Please use this identifier to cite or link to this item: http://103.99.128.19:8080/xmlui/handle/123456789/619
Full metadata record
DC FieldValueLanguage
dc.contributor.authorMahmud, Fahim-
dc.contributor.authorID:, 18MEE008P-
dc.date.accessioned2026-11-18T04:01:02Z-
dc.date.available2026-11-18T04:01:02Z-
dc.date.issued2025-09-03-
dc.identifier.urihttp://103.99.128.19:8080/xmlui/handle/123456789/619-
dc.descriptionA Master of Science (M.Sc) Thesis in Electrical and Electronic Engineering (EEE) Department at Chittagong University of Engineering and Technology (CUET).en_US
dc.description.abstractSepsis, defined by Sepsis-3 as life-threatening organ dysfunction caused by a dysregulated host response to infection, imposes profound physiological and socio-economic burdens. Globally, it is associated with roughly 50 million cases and 11 million deaths per year, prolonged ICU stays, and substantial healthcare costs and productivity loss. Because organ failure can evolve rapidly, timely detection is critical yet remains challenging in practice due to label noise, missingness, and severe class imbalance in clinical data. This work investigates lightweight machine-learning models for early sepsis detection using seven non-invasive vital signs (heart rate, temperature, systolic/diastolic/mean arterial pressure, oxygen saturation (SpO2), and respiratory rate). We took hourly data, addressed missingness and imbalance with a simple preprocessing pipeline, and used SHAP and LIME to provide model insight. We developed a low-cost point-of-care prototype (Raspberry Pi and commodity sensors with a Python/Django backend) that demonstrates feasibility for real-time prediction as well as telemedicine monitoring. On the PhysioNet Challenge dataset, the system achieved a utility score of ≈ 46%, which is an improvement of approximately 10% from the state-of-the art works. Our model also achieved an accuracy of 86.49%, and an AUROC of 0.94. For external validation on the MIMIC-IV demo cohort, our best model (XGBoost) attained 40% recall for the sepsis class, substantially outperforming a partial qSOFA baseline at 6% recall. These results indicate that simple vital-sign models can materially improve sensitivity over rule-based screening while remaining deployable on affordable hardware; future work should pursue validation on full-scale cohorts and prospective evaluation in real-world ICU workflows.en_US
dc.description.sponsorshipN/Aen_US
dc.language.isoenen_US
dc.publisherCUETen_US
dc.relation.ispartofseries;Tmail-12-
dc.subjectSepsis Predictionen_US
dc.subjectEarly Sepsis Detectionen_US
dc.subjectMachine Learningen_US
dc.subjectClinical Decision Support Systemen_US
dc.subjectVital Signs Monitoringen_US
dc.subjectExplainable Artificial Intelligence (XAI)en_US
dc.titleVital-sign based Real-time Sepsis Diagnosis Using an Interpretable Machine Learning Modelen_US
dc.typeThesisen_US
Appears in Collections:Thesis in EEE

Files in This Item:
File Description SizeFormat 
03.09.2025_Tmail-12_Fahim Mahmud_19MEE008P_.pdfA Master of Science (M.Sc) Thesis in Electrical and Electronic Engineering (EEE) Department at Chittagong University of Engineering and Technology (CUET).8.94 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.