scholarly journals Automated Interpretation of Blood Culture Gram Stains by Use of a Deep Convolutional Neural Network

2017 ◽  
Vol 56 (3) ◽  
Author(s):  
Kenneth P. Smith ◽  
Anthony D. Kang ◽  
James E. Kirby
Keyword(s):  
Neural Network ◽  
Blood Culture ◽  
Convolutional Neural Network ◽  
Clinical Laboratory ◽  
Clinical Microbiology Laboratory ◽  
Proof Of Concept ◽  
Gram Stain ◽  
Gram Positive ◽  
Gram Negative ◽  
Gram Positive Cocci

ABSTRACTMicroscopic interpretation of stained smears is one of the most operator-dependent and time-intensive activities in the clinical microbiology laboratory. Here, we investigated application of an automated image acquisition and convolutional neural network (CNN)-based approach for automated Gram stain classification. Using an automated microscopy platform, uncoverslipped slides were scanned with a 40× dry objective, generating images of sufficient resolution for interpretation. We collected 25,488 images from positive blood culture Gram stains prepared during routine clinical workup. These images were used to generate 100,213 crops containing Gram-positive cocci in clusters, Gram-positive cocci in chains/pairs, Gram-negative rods, or background (no cells). These categories were targeted for proof-of-concept development as they are associated with the majority of bloodstream infections. Our CNN model achieved a classification accuracy of 94.9% on a test set of image crops. Receiver operating characteristic (ROC) curve analysis indicated a robust ability to differentiate between categories with an area under the curve of >0.98 for each. After training and validation, we applied the classification algorithm to new images collected from 189 whole slides without human intervention. Sensitivity and specificity were 98.4% and 75.0% for Gram-positive cocci in chains and pairs, 93.2% and 97.2% for Gram-positive cocci in clusters, and 96.3% and 98.1% for Gram-negative rods. Taken together, our data support a proof of concept for a fully automated classification methodology for blood-culture Gram stains. Importantly, the algorithm was highly adept at identifying image crops with organisms and could be used to present prescreened, classified crops to technologists to accelerate smear review. This concept could potentially be extended to all Gram stain interpretive activities in the clinical laboratory.

scholarly journals Clinical Presentation and Bacterial Etiology of Adult Community Acquired Pneumonia

2017 ◽  
Vol 34 (3) ◽  
pp. 128-134
Author(s):  
Md Abdus Salam ◽  
Md Robed Amin ◽  
Quazi Tarikul Islam
Keyword(s):  
Streptococcus Pneumoniae ◽  
Blood Culture ◽  
Mycoplasma Pneumoniae ◽  
Chlamydia Pneumoniae ◽  
Community Acquired Pneumonia ◽  
Sputum Culture ◽  
Gram Positive ◽  
Gram Negative ◽  
The Mean ◽  
Gram Positive Cocci

Introduction: Pneumonia is a worldwide, serious threat to health and an enormous socio-economic burden for health care system. According to recent WHO data, each year 3-4 million patients die from pneumonia. The clinical presentations and bacterial agents responsible for community acquired pneumonia (CAP) varies according to geography and culture.Methods: A cross sectional observational study conducted among the 53 consecutive patients with a clinical diagnosis of CAP in admitted patient in the department of Medicine, DMCH, during January 2010 to December 2010. Hematological measurements (TC of WBC, Hb%, ESR, platelet count), blood culture, chest X-ray P/A view, sputum for Gram staining and culture sensitivity, sputum for AFB, blood urea and random blood sugar were done in all cases. ELISA for IgM antibody of Mycoplasma pneumoniae and Chlamydia pneumoniae were done in sputum culture negative cases.Results: The mean (±SD) age was 38.9±17.3 years and Male female ratio was 3:1. Fever, chest pain and productive cough were the most common clinical features. The mean (±SD) respiratory rate was 23.0±2.8 /minute . COPD and DM were found in 17.0% and 5.7% of patients respectively . Blood culture was found positive in only 1.9% of the study patients. Gram positive Cocci 62.26%, Gram negative Bacilli 9.43%, mixed Gram positive cocci and Gram negative bacilli 11.32% and Gram negative Cocco Bacilli 1.9% were observed and in 15.03 % cases, no bacteria could be seen. Sputum culture revealed 53.8% streptococcus pneumoniae, 26.9% Klebsiella pneumonia as predominant organism. Mycoplasma pneumoniae and Chlamydia pneumoniae were found in 7.4% and 3.7% respectively by serological test. For Streptococcus pneumoniae, sensitive antibiotics were Amoxyclav and Levofloxacin. For Gram negative bacilli and coccobacilli, more sensitive antibiotics were Meropenem, Ceftriaxone, and Clarithromycin. The best sensitive drug were found meropenem. The mean (±SD) duration of hospital stay was 5.0±1.7 days with ranging from 3 to 10 days.Conclusion: Region based bacteroiological diagnosis of Cap is important for selecting the best and sensitive drugs for complete cure.J Bangladesh Coll Phys Surg 2016; 34(3): 128-134

scholarly journals Comparing the Performance of the New Fully Automated Urine Particle Analyzer UF-5000 with UF-1000i and Gram Stain in Predicting Bacteria Growth Patterns in Women with Uncomplicated Urinary Tract Infections

2020 ◽  
Author(s):  
Stephen Shei-Dei Yang ◽  
Chun-Chun yang ◽  
Yi-Shen Chen ◽  
Shangjen chang
Keyword(s):  
Urinary Tract ◽  
Urine Analysis ◽  
Growth Patterns ◽  
Gram Stain ◽  
Gram Positive ◽  
Gram Negative ◽  
Bacterial Patterns ◽  
Tract Infections ◽  
Sensitivity Specificity ◽  
Gram Positive Cocci

Abstract BackgroudTo compare the performance of the new flow cytometer UF-5000 with UF-1000i (Sysmex, Kobe, Japan) and Gram stain in predicting the bacterial patterns in urine samples MethodsWomen with symptoms suggestive of urinary tract infection were enrolled. Mid-stream urine sample was collected for gram staining, urine analysis and urine culture. Bacterial patterns were classified though UF1000i (none, cocci bacteria or rods/mixed growth), UF-5000 (none, cocci, rods or mixed growth) and Gram stain. Results Among the 102 samples, there were 10 gram-positive cocci, 2 gram-positive bacilli, 66 gram-negative rods, and 24 mixed growth. The sensitivity/specificity of the UF-1000i was 81.8/91.1% for gram-negative rods and 23.5/96.9% for cocci/mixed. The sensitivity/specificity of the UF-5000 was 80.0/88.2% for gram negative rods and 70.0/86.5% for gram-positive cocci.ConclusionsThe UF-5000 demonstrated the good sensitivity and specificity for Gram-negative bacilli bacteria and demonstrated an improved sensitivity for detecting Gram-positive cocci.

scholarly journals Culturing Periprosthetic Tissues in BacT/Alert® Virtuo Blood Culture Bottles for a Short Duration of Post-operative Empirical Antibiotic Therapy

2020 ◽  
Vol 5 (3) ◽  
pp. 145-150
Author(s):  
Claire Duployez ◽  
Frédéric Wallet ◽  
Henri Migaud ◽  
Eric Senneville ◽  
Caroline Loiez
Keyword(s):  
Blood Culture ◽  
Antibiotic Therapy ◽  
Joint Infection ◽  
Length Of Hospital Stay ◽  
Empirical Antibiotic Therapy ◽  
Gram Positive ◽  
Gram Negative ◽  
Surgical Interventions ◽  
Joint Infections ◽  
Gram Positive Cocci

Abstract. Introduction: A post-operative empirical antibiotic therapy (PEAT) is required in periprosthetic joint infections. It commonly uses broad-spectrum antibiotics to cover most Gram-positive cocci and Gram-negative bacilli. It is currently continued until first microbiological results are available, no less than five days later.Methods: We performed a retrospective study in order to evaluate duration of incubation required for surgical samples using the BacT/Alert® Virtuo blood culture bottles system.Results: Among 216 surgical interventions and 199 clinical strains (53.8% staphylococci, 22,1% streptococci and enterococci, 14,6% Gram-negative bacilli, 5,5% anaerobes), 90.5% of the strains were detected between day 0 and day 2; 15 infective strains are cultured from day 3 including 8 Cutibacterium sp., 4 staphylococci, 2 streptococci and 1 Enterococcus.Conclusions: We suggest that the duration of PEAT in patients operated for a periprosthetic joint infection may be shortened to three days as Gram-negative rods are unlikely to grow after three days of culture by using BacT/Alert® Virtuo blood culture bottles. This is likely to shorten the overall length of hospital stay, to diminish the occurrence of adverse side effects, and the emergence of antimicrobial resistance. However, coverage of Gram-positive cocci should be maintained for 14 days until the definite culture results are available.

scholarly journals Study of the Comparative Activity of Piperacillin/Tazobactam with Currently Available Antibiotics against 8206 Aerobic Isolates

1997 ◽  
Vol 8 (3) ◽  
pp. 147-153 ◽  
Author(s):  
Kevin R Forward ◽  
Donald E Low ◽  
Michel Laverdiere ◽  
Robert Rennie ◽  
Andrew E Simor ◽  
...  
Keyword(s):  
Clinical Laboratory ◽  
National Committee ◽  
Community Hospitals ◽  
Gram Positive ◽  
Gram Negative ◽  
Beta Lactamases ◽  
Extended Spectrum ◽  
Disk Diffusion Testing ◽  
Extended Spectrum Beta Lactamases ◽  
Gram Positive Cocci

OBJECTIVES: To compare the activity of piperacillin-tazobactam with piperacillin and other parenterally administered antibiotics against aerobic Gram-negative bacilli and Gram-positive cocci isolated from across Canada, and to determine the prevalence of resistance mediated by extended-spectrum cephalosporinases.METHODS: Sixty-one laboratories participated. Disk diffusion testing was performed in accordance with methods outlined by the National Committee for Clinical Laboratory Standards. Susceptibilities were performed on 8206 strains.Escherichia coliandKlebsiella pneumoniaewith reduced susceptibilities to third-generation cephalosporins were screened for extended-spectrum beta-lactamases (ESBLs).RESULTS: Piperacillin-tazobactam was active against 92% of the strains, piperacillin against 81% and ticarcillin-clavulanic acid against 88%. Few differences were observed in the relative susceptibility of strains from teaching or community hospitals, from different anatomic sites or from different regions of the country. Aerobic Gram-negative bacilli tested tended to be more susceptible to all the agents than was recently reported in a similar American study. Only 43% ofEnterococcus faeciumwere susceptible to ampicillin and 42% to piperacillin piperacillin with and without tazobactam. Only two enterococcal strains were resistant to vancomycin, and 19 had intermediate zone sizes. Of the 10 strains ofE coliand eight strains ofK pneumoniaewith reduced susceptibility to extended spectrum cephalosporins, only one demonstrated typical ESBL activity.CONCLUSIONS: Canadian aerobic Gram-positive cocci and Gram-negative bacilli remain highly susceptible to many currently available antibiotics. The findings confirm a broad spectrum of activity of piperacillin-tazobactam and indicate that the pattern of susceptibility is quite uniform from different body sites, in both teaching and community hospitals, and across the country.

scholarly journals Use of thermonuclease testing to identify Staphylococcus aureus by direct examination of blood cultures

2003 ◽  
Vol 9 (1-2) ◽  
pp. 185-190
Author(s):  
N. M. Kaplan
Keyword(s):  
Staphylococcus Aureus ◽  
Blood Culture ◽  
Direct Detection ◽  
Blood Cultures ◽  
Military Hospital ◽  
Clinical Microbiology Laboratory ◽  
Microbiology Laboratory ◽  
Direct Examination ◽  
Gram Positive ◽  
Gram Positive Cocci

Blood cultures submitted to the Clinical Microbiology Laboratory, Queen Alia Military Hospital, Amman during 1999-2001 were examined to evaluate thermonuclease testing for identifying Staphylococcus aureus in blood culture broths growing gram-positive cocci. Of 170 cultures studied, 129 yielded gram-positive staphylococci and 41 yielded other gram-positive cocci. Toluidine blue-deoxynucleic acid agar plates were used to test for thermonuclease activity. St and ard tube coagulase tests were performed on the isolates. Direct detection of thermonuclease activity in 76 blood culture broths containing gram-positive staphylococci showed 100% correlation with subsequent tube coagulase tests. The thermonuclease test provides a fast, specific and reliable confirmation of S. aureus bacteraemia by direct examination of blood culture broths that contain gram-positive cocci. This allows for timely, optimal antibiotic therapy

What are the critical steps in processing blood cultures? A prospective audit evaluating current practice of reporting blood cultures in a centralised laboratory serving secondary care hospitals

2016 ◽  
Vol 70 (4) ◽  
pp. 361-366 ◽  
Author(s):  
Manjula Meda ◽  
James Clayton ◽  
Reela Varghese ◽  
Jayakeerthi Rangaiah ◽  
Clive Grundy ◽  
...  
Keyword(s):  
Blood Culture ◽  
Secondary Care ◽  
Rapid Identification ◽  
Antimicrobial Treatment ◽  
Blood Cultures ◽  
National Standards ◽  
Gram Stain ◽  
Gram Positive ◽  
Gram Negative ◽  
Common Intervention

AimsTo assess current procedures of processing positive blood cultures against national standards with an aim to evaluate its clinical impact and to determine the utility of currently available rapid identification and susceptibility tests in processing of blood cultures.MethodsBlood cultures from three secondary care hospitals, processed at a centralised laboratory, were prospectively audited. Data regarding processing times, communication with prescribers, changes to patient management and mortality within 30 days of a significant blood culture were collected in a preplanned pro forma for a 4-week period.ResultsOf 2206 blood cultures, 211 positive blood cultures flagged positive. Sixty-nine (3.1%) of all cultures were considered to be contaminated. Fifty per cent of blood cultures that flagged positive had a Gram stain reported within 2 hours. Two (0.99%) patients with a significant bacteraemia had escalation of antimicrobial treatment at the point of reporting the Gram stain that was subsequently deemed necessary once sensitivity results were known. Most common intervention was de-escalation of therapy for Gram-positive organisms at the point of availability of pathogen identification (25.6% in Gram positive vs 10% in Gram negative; p=0.012). For Gram-negative organisms, the most common intervention was de-escalation of therapy at the point of availability of sensitivity results (43% in Gram negatives vs 17.9% in Gram positive; p=0.0097). Overall mortality within 30 days of a positive blood culture was 10.9% (23/211). Antibiotic resistance may have contributed to mortality in four of these patients (three Gram negative and one Gram positive).ConclusionGram stain result had the least impact on antibiotic treatment interventions (escalation or de-escalation). Tests that improve identification time for Gram-positive pathogens and sensitivity time for Gram-negative pathogens had the greatest impact in making significant changes to antimicrobial treatment.

scholarly journals 285. Evaluation of the ePlex® Blood Culture Identification (BCID) Panels for Gram-positive/Gram-negative bacteria and yeasts

2020 ◽  
Vol 7 (Supplement_1) ◽  
pp. S143-S143
Author(s):  
Kimberle Chapin ◽  
Giannoula Tansarli
Keyword(s):  
Blood Culture ◽  
Gram Positive ◽  
Gram Negative ◽  
Soc Testing ◽  
Percent Agreement ◽  
Culture Identification ◽  
Reduced Time ◽  
Gram Positive Cocci ◽  
Pna Fish ◽  
Vitek Ms

Abstract Background Multiple methods used for blood culture identification create inconsistent to reporting of critical results. Study aim was to evaluate performance characteristics of the ePlex BCID panels compared to current standard of care (SOC) methods used in our lab. Methods Identification sensitivity and specificity were assessed across all targets detected by the ePlex as well as time to final identification (from time of bottle positive Gram stain) between ePlex and SOC testing. SOC included Xpert MRSA/SA or latex agglutination for Gram-positive cocci in clusters (GPCC), Vitek MS + Accelerate Pheno for Gram-negative rods (GNRs), serotyping or optochin disk ± Vitek MS for Gram-positive cocci in chains (GPC chains), Vitek MS or Vitek-2 for Gram-positive rods (GPR), and PNA-FISH or Vitek MS for yeasts. Results 313 unique prospective blood culture specimens were tested with ePlex BCID panels during a 3-month period (January-March 2020). The positive percent agreement was 100% for GNR (n= 98), S. aureus (n= 42), coagulase-negative staphylococci (n= 38), Group A Streptococcus (n= 3), Group B Streptococcus (n= 5), S. pneumoniae (n= 10), GPR (n= 21), and yeasts (n= 20). There was 1 false negative, (S.mutans) which should have been detected. The negative percent agreement was 100% across all targets except for 1 false positive Corynebacterium spp. In total, 6.7% of blood cultures had an off-panel organism which ePlex did not detect. The median time to final identification was 3 (2 – 4) hrs. for ePlex and calculated for all other SOC methods. Compared to SOC molecular methods, the ePlex reduced time to identification 0.5 h compared to Xpert MRSA/SA, 6.7 h compared to Accelerate Pheno for GNR (but Accelerate Pheno provides susceptibilities), and 3 h compared to PNA-FISH for yeasts (p< 0.05). ePlex compared to non-molecular techniques (MALDI-TOF), SOC for Streptococcus spp. and Enteroococcus spp., the time to final identification was reduced by 24 – 30 hours (p< 0.05). Workflow chart comparison eplex to SOC Time to results eplex vs SOC Conclusion The ePlex BCID system provided highly accurate identification results for GP and GN bacteria as well as for yeasts. Our evaluation showed that this system significantly reduced time to final identification compared to SOC testing methods. Disclosures Kimberle Chapin, MD, genmark (Scientific Research Study Investigator) Giannoula Tansarli, MD, GenMark (Grant/Research Support)

The Diagnostic Value of Gram Stain for Initial Identification of the Etiologic Agent of Peritonitis in Capd Patients

1997 ◽  
Vol 17 (3) ◽  
pp. 269-272 ◽  
Author(s):  
Dinoráh A. Bezerra ◽  
Mario B. Silva ◽  
Jaqueline S.T. Caramori ◽  
Maria F. Sugizaki ◽  
Teruê Sadatsune ◽  
...  
Keyword(s):  
Predictive Value ◽  
Medical Center ◽  
Single Agent ◽  
Etiologic Agent ◽  
Initial Diagnosis ◽  
Gram Stain ◽  
Gram Positive ◽  
Dialysis Unit ◽  
Gram Negative ◽  
Gram Positive Cocci

Objective To evaluate the effectiveness of the Gram stain in the initial diagnosis of the etiologic agent of peritonitis in continuous ambulatory peritoneal dialysis (CAPD). Design Retrospective study analyzing the sensitivity (S), specificity (SS), positive predictive value (+PV), and negative predictive value (-PV) of the Gram stain relating to the results of cultures in 149 episodes of peritonitis in CAPD. The data were analyzed in two studies. In the first, only the cases with detection of a single agent by Gram stain were taken (Study 1). In the second, only the cases with two agents in Gram stain were evaluated (Study 2). Setting Dialysis Unit and Laboratory of Microbiology of a tertiary medical center. Patients Sixty-three patients on regular CAPD who presented one or more episodes of peritonitis from May 1992 to May 1995. Results The positivity of Gram stain was 93.2% and the sensitivity was 95.7%. The values of S, SS, +PV, and -PV were respectively: 94.9%, 53.5%, 68.3%, and 90.9% for gram -positive cocci and 83.3%, 98.8%, 95.2%, and 95.6% for gram-negative bacilli. The association of grampositive cocci plus gram -negative bacilli were predictive of growth of both in 6.8%, growth of gram -positive cocci in 13.7%, and growth of gram -negative bacilli in 72.5%. Conclusions The Gram stain is a method of great value in the initial diagnosis of the etiologic agent of peritonitis in CAPD, especially for gram-negative bacilli.

scholarly journals MICROBIAL PATTERN AND ANTIBIOTIC SENSITIVITY TEST OF HOSPITALIZED CHILDREN

2016 ◽  
Vol 51 (3) ◽  
pp. 190
Author(s):  
John Wiwin ◽  
IGAA Putri Sri Rejeki
Keyword(s):  
Blood Culture ◽  
Rectal Swab ◽  
Antibiotic Sensitivity ◽  
Sensitivity Test ◽  
Sensitivity Testing ◽  
Gram Positive ◽  
Gram Negative ◽  
E Coli ◽  
Oncology Ward ◽  
Gram Positive Cocci

Infection often occurs in children with malignant hematology and causes morbidity and mortality in the children. Antibiotics given must be based on culture results and antibiotic sensitivity testing. This study was aimed to obtain the microbial pattern and sensitivity test in children hospitalized in the Hemato-Oncology Ward,  dr. Soetomo Hospital from September 2012 - February 2013. This was a descriptive study. Data were obtained from the  patients’ medical records  in Dr. Soetomo Hospital. There were 341 culture examinations (blood, urine, rectum swab, faecal, and others) from 88 patients (44 males and 44 females). Most of patients´ age was < 5 years (58%) and suffered from ALL (50%).There were microbial (83 of culture) and yeast (15 of culture) growth out of 98 cultures. Escherichia coli, Burkholderia cefacea, and Klebsiella oxytoca (Gram negative) dan CONS, Stapyloccocus aureus, and Stapylococcus sapropyticus (gram positive) were found in blood culture. S. aureus (gram positive) and E. coli, Klebsiella pneumoniae, and B. cefacea (gram negative) were found in urine culture. Only E. coli was found in rectal swab culture. CONS of gram positive cocci were mostly found in blood culture of children hospitalized in Hemato-Oncology Ward, Dr. Soetomo Hospital. E. coli was the mostly found gram negative rods. Gram positive cocci showed a high resistant to penicillin and co-trimoxazole. E. coli, mostly found in rectal swab and urine, has a high sensitivity to amikacin and meropenem, but highly resistant to  ampicillin and ampicillin sulbactam.

THE EFFECT OF SALMINE ON BACTERIA

1958 ◽  
Vol 4 (2) ◽  
pp. 65-71 ◽  
Author(s):  
Thomas D. Brock
Keyword(s):  
Sodium Chloride ◽  
Gram Negative Bacteria ◽  
Bacteriostatic Activity ◽  
Bactericidal Action ◽  
Gram Stain ◽  
Gram Positive ◽  
Gram Negative ◽  
Bacterial Surface ◽  
Bactericidal Effects

The bacteriostatic and bactericidal effects of salmine on various bacteria have been studied. Salmine has more bacteriostatic activity against Gram-positive than against Gram-negative bacteria. It is bactericidal in water but not in broth, and this bactericidal action occurs against both Gram-positive and Gram-negative bacteria. It has been shown that salmine causes agglutination of washed suspensions of certain bacteria and this agglutination is not correlated directly with the Gram stain. Salmine causes an increase in the turbidity of washed cells of all bacteria, Gram-positive and Gram-negative, and differs in this respect from the solutes sodium chloride and glucose, which affect only Gram-negative species.A comparison has been made of the effects of salmine and polymyxin and it has been concluded that salmine may also act by attachment to the bacterial surface.

Sign in / Sign up

Export Citation Format

Share Document