Deep Learning Fuzzy Inference: An Interpretable Model for Detecting Indirect Immunofluorescence Patterns Associated with Nasopharyngeal Cancer.

The detection of serum Epstein-Barr virus antibodies by immunofluorescence assay (IFA) is considered the gold standard screening test for nasopharyngeal cancer (NPC) in high-risk populations. Given the high survival rate after early detection in asymptomatic patients, compared to the poor prognosis in patients with late-stage NPC, screening using IFA has tremendous potential for saving lives in the general population. However, IFA requires visual interpretation of cellular staining patterns by trained pathology staff, making it labor intensive and hence nonscalable. In this study, an automated fuzzy inference (FI) system achieved high agreement with a human IFA expert in identifying cellular patterns associated with NPC (κ = 0.82). The integration of a deep learning module into FI further improved the performance of FI (κ = 0.90) and reduced the number of uncertain cases that required manual evaluation. The performance of the resulting hybrid model, termed deep learning FI (DeLFI), was then evaluated with a separate testing set of clinical samples. In this clinical validation, DeLFI outperformed human evaluation on the area under the curve (0.926 versus 0.821) and closely matched human performance on Youden J index (0.81 versus 0.80). Data from this study indicate that the combination of deep learning with FI in DeLFI has the potential to improve the scalability and accuracy of NPC detection.

Beta-Hemolytic Bacteria Selectively Trigger Liposome Lysis, Enabling Rapid and Accurate Pathogen Detection.

For more than a century, blood agar plates have been the only test for beta-hemolysis. Although blood agar cultures are highly predictive for bacterial pathogens, they are too slow to yield actionable information. Here, we show that beta-hemolytic pathogens are able to lyse and release fluorophores encapsulated in sterically stabilized liposomes whereas alpha and gamma-hemolytic bacteria have no effect. By analyzing fluorescence kinetics, beta-hemolytic colonies cultured on agar could be distinguished in real time with 100% accuracy within 6 h. Additionally, end point analysis based on fluorescence intensity and machine-extracted textural features could discriminate between beta-hemolytic and cocultured control colonies with 99% accuracy. In broth cultures, beta-hemolytic bacteria were detectable in under an hour while control bacteria remained negative even the next day. This strategy, called beta-hemolysis triggered-release assay (BETA) has the potential to enable the same-day detection of beta-hemolysis with single-cell sensitivity and high accuracy.

Increasing the accuracy and scalability of the Immunofluorescence Assay for Epstein Barr Virus by inferring continuous titers from a single sample dilution.

High Epstein Barr Virus (EBV) titers detected by the indirect Immunofluorescence Assay (IFA) are a reliable predictor of Nasopharyngeal Carcinoma (NPC). Despite being the gold standard for serological detection of NPC, the IFA is limited by scaling bottlenecks. Specifically, 5 serial dilutions of each patient sample must be prepared and visually matched by an evaluator to one of 5 discrete titers. Here, we describe a simple method for inferring continuous EBV titers from IFA images acquired from NPC-positive patient sera using only a single sample dilution. In the first part of our study, 2 blinded evaluators used a set of reference titer standards to perform independent re-evaluations of historical samples with known titers. Besides exhibiting high inter-evaluator agreement, both evaluators were also in high concordance with historical titers, thus validating the accuracy of the reference titer standards. In the second part of the study, the reference titer standards were IFA-processed and assigned an 'EBV Score' using image analysis. A log-linear relationship between titers and EBV Score was observed. This relationship was preserved even when images were acquired and analyzed 3days post-IFA. We conclude that image analysis of IFA-processed samples can be used to infer a continuous EBV titer with just a single dilution of NPC-positive patient sera. This work opens new possibilities for improving the accuracy and scalability of IFA in the context of clinical screening.

Identification of apnea during respiratory monitoring using support vector machine classifier: a pilot study.

To determine the use of photoplethysmography (PPG) as a reliable marker for identifying respiratory apnea based on time-frequency features with support vector machine (SVM) classifier. The PPG signals were acquired from 40 healthy subjects with the help of a simple, non-invasive experimental setup under normal and induced apnea conditions. Artifact free segments were selected and baseline and amplitude variabilities were derived from each recording. Frequency spectrum analysis was then applied to study the power distribution in the low frequency (0.04-0.15 Hz) and high frequency (0.15-0.40 Hz) bands as a result of respiratory pattern changes. Support vector machine (SVM) learning algorithm was used to distinguish between the normal and apnea waveforms using different time-frequency features. The algorithm was trained and tested (780 and 500 samples respectively) and all the simulations were carried out using linear kernel function. Classification accuracy of 97.22 % was obtained for the combination of power ratio and reflection index features using SVM classifier. The pilot study indicates that PPG can be used as a cost effective diagnostic tool for detecting respiratory apnea using a simple, robust and non-invasive experimental setup. The ease of application and conclusive results has proved that such a system can be further developed for use in real-time monitoring under critical care conditions.