Non-cardiac surgeries conducted globally each year often lead to cardiac complications,with myocardial injury commonly occurring within 30 days post-surgery. This thesis investigates the correlation between intra-operative signals and myocardial injuryusing machine learning models. The study focuses on analyzing intra-operative STelevation, heart rate, diastolic blood pressure, and pulse pressure variation. Multiplemachine learning models, including decision tree and random forest classifiers, weredeveloped and evaluated using two approaches: the sequence of event times and comprehensive event features.The results indicate that intra-operative physiological signals are valuable predictorsof myocardial injury, with random forest models generally outperforming decision treemodels. However, limited and unbalanced data posed challenges, affecting model performance variability. This research establishes a framework for enhancing predictivemodels and monitoring strategies to improve patient outcomes. / Icke-kirurgiska operationer som genomförs globalt varje år leder ofta till hjärtrelaterade komplikationer, där myokardskada vanligtvis uppstår inom 30 dagar efteroperationen. Denna avhandling undersöker sambandet mellan intraoperativa signaleroch myokardskada med hjälp av maskininlärningsmodeller. Studien fokuserar påatt analysera intraoperativa ST-höjning, hjärtfrekvens, diastoliskt blodtryck ochpulstrycksvariation. Flera maskininlärningsmodeller, inklusive beslutsträd- ochrandom forest-klassificerare, utvecklades och utvärderades med hjälp av två tillvägagångssätt: sekvensen av händelsetider och omfattande händelseegenskaper.Resultaten visar att intraoperativa fysiologiska signaler är värdefulla prediktorer förmyokardskada, där random forest-modeller generellt presterar bättre än beslutsträdmodeller. Begränsad och obalanserad data utgjorde dock utmaningar som påverkademodellens prestandavariabilitet. Denna forskning etablerar en ram för att förbättraprediktiva modeller och övervakningsstrategier för att förbättra patientutfall.

When blood sampling is performed, problems and complications occasionally occur. Sometimes the blood sampling fails. Statistically, blood sampling fails at the first attempt in 28 percent of all cases for adults. For children, the corresponding rate is 44 percent. The project was based on the need for aids to perform blood sampling. The need was identified and recognised by the taskmasters Frida Svensson and Lisa Eliasson during their degree project in 2021. The project group’s aim was to continue the taskmasters work developing a medical technology product that facilitates blood sampling and to bring the product one step closer to the market. Text removed due to confidentiality. The text mentions the functionality of the product. The product's effect is to increase the chance of succeeding with the blood sampling at the first try. This not only increases the patient’s comfort, but also is timesaving and decreases the need of disposable materials. In the long run this contributes to large financial savings for the healthcare department. The result of the project was a prototype that embodies the product Blood Simpling II.  The product’s functions reduce the risk of complications when blood sampling is performed. At the same time, the product increases the chance of succeeding with the blood sampling at the first attempt. The project group’s successful result resulted in its purpose of bringing the product one step closer to the market. Text removed due to confidentiality. The text mentions the functionality of the product. / Vid blodprovstagning förekommer det att problem och komplikationer uppstår. Det händer att blodprovstagningen helt misslyckas. Statistiskt sett misslyckas blodprovstagningen på första försöket för vuxna i 28 procent av fallen och för barn är motsvarande siffra 44 procent. Projektet utgick från det behov av hjälpmedel att utföra blodprovstagning som uppdragsgivarna Frida Svensson och Lisa Eliasson identifierade under sitt examensarbete från 2021. Syftet var att fortsatta deras arbete att utveckla en medicinteknisk produkt som underlättar vid blodprovstagning och föra den närmare marknaden. I syfte att föra produkten närmare marknaden var det projektgruppens mål att vidareutveckla produkten. Projektgruppen skulle färdigställa komponenterna var för sig och sedan sätta samman dem till en produkt. Produktens effekt är att den ökar chansen att lyckas med blodprovtagningen redan på första försöket. Det gör att det inte bara blir behagligare för patienten utan det sparar även på värdefull tid och engångsmaterial, vilket i längden leder till stora finansiella besparingar för vården. Resultatet av projektet blev en prototyp som gestaltar produkten Blood Simpling II vars funktioner gör att risken att komplikationer uppstår vid venös blodprovstagning minskar och samtidigt ökar produkten chansen att blodprovstagningen lyckas på första försöket. Genom det resultat som projektgruppen lyckades med uppfyllde de sitt syfte att föra produkten ett steg närmare marknaden. Text borttagen på grund av sekretess. Texten nämner produktens funktionalitet vars nyhetsvärde skyddas för att bevara patenteringsmöjligheten.

Atopic dermatitis (AD) and other atopic diseases are strongly related to skinbarrier dysfunction, a biomeasure which has limited and unsatisfactory assessmenttechniques. Machine learning (ML) powered electrical impedance spectroscopy(EIS) has been shown to differentiate defective barrier function in mice and adults. Techniques such as principal component analysis (PCA) andsupport vector classifiers (SVC) can be used as ML tools to evaluate EIS measurements. EIS measurements taken on unaffected skin of children aged between 4 monthsand 3 years were collected and analysed using the children’s AD status. Measurements were grouped into one of four groups based on this AD status; No AD (No AD was developed up until 2 years of age), Pre AD (measurements takenbefore the onset of AD), AD remission (measurements taken after the onset ofAD) and AD flare (measurements taken during active AD). A SVC model was trained on the raw EIS measurement data to distinguish measurements from twoof the binarized AD status groups; AD flare and No AD. An additional SVC model was trained on the No AD group, distinguishing measurements based onbinarized age groups: measurements taken at 4 months against measurements taken at 3 years of age. The AD model tested on AD flare against No AD within the test set yielded AUC of 0.92, with a sensitivity and specificity of 89.29% and 88.89% respectively. When testing on all groups of the test set, Pre AD and AD remission groups had group means between the AD flare and No AD groups. No data bias against age was detected in the model. The results of the age model showed that age could be chronologically identified by the age model. The AD model was able to differentiate active AD children from children never experiencing AD symptoms on visually unaffected skin and in turn detecting skin barrier dysfunction. Separate studies would need to be conducted to test the predictive power and external validity of the model. Age is a significant factor to consider when designing ML models using EIS data in children, with proper balance in the training set a data bias within the model can be avoided. EIS is a versatile technique due to its data rich nature. Machine learning powered electrical impedance spectroscopy measurements are able to detect skin barrier dysfunction. Age is a significant factor when measuring EIS on children, but can be managed. / Atopisk dermatit (AD) och andra atopiska sjudomar är starkt kopplade tillhudbarrärdefekter, en hudegenskap där diagnosverktygen är begränsade och otillräckliga. Maskininlärning (ML) i kombination med elektrisk impedansspektroskopi (EIS) har visats att kunna differentiera defekta hudbarrärer i möss och hos vuxna människor. Tekniker såsom principalkomponentanalys (PCA) och stödvektormaskiner (SVM) kan användas som ML verktyg för att utvärdera EIS-mätningar. Statistiska analysverktyg såsom reciever operator characteristics (ROC) tillsammans med arean under kurvan (AUC) och andra precisionsmått kan användas för att utvärdera ML modeller. EIS-mätningar utförda på opåverkad hud hos barn mellan 4 månader och 3 års ålder samlades och analyserades med hjälp av barnens atopiska status. Samtliga mätningarna grupperades i en av totalt fyra grupper baserade på deras atopiska status; Ingen AD (Ingen AD hade utvecklats upp till 2 års ålder), Pre AD (mätningarna togs innan sjukdomsförloppet), AD remission (mätningarna togs efter sjukdomsförloppet) och AD aktiv (mätningarn togs under sjukdomsförloppet). En SVM modell tränades på EIS-rådata för att särskilja mellan två av grupperna binärt; AD aktiv och ingen AD. En ytterliggare SVM modell tränades på endast ingen AD-gruppen för att särskilja mellan mätningarna tagna vid åldrarna 4 månader och 3 år, omgjorta till binära grupper.  AD modellen testades med en jämförelse mellan AD aktiv-gruppen och ingen AD-gruppen med test data och resulterade i en AUC på 0.92, med en respektive sensitivitet och specificitet på 89.29\% samt 88.89\%. Modellen testades även på de ytterliggare grupperna i test datan, Pre AD och AD remission som båda hade gruppmedelvärden som låg mellan AD aktiv-gruppens och ingen AD-gruppens. Inga systematiska fel med avseende på barnens ålder hittades i AD-modellen. Resultaten från ålders-modellen visade på att modellen kunde rangordna mätningarna baserat på ålder. AD-modellen kunde differentiera barn med aktiv AD mot barn som aldrig hade uppvisat atopiska symptom på opåverkad hud, vilket innebär en differentiering av hudbarrärdefekter. Separata studier är nödvändiga för att testa prediktiv prestanda hos modellen samt dess externa validitet. Ålder är en signifikant faktor när ML modeller med EIS-data ska designas, men med tillräcklig balans i träningsdatan kan systematiska fel förhindras. EIS är en versatil teknik med avseense på dess datakomplexitet. EIS tillsammans med maskininlärning kan särskilja defekta hudbarriärer hos barn. Ålder är en signifikant faktor när EIS mäts på barn, men kan hanteras.

Randomised controlled trials (RCT) are considered the gold standard for assessing the efficacy and safety of medical interventions. However, RCTs face unique challenges when applied to medical technologies, such as issues related to timing of assessment, eligible population, acceptability, blinding, choice of comparator group, and consideration for learning curves. To address these challenges, this thesis explores the adaptation of the model-informed drug development (MIDD) approach to the field of medical technology, using a case study on transurethral microwave thermotherapy (TUMT). The research employs non-linear mixed- effects (NLME) modelling and D-optimal design to optimise study designs and improve the reliability and efficiency of clinical trials. The impact of different sampling times, sample sizes, and learning curves on effect size estimates is analysed. The results show that optimising sampling points and sizes significantly improves the precision and reliability of effect size estimates and describes how MIDD can be a useful tool for this purpose. The study also highlights the limitations of the TUMT study design, suggesting ways in which the model-based approach could offer more robust and reliable clinical evidence generation. This research highlights the potential of the MIDD approach to streamline the medical technology clinical development process, enhance the quality of evidence, and address its inherent complexities. Future work should expand on these findings by exploring more complex error models and additional study designs and its related aspects. / Randomiserade kontrollerade studier (RCT) anses vara standard för att bedöma effekt och säkerhet i kliniska interventionsstudier. RCT:er står dock inför unika utmaningar när de tillämpas på medicinteknik såsom utmaningar relaterade till tidpunkt för bedömning, rekrytering av lämpliga studiedeltagare, acceptans, blindning, val av jämförelsegrupp och hänsyn till inlärningskurvor. För att hantera dessa utmaningar undersöker denna avhandling anpassningen av modellinformerad läkemedelsutveckling (MIDD) till området medicinteknik, med hjälp av en fallstudie om transuretral mikrovågstermoterapi (TUMT). I arbetet tillämpas icke-linjär, hierarkisk (NLME) modellering och D-optimal design för att optimera studiedesigner och förbättra tillförlitligheten i kliniska prövningar. Effekten av olika observationstider, antal studiedeltagare och inlärningskurvor på estimeringen av effektstorlek analyseras. Resultaten visar att optimering av observationstidpunkter och studiestorlek avsevärt förbättrar precisionen och tillförlitligheten av den estimerade effektstorleken och visar på hur MIDD kan vara ett användbart verktyg för detta ändamål inom medicinteknisk utveckling. Studien belyser också begränsningarna i studiedesignen för fallstudien och föreslår hur en modellbaserad metod skulle kunna erbjuda mer robust och tillförlitlig generering av klinisk evidens. Denna forskning belyser potentialen hos MIDD-metoder för att effektivisera den medicintekniska kliniska utvecklingsprocessen, förbättra kvaliteten av evidens, och hantera dess inneboende komplexitet. Framtida arbete bör utvidga dessa resultat genom att utforska mer komplexa modeller, ytterligare studiedesigner, och relaterade aspekter.

Developments in measurement technology, communication and data storage have resulted in measurement systems that produce large amount of data. Together with the long existing need for characterizing the performance of the power system this has resulted in demand for automatic and efficient information-extraction methods. The objective of the research work presented in this thesis was therefore to develop new robust methods that extract additional information from voltage and current measurements in power systems. This work has contributed to two specific areas of interest.The first part of the work has been the development of a measurement method that gives information how voltage flicker propagates (with respect to a monitoring point) and how to trace a flicker source. As part of this work the quantity of flicker power has been defined and integrated in a perceptionally relevant measurement method. The method has been validated by theoretical analysis, by simulations, and by two field tests (at low-voltage and at 130-kV level) with results that matched the theory. The conclusion of this part of the work is that flicker power can be used for efficient tracing of a flicker source and to determine how flicker propagates.The second part of the work has been the development of a voltage disturbance classification system based on the statistical learning theory-based Support Vector Machine method. The classification system shows always high classification accuracy when training data and test data originate from the same source. High classification accuracy is also obtained when training data originate from one power network and test data from another. The classification system shows, however, lower performance when training data is synthetic and test data originate from real power networks. It was concluded that it is possible to develop a classification system based on the Support Vector Machine method with “global settings” that can be used at any location without the need to retrain. The conclusion is that the proposed classification system works well and shows sufficiently high classification accuracy when trained on data that originate from real disturbances. However, more research activities are needed in order to generate synthetic data that have statistical characteristics close enough to real disturbances to replace actual recordings as training data.

In some medical cases it is of interest to study a persons gait. This can be done in several different ways, both simple and advanced. A project has been done at the department Medicinsk Teknik - Forskning och Utveckling at the University Hospital of Umeå. In that project a system with optical triangulation sensors was created to easily measure a persons step width and step height. It was found that the sensors had limitations, in particular with sample frequency, and it was therfore of interest to test new sensors for this application. The goal with this project was to evaluate two sensors, VL53L0X and VL6180, and investigate if they can be used to measure step width and step hight. A system was built to communicate with and read data from the sensors. The sensors were then tested both in a laborative and real environment. With good reflection both VL6180 and VL53L0X show good precision in the range of interest. The sensors has the resolution of millimeters. VL6180 works better at really close distances but has a much shorter maximum range, about 18 cm, and lower maximum sample frequency, under 43 Hz. VL53L0X was tested up to 70 cm and 50 Hz. To be able to measure step width from the ankle, a very high sample frequency is needed to get the right amount of data when the legs pass eachother. If the sensors can be used in that way has not been confirmed. When placed higher on the persons leg, VL6180 can be used to measure step width and VL53L0X can be used to measure both step width and step height. / Inom sjukvården kan det i vissa fall vara intressant att studera en persons gång. Detta kan göras mer eller mindre avancerat och av olika anledningar som för diagnostik eller uppföljning efter ingrepp. Det har tidigare skapats ett mätsystem för mätning av gångparametrar på avdelningen Medicinsk teknik - Forskning och Utveckling vid Norrlands Universitetssjukhus. I det arbetet testades optiska sensorer som använder en triangulerande metodik. Den existerande tekniken hade dock begränsningar i sitt mätutförande vilket ledde till ett behov av att ersätta dessa med en ny typ av sensor. Målet med det här arbetet var därför att utvärdera två andra sensorer, VL6180 och VL53L0X, och undersöka om de kan användas för mätning av stegbredd och steghöjd. Ett mätsystem skapades i för att kommunicera med och samla in data från sensorerna. Därefter utfördes tester, främst laborativa tester med även ett test med den ena sensorn i reell miljö på testperson. Båda sensorerna visar på bra precision och linjäritet på de intressanta avstånden och ger resultat med en upplösning på hela millimetrar. VL6180 opererar bättre på väldigt nära håll men har en begränsning i att det maximala mätavståndet ligger kring 18 cm och den maximala samplingsfrekvensen under 43 Hz. Den frekvensen går att programmera men är instabil. För att kunna placera sensorerna nere vid vristen vid breddmätning krävs en mycket hög samplingsfrekvens för att få in tillräckligt med mätdata. VL53L0X kan komma upp i en en samplingsfrekvens på 50 Hz men det på bekostnad av sämre precision än vid lägre frekvenser. Huruvida VL6180 och/eller VL53L0X kan placeras vid vristen har inte bekräftats. VL53L0X testades med testperson och fungerade då för höjdmätning och breddmätning vid vaden. VL6180 bedöms kunna användas som sensor för mätning av stegbredd och VL53L0X för mätning av stegbredd och steghöjd.

Background and Objective In stereotactic brain tumor biopsies, the combination of real-time fluorescence spectroscopy with the detection of microvascular perfusion using laser Doppler flowmetry provides an improved localization of the brain tumor while decreasing the risk of intra-cranial hemorrhage. The surgeon using the measurement probe is required to view signal values on a screen or usually, when her or his visual focus is directed at the patient, the verbal feedback of a biomedical engineer who is monitoring the measurement signals is needed. In this process possible important information can be overlooked and time is lost. The aim of the thesis was the development a visual and auditory user interface (UI) for use in stereotactic brain tumor biopsies. Materials and Methods The system translates the fluorescence intensity of protoporphyrin IX (PpIX) into sound and visual indicators that are easy and fast to recognize and transmits warning signals in case of signal error or the detection of microvascular perfusion. The increasing and de-creasing fluorescence values at tumor margins were reproduced to improve the precision of de-tecting varying fluorescence intensities when entering tumor tissue with color gradient models. The algorithm produced five signal values when specific fluorescence intensities were measured and compared at different wavelengths.For the development of the UI, a user-centered design was implemented. The user-, operating room- and safety requirements were gathered by communicating with the biomedical engineers and neurosurgeons who had experience in working with fluorescence guided brain tumor biop-sies. The requirements were considered when designing the UI’s features in LabVIEW and the auditory feedback was generated using OSC (Open Sound Control). The user interface intended to deliver measurement data to the user that triggered a high response accuracy by being easy to understand while inducing high user acceptance. The user interaction and function response accuracy of the visual and auditory interface were evaluated in statistical tests where operating room situations were mimicked. The user acceptance of the UI was evaluated. Results Signals for no, low (increasing and decreasing) and high fluorescence indicators, as well as two warning indicators for a blocked signal and vessel occurrence were represented visually and auditorially by the user interface. An intensity/time graph and intensity/wavelength graph, along with the option of recording measurement files and opening saved files allowed the inspec-tion of detailed measurement values. The user study exhibited auditory response accuracy of 95 ± 3% in the intuition test and 91±16% in a memory test. The testing of the response accuracy of the individual signal values displayed accurate responses in 84% to 100% of times a signal was played back. The user acceptance rating of the auditory and visual interface showed no negative results. Conclusion A UI was developed to visually and auditorially represent measurement values to a neurosurgeon performing a stereotactic brain tumor biopsy procedure and biomedical engineers monitoring the measurement signals. The visual display was successful in representing data in a way that was easy to understand. The auditory interface showed high response accuracies for the individual tones representing measurement values. The majority of the test subjects per-ceived the signals to be intuitive, easy to understand and easy to remember. The auditory and visual UI showed high user acceptance ratings, indicating that the user interface was useful and satisfactory in its application.

Medical technology is a field of work with connection to prevention, diagnosis, monitoring, treatment and care. The company Human care is a well-established company in Sweden with red rollators seen all over Sweden. The product family of Human care has four different units of mobility aid solutions. These are lifting devices, rollators, healthcare beds, and convertible chairs. In common they share the same brand, but they do not have any clear connection to each other when it comes to form, colour and texture. In this thesis a visual brand language is developed and created for the company. At the same time, it is implemented and developed on a lifting solution for patient mobility in the healthcare. To find a brand language which should be easier to implement on the products investigations into the brand and products has been done. By trying to find a common ground in the current products the step from creation to implementation of a guideline will be smaller. Further, research has been done in colour psychology, ergonomics, surface textures and forms. The result of this thesis is a guideline which the company can use as a base when developing new products. In this thesis, it has been implemented on a lifting device for patient transportation.

Studies have reported that around 20 % of all traffic accidents are caused by a sleepy driver. Sleepy driving has been compared to drunk driving. A sleepy driver is also an issue in the case of automated vehicles in the future. Handing back the control of the vehicle to a sleepy driver is a serious risk. This has increased the need for a sleepiness estimation system that can be used in the car to warn the driver when driving is not recommended. One commonly used method to estimate sleepiness is to study the heart rate variability, HRV, which is said to reflect the activity of the autonomous nervous system, the ANS. The HRV can be expressed through different measures obtained from a signal of RR-intervals. The aim with the thesis is to investigate how well the HRV translates into sleepiness estimation and how the experimental setup might affect the results. In this study, HRV data from 85 sleep deprived drivers was collected together with the drivers’ own ratings of their sleepiness according to the nine graded Karolinska sleepiness scale, KSS. An ANOVA test showed statistical significance for almost all of the used HRV measures when the Driver ID was set as random variable. In order to reduce the number of HRV measures, a feature selection step was performed before training a Support Vector Machine (SVM) used for classification of the data. SVM classifiers are trained to use the input features describing the data to optimize hyperplanes separating the discrete set of classes. Previous research has shown good results in using HRV for sleepiness detection, but common issues are the small data sets used and that most experiments are performed in a simulator instead of at real roads. In some cases, no sleep deprivation is used. The result from the classification in this study is a mean accuracy of around 58-59 %, mean sensitivity of 50-51 %, mean specificity of 75-76 % and mean F1 score of 50-51 % over the three classes ’Alert’, ’Getting sleepy’ and ’Sleepy’. This together with the results of the ANOVA test shows that the HRV measures performed relatively poor when used for classification of the data and that there are large inter-individual differences. This suggests the use of personalized algorithms when developing a sleepiness estimation system and an investigation regarding how other confounding factors could affect the estimation is also motivated.

Dialys är en medicinteknisk teknik som används för att rena blod vid nedsatt njurfunktion. Dialysvätska är en viktig komponent i utrustningen. Syftet med examensarbetet var att utvärdera förbättrade rutiner för kvalitetskontroll av vattnet som används för tillredning av dialysvätskan. Ett program för övervakning av desinfektionseffekt i vattenrenare för dialys har utvecklats i Microsoft Visual Studio, med Windows Forms som plattform och kod skriven i C#. Mjukvaran övervakar konduktivitet, dagar sedan desinfektion, desinfektionseffekt (A₀), aktivitet hos loggningsprogram samt kontrollprogram. Larm skickas via mail vid dåliga värden eller fel, samt visas lokalt på dator i form av en textruta. Ett ytterligare program skapades för att kontrollera att huvudprogrammet är igång, och en varning skickas när aktivitet uteblir under en viss längd. Utöver övervakning skapades även en grafisk del som tillåter visning av loggade data samt beräknade A₀-värden över tid. Analys utförs kring möjligheter att bibehålla mikrobiologisk trend med stabilt låg bakterienivå i vattnet vid minskad desinfektionsfrekvens från tre till två gånger i veckan. Provtagningar tyder på att marginal till gränsvärde inte hotas, men mer data behövs för at bekräfta resultatet. Analys utförs även kring möjligheten att använda beräkning av A₀-värden för att verifiera desinfektionseffekt tidigare än resultat från mikrobiologisk provtagning, vilket resultat från den studie som utförts tyder på är möjligt, med vissa undantag där det inte finns data att förhålla sig till. Vid användning av mjukvaran rekommenderas fortsatta mikrobiologiska provtagningar. / Dialysis is a health technology used for cleaning blood in people with kidney failure. Dialysis fluid is an important component used in the dialysis process. The aim of the project is to evaluate improved quality control routines for the water used in preparation of the dialysis fluid. A program for monitoring disinfection effect in water purification systems for dialysis has been developed in Microsoft Visual Studio, using Windows Forms as platform and the code was written in C#. The software monitors conductivity, days since last disinfection, effect of disinfection (A₀), activity of logging software as well as control software. Warnings are sent by mail for bad results or errors, and a message is shown as a textbox locally on the computer running the software. An additional program was made for monitoring the main programs activity, and a warning is sent after seeing no activity within a time period. Other than monitoring, the software also has a graphical view allowing the user to view logged data and calculated A₀-values over time. An analysis is made regarding the possibility to maintain microbiological trend of low amounts of bacteria in water after reducing the disinfection frequency from three to two times a week. Test results indicate that the margin to the maximum allowed value remains on safe levels, but more data is required to confirm the results. A further analysis is made regarding the possibility to use calculations of A₀-values to verify the disinfection effect earlier than results from microbiological testing. Results from studies indicate that this is possible, but with some exceptions in cases where not enough data is available to draw conclusions. When using the software, continued microbiological testing is recommended.