Skip to main content

Table 5 – Instances Corrected by Adding SDP-based Module

From: Integrating shortest dependency path and sentence sequence into a deep learning framework for relation extraction in clinical text

Relation Type

Sentence Sequence

SDP

TrWP

Subsequent discontinuance of [azithromycin]treatment, [trial_of_5-fc]treatment, with [increasing neutropenia]problem requiring discontinuance, change if [itraconazole]treatment to [voriconazole]treatment, given [continued neutropenia]treatment, and trial of [sulfadiazine]treatment, discontinued for [increasing ars]treatment

[trial_of_5-fc]treatment – appos →[azithromycin]treatment – nmod →discontinuance– nmod →[increasing neutropenia]problem

TrNAP

[His cast]treatment was removed by the orthopedic service in anticipation of [this edema]problem and to avoid [compartment syndrome]problem

[His cast]treatment –nsubjpass→ removed – nmod →service– acl → avoid –dobj→ [compartment syndrome]problem

TrIP

[His hypertension]problem; [his high blood pressure]problem was controlled with [intravenous nitroglycerine]treatment in the early going and then he was switched to [an oral regimen]treatment for better control after he was removed from the intensive care unit

[his high blood pressure]problem –nsubjpass→controlled– nmod →[intravenous nitroglycerine]treatment

  1. The italics in each sentence sequence are the candidate pair entities