This study is a part of a project at Q-linea that aims to present a rapid diagnostic instrument to speed up the process of identification of pathogens and determination of MIC-values (Minimal Inhibitory Concentration) of antibiotic needed to treat patients with sepsis. Specifically, this report is aimed to describe the development and implementation of algorithms that examine susceptibility profiles ofsepsis related pathogens where the bacteria have been exposed to different antibiotics and by different lapse of concentrations. The developed algorithms are based on a clustering technique that identify inhibited growth and present the lowest concentration needed to slow down the growth of the pathogen. The implemented solution was tested on sepsis related pathogens and the determined MIC values were compared to MIC values generated with a method commonly used in healthcare today. Approximately 90% of instances were correctly classified based on data from six hours long tests which is significantly faster than the reference method which takes 16-24 hours to complete. Furthermore, each result comes with a set of quality measures for validation of the algorithm results. Although, further studies are necessary to increase the performance at the four-hour target time, and more data is needed to validate the developed quality measures.
Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:uu-325414 |
Date | January 2017 |
Creators | Åhag, Stina |
Source Sets | DiVA Archive at Upsalla University |
Language | English |
Detected Language | English |
Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
Format | application/pdf |
Rights | info:eu-repo/semantics/openAccess |
Relation | UPTEC X ; 17 014 |
Page generated in 0.0015 seconds