Dataset Information

The clinical differentiation of bacterial and fungal keratitis: a photographic survey.

ABSTRACT:

Purpose

The purpose of this study was to determine whether clinical signs of infectious keratitis can be used to identify the causative organism.

Methods

Eighty photographs of eyes with culture-proven bacterial keratitis or smear-proven fungal keratitis were randomly selected from 2 clinical trials. Fifteen cornea specialists from the F. I. Proctor Foundation and the Aravind Eye Care System assessed the photographs for prespecified clinical signs of keratitis, and they identified the most likely causative organism.

Results

Clinicians were able to correctly distinguish bacterial from fungal etiology 66% of the time (P < 0.001). The Gram stain, genus, and species were accurately predicted 46%, 25%, and 10% of the time, respectively. The presence of an irregular/feathery border was associated with fungal keratitis, whereas a wreath infiltrate or an epithelial plaque was associated with bacterial keratitis.

Conclusions

Cornea specialists correctly differentiated bacterial from fungal keratitis more often than chance, but in fewer than 70% of cases. More specific categorization led to less successful clinical distinction. Although certain clinical signs of infectious keratitis may be associated with a bacterial or fungal etiology, this study highlights the importance of obtaining appropriate microbiological testing during the initial clinical encounter.

SUBMITTER: Dalmon C

PROVIDER: S-EPMC3342793 | biostudies-literature | 2012 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

The clinical differentiation of bacterial and fungal keratitis: a photographic survey.

Dalmon Cyril C Porco Travis C TC Lietman Thomas M TM Prajna N Venkatesh NV Prajna Lalitha L Das Mano Ranjan MR Kumar J Arun JA Mascarenhas Jeena J Margolis Todd P TP Whitcher John P JP Jeng Bennie H BH Keenan Jeremy D JD Chan Matilda F MF McLeod Stephen D SD Acharya Nisha R NR

Investigative ophthalmology & visual science 20120402 4

<h4>Purpose</h4>The purpose of this study was to determine whether clinical signs of infectious keratitis can be used to identify the causative organism.<h4>Methods</h4>Eighty photographs of eyes with culture-proven bacterial keratitis or smear-proven fungal keratitis were randomly selected from 2 clinical trials. Fifteen cornea specialists from the F. I. Proctor Foundation and the Aravind Eye Care System assessed the photographs for prespecified clinical signs of keratitis, and they identified ...[more]

PMID: 22395880

Similar Datasets

Project description:Purpose Develop computer vision models for image-based differentiation of bacterial and fungal corneal ulcers and compare their performance against human experts. Design Cross-sectional comparison of diagnostic performance. Participants Patients with acute, culture-proven bacterial or fungal keratitis from 4 centers in South India. Methods Five convolutional neural networks (CNNs) were trained using images from handheld cameras collected from patients with culture-proven corneal ulcers in South India recruited as part of clinical trials conducted between 2006 and 2015. Their performance was evaluated on 2 hold-out test sets (1 single center and 1 multicenter) from South India. Twelve local expert cornea specialists performed remote interpretation of the images in the multicenter test set to enable direct comparison against CNN performance. Main Outcome Measures Area under the receiver operating characteristic curve (AUC) individually and for each group collectively (i.e., CNN ensemble and human ensemble). Results The best-performing CNN architecture was MobileNet, which attained an AUC of 0.86 on the single-center test set (other CNNs range, 0.68–0.84) and 0.83 on the multicenter test set (other CNNs range, 0.75–0.83). Expert human AUCs on the multicenter test set ranged from 0.42 to 0.79. The CNN ensemble achieved a statistically significantly higher AUC (0.84) than the human ensemble (0.76; P < 0.01). CNNs showed relatively higher accuracy for fungal (81%) versus bacterial (75%) ulcers, whereas humans showed relatively higher accuracy for bacterial (88%) versus fungal (56%) ulcers. An ensemble of the best-performing CNN and best-performing human achieved the highest AUC of 0.87, although this was not statistically significantly higher than the best CNN (0.83; P = 0.17) or best human (0.79; P = 0.09). Conclusions Computer vision models achieved superhuman performance in identifying the underlying infectious cause of corneal ulcers compared with cornea specialists. The best-performing model, MobileNet, attained an AUC of 0.83 to 0.86 without any additional clinical or historical information. These findings suggest the potential for future implementation of these models to enable earlier directed antimicrobial therapy in the management of infectious keratitis, which may improve visual outcomes. Additional studies are ongoing to incorporate clinical history and expert opinion into predictive models.

Project description:Microbial keratitis is a major cause of blindness worldwide. An excessive host inflammatory response can occur even after adequate antimicrobial treatment. This results in tissue damage with corneal thinning and even perforation, which may require corneal transplantation. In this study we investigated the pathways involved in the pathophysiology of this disease by comparing the human transcriptome profile of tissue from culture-proven bacterial and fungal keratitis (n=7 and n=8 respectively) with normal non-infected cadaveric corneal tissue (C, n=12) using Illumina HT12 v4 microarrays. The causative organisms were Streptococcus pneumoniae (n=6) and Pseudomonas aeruginosa (n=1) for bacterial keratitis (BK). Fungal keratitis (FK) was caused by Fusarium sp. (n=5), Aspergillus sp. (n=2, A. flavus and terreus) and Lasiodiplodia sp. (n=1). Differential expression (DE) analysis revealed 2310 significantly altered probes in the BK v C comparison, and 1813 probes for FK v C. The most highly upregulated gene in both comparisons was MMP9 with fold changes (FC) of 64 (fdr-adjusted p<6 x10-11) for FK v C and 89 for BK v C (fdr-adjusted p<4 x10-11) respectively. Network co-expression analyses revealed the defense response, inflammatory response and extracellular matrix mechanisms to be the main functional pathways involved. Microarray results were validated by performing real-time quantitative PCR (RTqPCR) for 46 DE genes using RNA extracted from the same samples. There was a high correlation between log2 FC values from microarray and RTqPCR. Further studies are needed to evaluate the most highly differentially expressed genes as possible biomarkers of disease progression or therapeutic targets. Case - control study design. Corneal ulcer tissue from 8 bacterial and 9 fungal ulcers was excised at the time of corneal transplantation surgery and immediately preserved in RNALater. Non-infected corneal tissue from 13 cadaver corneas were the control tissue. Transcriptome profile generated using Illumina HT12 v4 beadchips. Differential expression analysis was performed with pairwise comparisons: bacterial ulcers versus controls, fungal ulcers versus controls and bacterial versus fungal ulcers. Microarray results validated with RTqPCR.

Project description:IntroductionInfectious keratitis is a vision threatening disease. Bacterial and fungal keratitis are often confused in the early stages, so right diagnosis and optimized treatment for causative organisms is crucial. Antibacterial and antifungal medications are completely different, and the prognosis for fungal keratitis is even much worse. Since the identification of microorganisms takes a long time, empirical treatment must be started according to the appearance of the lesion before an accurate diagnosis. Thus, we developed an automated deep learning (DL) based diagnostic system of bacterial and fungal keratitis based on the anterior segment photographs using two proposed modules, Lesion Guiding Module (LGM) and Mask Adjusting Module (MAM).MethodsWe used 684 anterior segment photographs from 107 patients confirmed as bacterial or fungal keratitis by corneal scraping culture. Both broad- and slit-beam images were included in the analysis. We set baseline classifier as ResNet-50. The LGM was designed to learn the location information of lesions annotated by ophthalmologists and the slit-beam MAM was applied to extract the correct feature points from two different images (broad- and slit-beam) during the training phase. Our algorithm was then externally validated using 98 images from Google image search and ophthalmology textbooks.ResultsA total of 594 images from 88 patients were used for training, and 90 images from 19 patients were used for test. Compared to the diagnostic accuracy of baseline network ResNet-50, the proposed method with LGM and MAM showed significantly higher accuracy (81.1 vs. 87.8%). We further observed that the model achieved significant improvement on diagnostic performance using open-source dataset (64.2 vs. 71.4%). LGM and MAM module showed positive effect on an ablation study.DiscussionThis study demonstrated that the potential of a novel DL based diagnostic algorithm for bacterial and fungal keratitis using two types of anterior segment photographs. The proposed network containing LGM and slit-beam MAM is robust in improving the diagnostic accuracy and overcoming the limitations of small training data and multi type of images.

Project description:Dysbiosis in the gut microbiome has been implicated in several diseases including auto-immune diseases, inflammatory diseases, cancers and mental disorders. Keratitis is an inflammatory disease of the eye significantly contributing to corneal blindness in the developing world. It would be worthwhile to investigate the possibility of dysbiosis in the gut microbiome being associated with Keratitis. Here, we have analyzed fungal and bacterial populations in stool samples through high-throughput sequencing of the ITS2 region for fungi and V3-V4 region of 16S rRNA gene for bacteria in healthy controls (HC, n = 31) and patients with fungal keratitis (FK, n = 32). Candida albicans (2 OTUs), Aspergillus (1 OTU) and 3 other denovo-OTUs were enriched in FK samples and an unclassified denovo-OTU was enriched in HC samples. However, the overall abundances of these 'discriminatory' OTUs were very low (< 0.001%) and not indicative of significant dysbiosis in the fungal community inhabiting the gut of FK patients. In contrast, the gut bacterial richness and diversity in FK patients was significantly decreased when compared to HC. 52 OTUs were significantly enriched in HC samples whereas only 5 OTUs in FK. The OTUs prominently enriched in HC were identified as Faecalibacterium prausnitzii, Bifidobacterium adolescentis, Lachnospira, Mitsuokella multacida, Bacteroides plebeius, Megasphaera and Lachnospiraceae. In FK samples, 5 OTUs affiliated to Bacteroides fragilis, Dorea, Treponema, Fusobacteriaceae, and Acidimicrobiales were significantly higher in abundance. The functional implications are that Faecalibacterium prausnitzii, an anti-inflammatory bacterium and Megasphaera, Mitsuokella multacida and Lachnospira are butyrate producers, which were enriched in HC patients, whereas Treponema and Bacteroides fragilis, which are pathogenic were abundant in FK patients, playing a potential pro-inflammatory role. Heatmap, PCoA plots and functional profiles further confirm the distinct patterns of gut bacterial composition in FK and HC samples. Our study demonstrates dysbiosis in the gut bacterial microbiomes of FK patients compared to HC. Further, based on inferred functions, it appears that dysbiosis in the gut of FK subjects is strongly associated with the disease phenotype with decrease in abundance of beneficial bacteria and increase in abundance of pro-inflammatory and pathogenic bacteria.

Dataset Information

The clinical differentiation of bacterial and fungal keratitis: a photographic survey.

Purpose

Methods

Results

Conclusions

Publications

The clinical differentiation of bacterial and fungal keratitis: a photographic survey.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets