) better than the conventional method. We further applied our method to a cross domain plaque detection task and show significant improvement compared to the state-of-the art method., Comment: accepted as a short paper in the 2020 Medical Imaging with Deep Learning (MIDL) conference

The ability to explain decisions made by AI systems is highly sought after, especially in domains where human lives are at stake such as medicine or autonomous vehicles. While it is often possible to approximate the input-output relations of deep neural networks with a few human-understandable rules, the discovery of the double descent phenomena suggests that such approximations do not accurately capture the mechanism by which deep neural networks work. Double descent indicates that deep neural networks typically operate by smoothly interpolating between data points rather than by extracting a few high level rules. As a result, neural networks trained on complex real world data are inherently hard to interpret and prone to failure if asked to extrapolate. To show how we might be able to trust AI despite these problems we introduce the concept of self-explaining AI. Self-explaining AIs are capable of providing a human-understandable explanation of each decision along with confidence levels for both the decision and explanation. For this approach to work, it is important that the explanation actually be related to the decision, ideally capturing the mechanism used to arrive at the explanation. Finally, we argue it is important that deep learning based systems include a "warning light" based on techniques from applicability domain analysis to warn the user if a model is asked to extrapolate outside its training distribution. For a video presentation of this talk see https://www.youtube.com/watch?v=Py7PVdcu7WY& ., Comment: 10pgs, 2 column format

The vertebral levels of the spine provide a useful coordinate system when making measurements of plaque, muscle, fat, and bone mineral density. Correctly classifying vertebral levels with high accuracy is challenging due to the similar appearance of each vertebra, the curvature of the spine, and the possibility of anomalies such as fractured vertebrae, implants, lumbarization of the sacrum, and sacralization of L5. The goal of this work is to develop a system that can accurately and robustly identify the L1 level in large heterogeneous datasets. The first approach we study is using a 3D U-Net to segment the L1 vertebra directly using the entire scan volume to provide context. We also tested models for two class segmentation of L1 and T12 and a three class segmentation of L1, T12 and the rib attached to T12. By increasing the number of training examples to 249 scans using pseudo-segmentations from an in-house segmentation tool we were able to achieve 98% accuracy with respect to identifying the L1 vertebra, with an average error of 4.5 mm in the craniocaudal level. We next developed an algorithm which performs iterative instance segmentation and classification of the entire spine with a 3D U-Net. We found the instance based approach was able to yield better segmentations of nearly the entire spine, but had lower classification accuracy for L1., Comment: Accepted for publication in Proceedings of SPIE 2020: Medical Imaging

The existence of the exclusion zone (EZ), a layer of water in which plastic microspheres are repelled from hydrophilic surfaces, has now been independently demonstrated by several groups. A better understanding of the mechanisms which generate EZs would help with understanding the possible importance of EZs in biology and in engineering applications such as filtration and microfluidics. Here we review the experimental evidence for EZ phenomena in water and the major theories that have been proposed. We review experimental results from birefringence, neutron radiography, nuclear magnetic resonance, and other studies. Pollack and others have theorized that water in the EZ exists has a different structure than bulk water, and that this accounts for the EZ. We present several alternative explanations for EZs and argue that Schurr's theory based on diffusiophoresis presents a compelling alternative explanation for the core EZ phenomenon. Among other things, Schurr's theory makes predictions about the growth of the EZ with time which have been confirmed by Florea et al. and others. We also touch on several possible confounding factors that make experimentation on EZs difficult, such as charged surface groups, dissolved solutes, and adsorbed nanobubbles., Comment: 14 pgs

The heat transfer properties of the organic molecular crystal ${\alpha}$-RDX were studied using three phonon-based thermal conductivity models. It was found that the widely used Peierls-Boltzmann model for thermal transport in crystalline materials breaks down for ${\alpha}$-RDX. We show this breakdown is due to a large degree of anharmonicity that leads to a dominance of diffusive-like carriers. Despite being developed for disordered systems, the Allen-Feldman theory for thermal conductivity actually gives the best description of thermal transport. This is likely because diffusive carriers contribute to over 95% of the thermal conductivity in ${\alpha}$-RDX. The dominance of diffusive carriers is larger than previously observed in other fully ordered crystalline systems. These results indicate than van-der Waals bonded organic crystalline solids conduct heat in a manner more akin to amorphous materials than simple atomic crystals., Comment: Presented at 2019 MRS Spring Meeting & Exhibit, Phoenix Arizona, USA Session QN04.08.02, https://www.mrs.org/spring2019/symposium-sessions/symposium-sessions-detail?code=QN04 8 pages, 3 figures

In the space of only a few years, deep generative modeling has revolutionized how we think of artificial creativity, yielding autonomous systems which produce original images, music, and text. Inspired by these successes, researchers are now applying deep generative modeling techniques to the generation and optimization of molecules - in our review we found 45 papers on the subject published in the past two years. These works point to a future where such systems will be used to generate lead molecules, greatly reducing resources spent downstream synthesizing and characterizing bad leads in the lab. In this review we survey the increasingly complex landscape of models and representation schemes that have been proposed. The four classes of techniques we describe are recursive neural networks, autoencoders, generative adversarial networks, and reinforcement learning. After first discussing some of the mathematical fundamentals of each technique, we draw high level connections and comparisons with other techniques and expose the pros and cons of each. Several important high level themes emerge as a result of this work, including the shift away from the SMILES string representation of molecules towards more sophisticated representations such as graph grammars and 3D representations, the importance of reward function design, the need for better standards for benchmarking and testing, and the benefits of adversarial training and reinforcement learning over maximum likelihood based training., Comment: 24 pages, new title, published in RSC MSDE

The number of scientific journal articles and reports being published about energetic materials every year is growing exponentially, and therefore extracting relevant information and actionable insights from the latest research is becoming a considerable challenge. In this work we explore how techniques from natural language processing and machine learning can be used to automatically extract chemical insights from large collections of documents. We first describe how to download and process documents from a variety of sources - journal articles, conference proceedings (including NTREM), the US Patent & Trademark Office, and the Defense Technical Information Center archive on archive.org. We present a custom NLP pipeline which uses open source NLP tools to identify the names of chemical compounds and relates them to function words ("underwater", "rocket", "pyrotechnic") and property words ("elastomer", "non-toxic"). After explaining how word embeddings work we compare the utility of two popular word embeddings - word2vec and GloVe. Chemical-chemical and chemical-application relationships are obtained by doing computations with word vectors. We show that word embeddings capture latent information about energetic materials, so that related materials appear close together in the word embedding space., Comment: accepted for publication in the 2019 NTREM proceedings

Due to its high computational speed and accuracy compared to ab-initio quantum chemistry and forcefield modeling, the prediction of molecular properties using machine learning has received great attention in the fields of materials design and drug discovery. A main ingredient required for machine learning is a training dataset consisting of molecular features\textemdash for example fingerprint bits, chemical descriptors, etc. that adequately characterize the corresponding molecules. However, choosing features for any application is highly non-trivial. No "universal" method for feature selection exists. In this work, we propose a data fusion framework that uses Independent Vector Analysis to exploit underlying complementary information contained in different molecular featurization methods, bringing us a step closer to automated feature generation. Our approach takes an arbitrary number of individual feature vectors and automatically generates a single, compact (low dimensional) set of molecular features that can be used to enhance the prediction performance of regression models. At the same time our methodology retains the possibility of interpreting the generated features to discover relationships between molecular structures and properties. We demonstrate this on the QM7b dataset for the prediction of several properties such as atomization energy, polarizability, frontier orbital eigenvalues, ionization potential, electron affinity, and excitation energies. In addition, we show how our method helps improve the prediction of experimental binding affinities for a set of human BACE-1 inhibitors. To encourage more widespread use of IVA we have developed the PyIVA Python package, an open source code which is available for download on Github.

The microscale thermal transport properties of $\alpha$RDX are believed to be major factors in the initiation process. In this study we present a thorough examination of phonon properties which dominate energy storage and transport in $\alpha$RDX. The phonon lifetimes are determined for all phonon branches, revealing the characteristic time scale of energy transfer amongst phonon modes. The phonon parameters also serve as inputs to a full Brillouin zone three dimensional phonon transport simulation in the presence of a hotspot. In addition to identifying the phonon mode contributions to thermal transport, and as N-N bond breaking is integral to disassociation, we identify phonon modes corresponding to large N-N bond stretch analyzing the manner in which these modes store and transfer energy., Comment: Preprint for publication in the Proceedings of the 16th International Detonation Symposium

This short report details the mathematical properties of the stretched exponential function and some of its applications in materials science. G(tau) distributions for different values of the stretching parameter beta are provided.

In this work, we discuss use of machine learning techniques for rapid prediction of detonation properties including explosive energy, detonation velocity, and detonation pressure. Further, analysis is applied to individual molecules in order to explore the contribution of bonding motifs to these properties. Feature descriptors evaluated include Morgan fingerprints, E-state vectors, a custom "sum over bonds" descriptor, and coulomb matrices. Algorithms discussed include kernel ridge regression, least absolute shrinkage and selection operator ("LASSO") regression, Gaussian process regression, and the multi-layer perceptron (a neural network). Effects of regularization, kernel selection, network parameters, and dimensionality reduction are discussed. We determine that even when using a small training set, non-linear regression methods may create models within a useful error tolerance for screening of materials., Comment: Preprint for publication in the Proceedings of the 16th International Detonation Symposium

It is now established that nuclear quantum motion plays an important role in determining water's hydrogen bonding, structure, and dynamics. Such effects are important to include in density functional theory (DFT) based molecular dynamics simulation of water. The standard way of treating nuclear quantum effects, path integral molecular dynamics (PIMD), multiplies the number of energy/force calculations by the number of beads required. In this work we introduce a method whereby PIMD can be incorporated into a DFT simulation with little extra cost and little loss in accuracy. The method is based on the many body expansion of the energy and has the benefit of including a monomer level correction to the DFT energy. Our method calculates intramolecular forces using the highly accurate monomer potential energy surface developed by Partridge-Schwenke, which is cheap to evaluate. Intermolecular forces and energies are calculated with DFT only once per timestep using the centroid positions. We show how our method may be used in conjunction with a multiple time step algorithm for an additional speedup and how it relates to ring polymer contraction and other schemes that have been introduced recently to speed up PIMD simulations. We show that our method, which we call "monomer PIMD", correctly captures changes in the structure of water found in a full PIMD simulation but at much lower computational cost., Comment: published in PCCP

We present a proof of concept that machine learning techniques can be used to predict the properties of CNOHF energetic molecules from their molecular structures. We focus on a small but diverse dataset consisting of 109 molecular structures spread across ten compound classes. Up until now, candidate molecules for energetic materials have been screened using predictions from expensive quantum simulations and thermochemical codes. We present a comprehensive comparison of machine learning models and several molecular featurization methods - sum over bonds, custom descriptors, Coulomb matrices, bag of bonds, and fingerprints. The best featurization was sum over bonds (bond counting), and the best model was kernel ridge regression. Despite having a small data set, we obtain acceptable errors and Pearson correlations for the prediction of detonation pressure, detonation velocity, explosive energy, heat of formation, density, and other properties out of sample. By including another dataset with 309 additional molecules in our training we show how the error can be pushed lower, although the convergence with number of molecules is slow. Our work paves the way for future applications of machine learning in this domain, including automated lead generation and interpreting machine learning models to obtain novel chemical insights., Comment: published in Scientific Reports

We critically review the literature on the Debye absorption peak of liquid water and the excess response found on the high frequency side of the Debye peak. We find a lack of agreement on the microscopic phenomena underlying both of these features. To better understand the molecular origin of Debye peak we ran large scale molecular dynamics simulations and performed several different distance-dependent decompositions of the low frequency dielectric spectra, finding that it involves processes that take place on scales of 1.5-2.0 nm. We also calculated the $k$-dependence of the Debye relaxation, finding it to be highly dispersive. These findings are inconsistent with models that relate Debye relaxation to local processes such as the rotation/translation of molecules after H-bond breaking. We introduce the spectrumfitter Python package for fitting dielectric spectra and analyze different ways of fitting the high frequency excess, such as including one or two additional Debye peaks. We propose using the generalized Lydanne-Sachs-Teller (gLST) equation as a way of testing the physicality of model dielectric functions. Our attempts at fitting the experimental spectrum using the gLST relation as a constraint indicate that the traditional way of fitting the excess response with secondary and tertiary Debye relaxations is problematic. All of our work is consistent with the recent theory of Popov et al. (2016) that Debye relaxation is due to the migration of Bjerrum-like defects in the hydrogen bond network. Under this theory, the mechanism of Debye relaxation in liquid water is similar to the mechanism in ice, but the heterogeneity and power-law dynamics of the H-bond network in water results in excess response on the high frequency side of the peak., Comment: Phys. Chem. Chem. Phys., 2017, Accepted Manuscript

The local structure of liquid water as a function of temperature is a source of intense research. This structure is intimately linked to the dynamics of water molecules, which can be measured using Raman and infrared spectroscopies. The assignment of spectral peaks depends on whether they are collective modes or single molecule motions. Vibrational modes in liquids are usually considered to be associated to the motions of single molecules or small clusters. Using molecular dynamics simulations we find dispersive optical phonon-like modes in the librational and OH stretching bands. We argue that on subpicosecond time scales these modes propagate through water's hydrogen bond network over distances of up to two nanometers. In the long wavelength limit these optical modes exhibit longitudinal-transverse splitting, indicating the presence of coherent long range dipole-dipole interactions, as in ice. Our results indicate the dynamics of liquid water have more similarities to ice than previously thought., Comment: accepted into Nat. Comm

We present a critical comparison of the dielectric properties of three models of water - TIP4P/2005, TIP4P/2005f and TTM3F. Dipole spatial correlation is measured using the distance dependent Kirkwood function along with one dimensional and two dimensional dipole correlation functions. We find that the introduction of flexibility alone does not significantly affect dipole correlation and only affects $\varepsilon(\omega)$ at high frequencies. By contrast the introduction of polarizability increases dipole correlation and yields a more accurate $\varepsilon(\omega)$. Additionally the introduction of polarizability creates temperature dependence in the dipole moment even at fixed density, yielding a more accurate value for $d \varepsilon / d T$ compared to non-polarizable models. To better understand the physical origin of the dielectric properties of water we make analogies to the physics of polar nanoregions in relaxor ferroelectric materials. We show that $\varepsilon(\omega,T)$ and $\tau_D(T)$ for water have striking similarities with relaxor ferroelectrics, a class of materials characterized by large frequency dispersion in $\varepsilon(\omega,T)$, Vogel-Fulcher-Tamann behaviour in $\tau_D(T)$, and the existence of polar nanoregions., Comment: 13 pages, 16 figures. Accepted in JCP. Updated figs 1, 3, & 4, added acknowledgement