Injury pattern recognition to discriminate competing causes of liver injury

Abstract

Background Competing causes of liver injury may be difficult to discriminate. Characterization of the typical phenotype of each injury defined by latency, time to improvement and biochemical pattern, could be helpful to identify the most likely of competing causes. Methods Liver injury characteristics of both bortezomib‐associated drug‐induced liver injury (DILI) and hepatitis B virus (HBV) reactivation associated with bortezomib are derived from PubMed listed publications. Results Bortezomib‐associated DILI has very short latency of days and AP is found elevated, while liver injury due to HBV reactivation occurs after several months of bortezomib therapy. Therefore, a patient's liver injury pattern occurring 3 months into bortezomib therapy should be attributed to HBV reactivation. Discussion Identification of liver injury characteristics for competing causes of liver injury can be helpful to identify the most likely cause and improve clinical outcome.

A Novel Computer-Assisted Drug-Induced Liver Injury Causality Assessment Tool (DILI-CAT)

“…We speculate that this will be helpful to experts and, perhaps to an even greater degree, to nonexperts. 19 An algorithmic data-driven and drug-specific diagnostic tool such as DILI-CAT has a number of advantages. DILI-CAT's data-driven approach is objective.…”

“…We speculate that this will be helpful to experts and, perhaps to an even greater degree, to nonexperts. 19…”

“…In this regard, we speculate that the number of cases needed to develop a robust signature will depend on the consistency of the drug’s phenotype. 19 The more variable the phenotype, the more cases that are likely to be required to generate a precise picture of a drug’s signature.…”

Pattern Recognition of Grating Perimeter Intrusion Behavior in Deep Learning Method

“…Template adjustment module is mainly responsible for recording various event templates. The event template is obtained and stored in the database through cluster analysis of recorded event information through machine learning algorithm [11]. Users can view the frequency domain waveform of the template to understand the main intrusion and environmental noises in the environment.…”

“…In Equation (11), represents the number of points of ζ(τ), and ζ(τ) represents the kurtosis value corresponding to the IMF component. Through the above steps, selecting the time domain feature can effectively identify different intrusion behaviors, thereby obtaining the time domain feature vector of the intrusion signal.…”