Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Efficiency of TrueNat: Use as the point of care test
0
Zitationen
10
Autoren
2025
Jahr
Abstract
Background: Tuberculosis (TB) remains a significant health challenge, particularly in a populous country like India. Efficient diagnosis and treatment are crucial in the age of advanced diagnostic methods such as CBNAAT and line probe assay (LPA). They are proven effective; however, they are often resource-intensive. TrueNat, a chip-based nucleic acid amplification technique, offers a viable alternative due to its affordability, ease of use, and portability. Aims and Objective: This study evaluates the accuracy of TrueNat for Mycobacterium tuberculosis (Mtb) and resistance detection in comparison to other diagnostic methods, highlighting its potential to enhance accessibility and reliability in diverse settings. Materials and Methods: Samples from suspected TB patients in a tertiary care hospital in western Maharashtra underwent Ziehl–Neelsen staining, TrueNat testing for Mtb and rifampicin (RIF) resistance, and LPA for drug resistance. Data analysis was performed to evaluate diagnostic effectiveness. Results: Among 1291 samples, 18.9% tested positive by TrueNat, whereas 30.07% yielded indeterminate results, predominantly in low bacterial load samples (65% in very low load). Detection of acid-fast bacilli (AFB) was observed mainly in high and medium bacterial loads. TrueNat identified Mtb in 47% of samples that did not show AFBs. LPA locus detection was correlated with high and medium loads. Conclusion: TrueNat demonstrates higher accuracy in detecting Mtb and RIF resistance in samples with high bacterial loads. It surpasses smear microscopy and LPA in sensitivity, though indeterminate RIF resistance results should be confirmed by other methods.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.245 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.102 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.468 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.776 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.429 Zit.