High-throughput and automated diagnosis of antimicrobial resistance using a cost-effective cellphone-based micro-plate reader

• Published in: Scientific reports Vol. 6; no. 1; p. 39203
• Main Authors: Feng, Steve, Tseng, Derek, Di Carlo, Dino, Garner, Omai B., Ozcan, Aydogan
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 15-12-2016; Nature Publishing Group
• Subjects: 631/326/107; 639/166/985; 639/624/1107/510; 692/699/255/1318; Accuracy; Anti-Bacterial Agents - pharmacology; Antibiotics; Antimicrobial agents; Automation; Bacteria; Cell Phone; Cellular telephones; Drug resistance; Drug Resistance, Bacterial; Gram-negative bacteria; Gram-Negative Bacteria - drug effects; Gram-Negative Bacteria - isolation & purification; Gram-Negative Bacterial Infections - diagnosis; Gram-Negative Bacterial Infections - microbiology; High-Throughput Screening Assays; Humanities and Social Sciences; Humans; Klebsiella pneumoniae - drug effects; Klebsiella pneumoniae - isolation & purification; Microarray Analysis - economics; Microarray Analysis - instrumentation; Microarray Analysis - methods; Microbial Sensitivity Tests - economics; Microbial Sensitivity Tests - instrumentation; Microbial Sensitivity Tests - methods; multidisciplinary; Nephelometry and Turbidimetry; Preventable deaths; Science; Smartphones; Turbidity
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/srep39203

Routine antimicrobial susceptibility testing (AST) can prevent deaths due to bacteria and reduce the spread of multi-drug-resistance, but cannot be regularly performed in resource-limited-settings due to technological challenges, high-costs, and lack of trained professionals. We demonstrate an automated and cost-effective cellphone-based 96-well microtiter-plate (MTP) reader, capable of performing AST without the need for trained diagnosticians. Our system includes a 3D-printed smartphone attachment that holds and illuminates the MTP using a light-emitting-diode array. An inexpensive optical fiber-array enables the capture of the transmitted light of each well through the smartphone camera. A custom-designed application sends the captured image to a server to automatically determine well-turbidity, with results returned to the smartphone in ~1 minute. We tested this mobile-reader using MTPs prepared with 17 antibiotics targeting Gram-negative bacteria on clinical isolates of
Klebsiella pneumoniae,
containing highly-resistant antimicrobial profiles. Using 78 patient isolate test-plates, we demonstrated that our mobile-reader meets the FDA-defined AST criteria, with a well-turbidity detection accuracy of 98.21%, minimum-inhibitory-concentration accuracy of 95.12%, and a drug-susceptibility interpretation accuracy of 99.23%, with no very major errors. This mobile-reader could eliminate the need for trained diagnosticians to perform AST, reduce the cost-barrier for routine testing, and assist in spatio-temporal tracking of bacterial resistance.