Ovarian cancer is the most lethal gynaecological malignancy; however, ovarian lesions are very common and only around 1% are malignant. Due to the large number of cases, patients are triaged by gynaecologists having a high variability in diagnostic accuracy. The aim of this study is to train and validate deep neural networks and, by comparison to subjective expert assessment, determine their potential in the triage of patients with ovarian tumours. We used a transfer learning approach on pre-trained networks (VGG16, ResNet50, MobileNet), and a post-processing calibration to better align their confidence scores with the true certainty of their predictions. Our dataset contained 3077 transvaginal ultrasound images from 758 patients with ovarian tumours, where histological outcome from surgery or long-time follow-up (> 3 years) served as diagnostic ground truth. From our dataset, 150 cases (75 benign, 75 malignant), each containing 3 images, were held out for testing, while the remaining cases were used for training and model selection. The models were assessed bases on sensitivity, specificity, and AUC, along with their corresponding 95% confidence intervals. On the test set, our final model had a sensitivity of 96.0% (0.897–0.989), specificity of 86.7% (0.776–0.929), and AUC of 0.950 (0.906–0.985). When excluding the 12.7% (0.073–0.180) of cases most difficult to classify (based on the confidence score of the model output), our model had a sensitivity of 97.1% (0.909–0.994), specificity of 93.7% (0.856–0.978), and AUC of 0.958 (0.911–0.993). As comparison, the subjective expert assessment had a sensitivity and specificity of 96.0% and 88.0% respectively. We show that neural networks can be used to predict ovarian malignancy with high diagnostic accuracy, comparable to that of human experts, and thus have potential in the triage of patients with ovarian tumours. / Äggstockscancer har högst dödlighet bland gynekologiska cancersjukdomar. Äggstocksförändringar är dock vanligt förekommande och endast omkring 1% är maligna. På grund av den höga förekomsten görs initialt en bedömning lokalt (triage) huruvida patienten skall remitteras vidare för expertbedömning, eller om uppföljning på lokal vårdinrättning är tillräcklig. Triagen utförs av gynekologer som saknar utan expertkompetens inom äggstockscancer, och därav har stor variation i diagnostisk precision. Syftet med denna studie är att, genom jämförelse med subjektiv expertbedömning, utvärdera potentialen hos artificiella neurala nätverk för triagering av kvinnor med äggstockstumörer. Vi använde transfer learning av förtränade modeller (VGG16, ResNet50, MobileNet) och en kalibreringsmetod för bättre probabilistisk överensstämmelse mellan modellernas svar och deras underliggande diagnostiska precision. Vårt bildmaterial bestod av 3077 transvaginala ultraljudsbilder från 758 kvinnor med äggstockstumörer. Samtliga fall hade säkerställd diagnos genom resultat från operation eller långvarig uppföljning (> 3 år). Av detta material lades 150 fall (75 benigna, 75 maligna) à 3 bilder åt sidan för slutgiltig validering av modellen, medan resterande fall användes till träning och val av modell. Modellerna bedömdes baserat på sensitivitet, specificitet och AUC, ihop med deras 95\% konfidensintervall. Vid validering hade vår slutgiltiga modell en sensitivitet på 96,0% (0,897–0,989), specificitet på 86,7% (0,776–0,929), och AUC på 0,950 (0,906–0,985). Vid uteslutande av 12,7% (0,073–0,180) av de fall som var svårast att klassificera hade vår modell en sensitivitet på 97,1% (0,909–0,994), specificitet på 93,7% (0,856–0,978) och AUC på 0,958 (0,911–0,993). Som jämförelse hade den subjektiva expertbedömningen en sensitivitet och specificitet på 96,0%, respektive 88,0%. Vår studie visar att artificiella neurala nätverk kan användas för differentiering av benigna och maligna äggstockstumörer med hög diagnostisk precision, jämförbar med den hos erfarna läkare på området. Således bedömer vi att det finns potential för användning av dessa modeller för triagering av kvinnor med äggstockstumörer.

Background:Approximately 15% of the world’s population are affected bysome kind of disability where over 150 conditions may affect the human gaitpattern. The ability to ambulate with ease is important for overall well-being.Various assistive devices have been developed to improve mobility of theirusers. A lot of research is currently focused on ankle exoskeletons, showingpromising results in providing important assistance during stance phase of gait. Objective:To investigate how different combinations of active and passiveelements in an ankle exoskeleton affects the metabolic cost of walking. Methods:Musculoskeletal simulations were carried out in OpenSim Moco.Different assistive configurations were tested over one gait cycle using a pas-sive element, an active element, and a parallel connection of the both. Parame-ter values were modified to find the most optimal setup for reducing metaboliccost. Results:All assistive configurations were found successful in reducing bothwhole-body metabolic cost and the metabolic cost of the plantarflexors whencompared to the unassisted gait. Most whole-body metabolic cost reductionwas found when using a passive spring with resting length of 0.28 m and stiff-ness of 6 kN/m in parallel with an active motor capable of providing forceequal to 150% of body weight. The most reduction in metabolic cost of theplantarflexors was also found for a parallel connection of elements, but herewith a 100% body weight motor and spring with rest length of 0.19 m andstiffness of 10 kN/m. With higher assistance, more reduction in ankle mo-ment generated by the muscles was observed. Conclusion:Powered ankle exoskeletons are promising in terms of minimiz-ing metabolic cost during walking due to assistance during late stance phaseof gait for ambulators requiring plantarflexor assistance. Keywords:Simulation, exoskeleton, ankle, moco.

Oral mucositis is an inflammation that affects the mucosa in the mouth. It often affects patients undergoing chemotherapy for cancer. The risk of inflammation can be reduced by cooling the mouth. In this project, ten different mouthpieces for cryotherapy of the oral cavity with air as cooling medium have been modeled in the SolidWorks. Flow simulations have been made on the different mouthpieces to investigate how the air is distributed in the models. The goal of the simulations is that the outflow of air should not differ more than 10 % between the largest and the smallest flow. With the results, the mouthpieces could be fine-tuned. Flow simulation on the ten prototypes gave varying results, some have a good distribution of air between the holes. Some, however, do not evenly spread the air and it flows much less air through certain holes. Three prototypes were printed in hard plastic at an early stage to see how they fit the mouth. At the end of the project, a total of seven models were printed by TADA medical's partner in a softer plastic. Several models then showed some points that need to be considered when designing nozzles. A handful of the models have been tested with the cooling device and the flow has been examined with a heating camera. The other prototypes have not been 3D-printed, so there is nothing about their convenience. To get better results, more advanced simulation requires the environment of the oral cavity with the exact dimensions.

Infectious diseases pose a serious threat to global health. Molecular diagnostics provide solutions for effective control and prevention of infections, however suffers from expensive laboratory equipment, and infrastructure to be fully implemented at point of care (POC), especially at low-resource settings. Lab-on-a-chip that aims to integrate complex biochemical analyses into automated systems is promising for POC analysis. A major challenge is the integration of a complete molecular diagnostic assay, generally translating into complex microfluidics, with the requirement of low fabrication cost. This thesis explores the use of flexible electronics, plastic foils and roll-to-roll manufacturing to enable low-cost microfluidic systems, for molecular diagnostic assays especially targeted towards infectious diseases. Many biochemical assays rely on heat; hence a first aspect in this thesis is the integration of a microheater into microfluidics. In a first project a system for SNP-genotyping is presented using solid phase melting curve analysis to discriminate mutations at a single base resolution. Starting with a glass based concept (paper I) which is further developed to a foil based system (paper II), detection of the polymorphism in the neuropeptide Y associated with increased risk of type II diabetes is demonstrated as a proof of principle. Further development and optimization of the microheater concept has enabled roll-to-roll manufacturing compatibility and multiplexing of targets (paper III). A bacterial sub-typing and multiresistance detection in clinical Staphylococcus Aureus samples is demonstrated for applications in infectious diseases diagnostics. Finally, the microheater concept is further developed to enable μPCR (paper IV). Detection of genomic HIV-1 is demonstrated and a portable detection setup based on an LED light source and low cost CMOS camera for detection was developed. A second aspect of this thesis is integration of light sources and optical detection (paper V-VI). A multilayer system integrating an electroluminescent light source, reactive sensor dyes and organic semiconductor transistor for detection is demonstrated. The system could be used for amine detection in gases (paper V). System was made further roll-to-roll compatible. The system uses an external LED light source and a photodetector processed in only one screen printing- and one dispensing step (paper VI). As a proof of principle, absorbance based DNA hybridization was detected. Collectively, roll-to-roll manufacturing compatible “lab on foil” systems have the potential to improve our ability to diagnose at POC especially at resource-limited settings. / <p>QC 20170426</p>

Rapporten behandlar utvecklingen och konstruktionen av ett laparoskopiskt blodkärlsidentifierande instrument. Identifiering sker med hjälp av fotopletysmografi. Fotopletysmografi betyder ungefär "detektion av volymförändrings med hjälp av ljus". Laparoskopi kan översättas med titthålskirurgi. Laparoskopioperationer sker med små runda instrument som förs in i kroppen. Rapporten diskuterar olika probmodeller, det vill säga hur sensorn skall konstrueras för att erhålla en bra mätsignal. Konkurrerande tekniker och framtida applikationer redovisas på flera ställen i rapporten. Vid mätningar på blodmodell studerades och analyserades signalkvaliten. Till slut skall tilläggas att proben fungerade tillfredsställande. Proben kunde detektera pulsationer i blodmodellen.

With a longer-living population and an increase in cancer incidence the health care’s workload has increased over the past decade. The treatment process of a cancer patient is dependant on clinical information collected and communicated from the pathology department. With a standardised and structured pathology report the information communicated can become easier to interpret and will fa- cilitate the search for important parameters. This master thesis aims to develop a template prototype to replace four static free-text templates used in the area of breast cancer pathology at the pathology department at Region Östergötland. The end product was intends to store docu- mented information in a structured manner through structured data, in order to obtain semantic interoperability. Semantic interoperability means that different systems are able to communicate with each other in such a way that the information is handled and interpreted equally by the systems. By using certain standards such as openEHR archetypes and SNOMED CT concepts, the data becomes uniform and unambiguous. When that is achieved, information can be sent more easily between systems such as patient health data if an individual moves between different cities where the hos- pitals have different medical records systems. The result of the master thesis is a single template that incorporates all the parts from the four static templates currently used at Region Östergötland. To avoid a large and cumbersome template for the end-user the template is built with con- ditions that changed the appearance of the template while it is being filled in, making it dynamic.

Metamers of the ventral stream is a model which tries to describe what information we gather from our visual field. It have previously only been tested on static images. This thesis have continued the research and applied it to dynamic images in order to investigate if the model can be seen as a functional representation of our visual field. The results show that the model, at this stage, can not be seen as a fully functional representation of the visual field, but it can be used to determine the detectability of objects in the periphery. It also shows that what we humans perceive as motion is, at least to some extent, merely a change of the statistics in our visual field.

No description available.

<p>Svenska medicintekniska företag kan inte konkurrera med de globala företagen med stordrift och låga priser, de måste därför anpassa sina produkter efter de specifika behoven i den svenska vården.</p><p>Syftet med studien är att analysera och utvärdera svenska medicintekniska företags marknadsföringsstrategier med avseende på sambandet mellan samarbetsformer och produkters anpassning till specifika behov.</p>

<p>Många små företag upplever svårigheter att finansiera företaget under uppstarten, det krävs både kontakter, tur och tålamod för att lyckas. Att finna finansiering är ofta ett omfattande och tidsödande arbete för småföretagaren. Medicintekniska företag tillverkar ofta forskningsbaserade produkter och har ett stort kapitalbehov långt innan de kan visa upp en färdig produkt eller försäljningsintäkter. Detta är ett stort problem då externa finansiärer helst inte vill riskera för mycket vid en investering och föredrar att investera i ett senare skede.</p><p><strong>Syfte: </strong>Syftet med uppsatsen är att ta reda på vilken attityd småföretag i medicintekniska branschen har på externt kapital samt i vilken utsträckning det används för finansiering av företagen i introduktionsfasen.</p><p><strong>Teori: </strong>Finansieringsteori och tidigare forskning kring finansiering av småföretag samt riskkapital har studerats för att skapa en teoretisk referensram.</p><p><strong>Metod: </strong>En kvalitativ metod användes och undersökningen bygger på intervjuer med åtta små medicintekniska företag, en affärsinkubator och en investerare. Vid analysen användes en deduktiv ansats och teorier och modeller testades mot det empiriska underlaget.</p><p><strong>Resultat och analys: </strong>Finansieringen under uppstarten består till största delen av offentlig finansiering och riskkapital. Företagen har en lång introduktionsfas då det tar lång tid att ta en produkt till marknad. Det förekommer gap mellan företag och investerare av olika karaktär, förväntningsgap, informationsgap och skillnad i målsättning är några. Det är relativt lätt att få de små belopp som krävs för att starta ett företag men småföretagen har sedan svårt att komma över tröskeln hos riskkapitalbolagen då de söker ett för litet kapitalbelopp.</p><p><strong>Slutsatser: </strong>Offentlig finansiering är betydande för företagen. Det stora kapitalbehovet gör det nödvändigt att finansiera företaget med riskkapital. Riskkapital bidrar förutom rent finansiellt med kompetens och kontakter. Trots att investerare helst vill investera i senare faser kan dessa företag attrahera externt kapital.</p>