Deep learning algorithm for detection of aortic dissection on non-contrast-enhanced CT.

OBJECTIVES: To develop a deep learning-based algorithm to detect aortic dissection (AD) and evaluate the diagnostic ability of the algorithm compared with those of radiologists. METHODS: Included in the study were 170 patients (85 with AD and 85 without AD). An AD detection algorithm was developed using a convolutional neural network with Xception architecture. Of the patient data, 80% were used for training and validation and 20% were used for testing. Fivefold cross-validation was performed to evaluate the method. An average of 6688 non-contrast-enhanced CT images (slice thickness, 5 mm) were used for training. A radiologist reviewed both contrast-enhanced and non-contrast-enhanced images and identified the slices of AD. The identified slices were used as ground truth. Receiver operating characteristic curve and area under the curve (AUC) analysis was performed. Five radiologists independently evaluated the images. The accuracy, sensitivity, and specificity of the algorithm and those of the radiologists were compared. RESULTS: The AUC of the developed algorithm was 0.940, and a cutoff value of 0.400 provided accuracy of 90.0%, sensitivity of 91.8%, and specificity of 88.2%. For the radiologists, median (range) accuracy, sensitivity, and specificity were 88.8 (83.5-94.1)%, 90.6 (83.5-94.1)%, and 94.1 (72.9-97.6)%, respectively. There was no significant difference in performance in terms of accuracy, sensitivity, or specificity between the algorithm and the average performance of the radiologists (p > 0.05). CONCLUSIONS: The developed algorithm showed comparable diagnostic performance to radiologists for detecting AD, which suggests the potential of the proposed method to support clinical practice by reducing missed ADs. KEY POINTS: • A deep learning-based algorithm for detecting aortic dissection was developed using the non-contrast-enhanced CT images of 170 patients. • The algorithm had an AUC of 0.940 for detecting aortic dissection. • The accuracy, sensitivity, and specificity of the algorithm were comparable to those of radiologists.

Contribution of ß-lactamase and efflux pump overproduction to tazobactam-piperacillin resistance in clinical isolates of Escherichia coli.

INTRODUCTION: Tazobactam-piperacillin (TZP) is a mixture of a broad-spectrum penicillin and an irreversible ß-lactamase inhibitor. TZP is effective against Gram-negative bacteria that produce extended-spectrum ß-lactamases, and it is used as a first-line or second-line drug to treat serious infections. METHODS: This study identified three TZP-resistant and two TZP-intermediate strains among 514 clinical isolates of Escherichia coli. RESULTS: These five isolates possessed one or more ß-lactamase genes, blaTEM-1, blaCTX-M-2, blaCTX-M-14, and/or blaCMY-8. The expression levels of ß-lactamase genes and acrAB genes in the strains were examined by using real-time reverse transcription PCR. The total enzymatic piperacillin-degrading activity in cells was determined. Two TZP-resistance mechanisms were identified: hyperproduction of TEM-1 in the two resistant strains; and simultaneous high production of ß-lactamase and efflux pump AcrAB in the two TZP-intermediate isolates. The latter are an international high-risk clone O25b:H4-ST131-H30R. CONCLUSION: TZP resistance is still rare in clinical isolates of E. coli. However, resistance can develop on high production and/or combinations of known antimicrobial resistance mechanisms in different ways.

Mechanism of Reduced Susceptibility to Fosfomycin in Escherichia coli Clinical Isolates.

In recent years, multidrug resistance of Escherichia coli has become a serious problem. However, resistance to fosfomycin (FOM) has been low. We screened E. coli clinical isolates with reduced susceptibility to FOM and characterized molecular mechanisms of resistance and reduced susceptibility of these strains. Ten strains showing reduced FOM susceptibility (MIC ≥ 8 µg/mL) in 211 clinical isolates were found and examined. Acquisition of genes encoding FOM-modifying enzyme genes (fos genes) and mutations in murA that underlie high resistance to FOM were not observed. We examined ability of FOM incorporation via glucose-6-phosphate (G6P) transporter and sn-glycerol-3-phosphate transporter. In ten strains, nine showed lack of growth on M9 minimum salt agar supplemented with G6P. Eight of the ten strains showed fluctuated induction by G6P of uhpT that encodes G6P transporter expression. Nucleotide sequences of the uhpT, uhpA, glpT, ptsI, and cyaA shared several deletions and amino acid mutations in the nine strains with lack of growth on G6P-supplemented M9 agar. In conclusion, reduction of uhpT function is largely responsible for the reduced sensitivity to FOM in clinical isolates that have not acquired FOM-modifying genes or mutations in murA. However, there are a few strains whose mechanisms of reduced susceptibility to FOM are still unclear.

Tigecycline Nonsusceptibility Occurs Exclusively in Fluoroquinolone-Resistant Escherichia coli Clinical Isolates, Including the Major Multidrug-Resistant Lineages O25b:H4-ST131-H30R and O1-ST648.

Tigecycline (TGC) is a last-line drug for multidrug-resistant Enterobacteriaceae We investigated the mechanism(s) underlying TGC nonsusceptibility (TGC resistant/intermediate) in Escherichia coli clinical isolates. The MIC of TGC was determined for 277 fluoroquinolone-susceptible isolates (ciprofloxacin [CIP] MIC, <0.125 mg/liter) and 194 fluoroquinolone-resistant isolates (CIP MIC, >2 mg/liter). The MIC50 and MIC90 for TGC in fluoroquinolone-resistant isolates were 2-fold higher than those in fluoroquinolone-susceptible isolates (MIC50, 0.5 mg/liter versus 0.25 mg/liter; MIC90, 1 mg/liter versus 0.5 mg/liter, respectively). Two fluoroquinolone-resistant isolates (O25b:H4-ST131-H30R and O125:H37-ST48) were TGC resistant (MICs of 4 and 16 mg/liter, respectively), and four other isolates of O25b:H4-ST131-H30R and an isolate of O1-ST648 showed an intermediate interpretation (MIC, 2 mg/liter). No TGC-resistant/intermediate strains were found among the fluoroquinolone-susceptible isolates. The TGC-resistant/intermediate isolates expressed higher levels of acrA and acrB and had lower intracellular TGC concentrations than susceptible isolates, and they possessed mutations in acrR and/or marR The MICs of acrAB-deficient mutants were markedly lower (0.25 mg/liter) than those of the parental strain. After continuous stepwise exposure to CIP in vitro, six of eight TGC-susceptible isolates had reduced TGC susceptibility. Two of them acquired TGC resistance (TGC MIC, 4 mg/liter) and exhibited expression of acrA and acrB and mutations in acrR and/or marR In conclusion, a population of fluoroquinolone-resistant E. coli isolates, including major extraintestinal pathogenic lineages O25b:H4-ST131-H30R and O1-ST648, showed reduced susceptibility to TGC due to overexpression of the efflux pump AcrAB-TolC, leading to decreased intracellular concentrations of the antibiotics that may be associated with the development of fluoroquinolone resistance.