No description available.

No description available.

Explainability is a key component in building trust for computer calculated predictions when they are applied to areas with influence over individual people. This bachelor thesis project report focuses on the explanation regarding the decision making process of the machine learning method Logistic Regression when predicting mortality. The aim is to present theoretical information about the predictive model as well as an explainable interpretation when applied on the clinical MIMIC-III database. The project found that there was a significant difference between particular features considering the impact of each individual feature on the classification. The feature that showed the greatest impact was the Glasgow Coma Scale value, which could be proven through the fact that a good classifier could be constructed with only that and one other feature. An important conclusion from this study is that a great focus should be enforced early in the implementation process when the features are selected. In this specific case, when medical artificial intelligence is implemented, medical expertise is desired in order to make a good feature selection. / Förklarbarhet är en viktig komponent för att skapa förtroende för datorframtagna prognoser när de appliceras på områden som påverkar individuella personer. Denna kandidatexamensarbetesrapport fokuserar på förklarandet av beslutsprocessen hos maskininlärningsmetoden Logistic Regression när dödlighet ska förutsägas. Målet är att presentera information om den förutsägande modellen samt en förklarbar tolkning av resultaten när modellen appliceras på den kliniska databasen MIMIC-III. Projektet fann att det fanns signifikanta skillnader mellan särskilda egenskaper med hänsyn till den påverkan varje enskild egenskap har på klassificeringen. Den egenskapen som visade ha störst inverkan var Glascow Coma Scale värdet, vilket kunde visas via det faktum att en god klassificerare kunde konstrueras med endast den och en annan egenskap. En viktig slutsats av denna studie är att stort fokus bör läggas tidigt i implementationsprocessen då egenskaperna väljs. I detta specifika fall, då medicinsk artificiell intelligens implementeras, krävs medicinsk expertis för att göra ett gott egenskapsurval. / Kandidatexjobb i elektroteknik 2022, KTH, Stockholm

Machine learning has the potential of significantly improving daily operations in health care institutions but many persistent barriers are to be faced in order to ensure its wider acceptance. Among such obstacles are the accuracy and reliability. For a decision support system to be entrusted by the medical staff in clinical situations, it must perform with an accuracy comparable to or surpassing that of human medics, as well ashaving a universal applicability and not being subject to any bias. In this paper the MIMIC-IV Clinical Database will be utilized in order to: (1) Predict patient mortality and its associated risk factors in intensive care units (ICU) and: (2) Assess the reliability of utilizing the database as a basis for a clinical decision system. The cohort consisted of 523,740 hospitalizations, matched with each respective admitting diagnoses in ICD-9 format. The diagnoses were then converted from code to text-format, with the most frequently occurring factors (words) observed in deceased and surviving patients being analyzed with an Natural language Processing (NLP) algorithm. The results concluded that many of the observed risk factors were self-evident while others required further explanation, and that the performance was highly by selection of hyperparameters. Finally, the MIMIC-IV database can serve as a stable foundation for a clinical decision system but its reliability and universality shall also be taken into consideration. / Maskininlärninstekniker har en stor potential att gynna sjukvården men står inför ett flertal hinder för att fullständigt kunna tillämpas. Framförallt bör modellernas tolkningsbarhet och reproducerbarhet beaktas. För att att ett kliniskt beslutstodssystem skall vara fullständigt anförtrott av sjukvårdspersonal måste det kunna prestera med en jämförbar eller högre träffsäkerhet än sjukvårdspersonal, samt kunna tillämpas i åtskilliga sammanhang utan någon subjektivitet. Syftet med denna studie är att: (1) Förutspå patientdödsfall i intensivvårdsavdelningar och utreda dess riskfaktorer genom journalförd information från databasen MIMIC-IV och: 2) Bedöma databasens tillförlitlighet som underlag för ett kliniskt beslutstödssystem. Kohorten bestod av 523,740 insjuknanden som matchades med de diagnoser som ställdes vid deras sjukhusintag. Eftersom diagnoserna inskrevs i ICD-9-format omvandlades dessa till ord och de mest förekommande faktorerna (orden) för avlidna och överlevande patienter analyserades med en NLP-model (Natural Language Processing). Resultaten konkluderade att många av de förutspådda riskfaktorerna var uppenbara medan andra krävde ytterligare klargöranden. Dessutom kunde val av hyperparametrar stort påverka modellens kvalitet. MIMIC-IV-databasen kan utgöra ett gediget underlag för ett kliniskt beslutsystem men dess tillförlitlighet och relevans bör även tas i beaktande. / Kandidatexjobb i elektroteknik 2022, KTH, Stockholm

No description available.

Pin1 (protein interacting with NIMA 1) is a peptidyl-prolyl isomerase involved in mitosis. As a potential anti-cancer drug target, Pin1 interacts and regulates the activity of an increasing number of cell cycle enzymes by an unknown mechanism. These cell cycle enzymes include Cdc25, Cdc27, Cyclin D1, Myt1, Wee1, NIMA, Cdc2, Plk1 and c-Myc. Recent research has revealed that Pin1 is overexpressed in a variety of cancer cell lines and Pin1 inhibitors inhibit proliferation activity of several cancer cells overexpressing Pin1. The most potent Pin1 inhibitors identified so far are in the micromolar range and no pharmacophore has been identified. In order to assist the understanding of the biological function of Pin1 using molecular probes, two amide isosteres of Ser-<i>trans</i>-Pro and Ser-<i>cis</i>-Pro dipeptides were designed and stereoselectively synthesized. The conformationally locked Ser–<i>trans</i>–Pro mimic, Boc-SerΨ[(<i>E</i>)CH=C]Pro–OH, was synthesized through the use of an Ireland-Claisen [3,3]-sigmatropic rearrangement in nine steps with 13% overall yield from a serine derivative. The Ser-<i>cis</i>-Pro mimic, Boc-SerΨ[(<i>Z</i>)CH=C]Pro–OH, was synthesized through the use of a Still-Wittig [2,3]-sigmatropic rearrangement in 11 steps with an overall yield of 20% from the same starting material. Conformationally locked peptidomimetics, including two exactly matched peptidomimetics, Ac–Phe–Phe–pSer–Ψ(<i>E</i>)CH=C]Pro–Arg–NH2 and Ac–Phe–Phe–pSer–Ψ[(<i>Z</i>)CH=C]Pro–Arg–NH2, were synthesized from these Ser-Pro isosteres using Fmoc SPPS. A protocol for in vitro Pin1 inhibition assay was established for measuring the inhibition constant for these peptidomimetics. A conformationally locked cis peptidomimetic inhibits Pin1 with a <i>K</i>_i of 1.7 <i>μ</i>M, 23-fold more potent than its trans counterpart, illustrating the preference of Pin1 for a cis amide bond in its PPIase domain. The A2780 ovarian cancer cell antiproliferation activity of these peptidomimetics parallels their respective Pin1 inhibition data. This research provides a start toward more drug-like Pin1 inhibitor design. Gly–<i>trans</i>–Pro isosteres were synthesized using the Ireland-Claisen route. The construction of a non-peptidic (Z)-alkene library for Pin1 inhibition was attempted using the Ser-<i>cis</i>-Pro mimic, Boc—SerΨ[(Z)CH=C]Pro–OH as the core. / Ph. D.

Pin1 is a peptidyl prolyl isomerase (PPIase) enzyme with two domains, the catalytic domain and the WW domain. Both domains specifically bind pSer/pThr–Pro motifs. Pin1 plays an important role in regulating the cell cycle, and it is involved in many diseases, such as cancer, HIV-1, Alzheimer's disease, asthma, hepatitis B, and rheumatoid arthritis. Pin1 is a very promising target for new drug development. Three stereoisomers: (R,S)-, (S,R)- and (S,S)-Ac–pSer–Ψ[(Z)CH=C]–Pip–2-(2-naphthyl)ethylamine were synthesized as inhibitors binding to the Pin1 catalytic domain. The (R,S)- and (S,R)-isomers were synthesized via a 13-step route, with overall yields of 2.0% and 1.4%, respectively. The newly formed stereogenic center in the piperidyl ring was introduced by a Luche reduction, followed by a stereoselective [2,3]-Still-Wittig rearrangement. The configuration of the stereocenter was determined by NOESY of a bicyclic derivative. The (Z)- to (E)-alkene ratio in the rearrangement was (5.5:1). The (S,S)-isomer was obtained as the epimerized by-product resulting from the (S,R)-isomer in the Na/NH3 deprotection step. The IC50 values for Pin1 inhibition were: 52, 85, and 141 μM, respectively. We concluded that in this Z-alkene isostere, the R-configuration would be preferred at both stereogenic centers, as mimics of L-Ser and L-Pip, to improve the affinity. Combinatorial chemistry is a powerful method to discover biologically active compounds, and solid-phase synthesis is most commonly used to synthesize combinatorial libraries. To identify ligands for the Pin1 WW domain, a library, R1CO–pSer–Pro–NHR2, was designed. A new solid-phase phosphorylating reagent (SPPR) containing a phosphoramidite function was synthesized in one step from commercially available Wang resin. The SPPR was applied in the preparation of a designed library through parallel synthesis. The library contained 357 members (17 Ã 21), and was screened by an enzyme-linked enzyme binding assay (ELEBA). The best hits were resynthesized, and the competitive dissociation constants, Kd-rel, were measured by ELEBA, with a Kd-rel value of 130 μM for the best ligand. The absolute dissociation constants will be measured by our collaborator, Prof. Jefferey Peng, University of Notre Dame, using NMR methods. Besides the identification of the Pin1 WW domain ligands, I created a practical method for solid-phase synthesis of phosphopeptides. / Ph. D.

Em interações humano-robô ainda existem muitas limitações a serem superadas referentes à provisão de uma comunicação natural quanto aos sentidos humanos. A capacidade de interagir com os seres humanos de maneira natural em contextos sociais (pelo uso da fala, gestos, expressões faciais, movimentos do corpo) é um ponto fundamental para garantir a aceitação de robôs em uma sociedade de pessoas não especialistas em manipulação de engenhos robóticos. Outrossim, a maioria dos robôs existentes possui habilidades limitadas de percepção, cognição e comportamento em comparação com seres humanos. Nesse contexto, este projeto de pesquisa investigou o potencial da arquitetura robótica do humanoide NAO, no tocante à capacidade de realizar interações com seres humanos através de imitação de movimentos do corpo de uma pessoa e pelo controle do robô. Quanto a sensores, foi utilizado um sensor câmera não-intrusivo de profundidade incorporado no dispositivo Kinect. Quanto às técnicas, alguns conceitos matemáticos foram abordados para abstração das configurações espaciais de algumas junções/membros do corpo humano essas configurações foram capturadas por meio da utilização da biblioteca OpenNI. Os experimentos realizados versaram sobre a imitação e o controle do robô por meio da avaliação de vários usuários. Os resultados desses experimentos revelaram um desempenho satisfatório quanto ao sistema desenvolvido / In human-robot interactions there are still many restrictions to overcome regarding the provision of a communication as natural to the human senses. The ability to interact with humans in a natural way in social contexts (the use of speech, gestures, facial expressions, body movements) is a key point to ensure the acceptance of robots in a society of people not specialized in manipulation of robotic devices. Moreover, most existing robots have limited abilities of perception, cognition and behavior in comparison with humans. In this context, this research project investigated the potential of the robotic architecture of the NAO humanoid robot, in terms of ability to perform interactions with humans through imitation of body movements of a person and the robot control. As for sensors, we used a non-intrusive sensor depth-camera built into the device Kinect. As to techniques, some mathematical concepts were discussed for abstraction of the spatial configurations of some joints/members of the human body these configurations were captured through the use of the OpenNI library. The performed experiments were about imitation and the control of the robot through the evaluation of various users. The results of these experiments showed a satisfactory performance for the developed system

How are neuromodulatory networks organised to adapt sensory discrimination for different contexts? I hypothesised that neurons within a sensory circuit express different neuromodulatory receptors for differential modulation. Here I aimed to use the simple and genetically amenable Drosophila larval Mushroom Body (MB) calyx, a higher order processing area involved in learned odour discrimination, as a model to map octopamine (OA) neuromodulatory circuitry. I first identified olfactory projection neurons (PNs), a GABAergic feedback neuron and cholinergic extrinsic neurons as putative postsynaptic partners to OA neurons in the MB calyx using GFP reconstitution across synaptic partners. Next, I used novel EGFP-tagged OA receptors generated from recombination-mediated cassette exchange with MiMIC insertions in receptor genes to visualise endogenous expression patterns of OA receptors. Most notably, this is the first report of α2-adrenergic-like OA receptor localisation in any insect. For the first time, I showed that the α1-adrenergic-like OAMB localised to PN presynaptic terminals in the calyx; while Octβ1R localised diffusely in the calyx, resembling the innervation pattern of MB neuron dendrites. I detected EGFP-tagged Octα2R and Octβ2R in some PN cell bodies but not in neuron terminals - suggesting that Octα2R and Octβ2R may be expressed in some PNs, provided the misfolded fusion proteins are retained in the cell bodies of the neurons they are normally expressed in. Furthermore, I found that Octα2R and GABAAR fusion proteins localised to OA cell bodies but not to neuronal terminals, suggesting that OA neurons are subjected to inhibition, again given that these are not artefacts of the fusion proteins. To obtain tools to study OA modulation in the larval calyx, I then confirmed the expression patterns of driver lines that more specifically labelled calyx-innervating OA and extrinsic neurons, and tested the efficacy of three OAMB receptor knockdown lines. This initial attempt of mapping OA receptors, while subjected to further verification and development, is consistent with my hypothesis that a single neuromodulatory source can regulate multiple neuronal types in the same circuit through the distribution of different types of neuromodulatory receptors. This provides a new perspective in how the anatomical organisation of neuromodulation within a sensory network may translate to flexible outputs.

<p>A molecular system capable of photoinduced water splitting is an attractive approach to solar energy conversion. This thesis deals with the functional characterization of molecular building blocks for the three principal functions of such a molecular system: Photoinduced accumulative charge separation, catalytic water oxidation, and catalytic proton reduction. </p><p>Systems combining a ruthenium-trisbipyridine photosensitizer with multi-electron donors in form of dinuclear ruthenium or manganese complexes were investigated in view of the rate constants of electron transfer and excited state quenching. The kinetics were studied in the different oxidation states of the donor unit by combination of electrochemistry and time resolved spectroscopy. The rapid excited state quenching by the multi-electron donors points to the importance of redox intermediates for efficient accumulative photooxidation of the terminal donor.</p><p>The redox behavior of manganese complexes as mimics of the water oxidizing catalyst in the natural photosynthetic reaction center was studied by electrochemical and spectroscopic methods. For a dinuclear manganese complex ligand exchange reactions were studied in view of their importance for the accumulative oxidation of the complex and its reactivity towards water. With the binding of substrate water, multiple oxidation in a narrow potential range and concomitant deprotonation of the bound water it was demonstrated that the manganese complex is capable of mimicking multiple aspects of photosynthetic water oxidation.</p><p>A dinuclear iron complex was investigated as biomimetic proton reduction catalyst. The complex structurally mimics the active site of the iron-only hydrogenase enzyme and was designed to hold a proton on the bridging ligand and a hydride on the iron centers. Thermodynamics and kinetics of the protonation reactions and the electrochemical behavior of the different protonation states were studied in view of their potential catalytic performance.</p>