We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.AP

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Applications

Title: Augmenting expert detection of early coronary artery occlusion from 12 lead electrocardiograms using deep learning

Abstract: Early diagnosis of acute coronary artery occlusion based on electrocardiogram (ECG) findings is essential for prompt delivery of primary percutaneous coronary intervention. Current ST elevation (STE) criteria are specific but insensitive. Consequently, it is likely that many patients are missing out on potentially life-saving treatment. Experts combining non-specific ECG changes with STE detect ischaemia with higher sensitivity, but at the cost of specificity. We show that a deep learning model can detect ischaemia caused by acute coronary artery occlusion with a better balance of sensitivity and specificity than STE criteria, existing computerised analysers or expert cardiologists.
Comments: Our attempts to produce what we considered to be an acceptable level of explainability from our algorithm have not yielded a satisfactory account of its internal logic and we do not feel this is acceptable from a clinical application. We will publish a fuller account of our work on this issue and its implications on the validation of clinical deep learning algorithms in the near future
Subjects: Applications (stat.AP); Artificial Intelligence (cs.AI); Quantitative Methods (q-bio.QM)
ACM classes: I.2.1
Cite as: arXiv:1903.04421 [stat.AP]
  (or arXiv:1903.04421v4 [stat.AP] for this version)

Submission history

From: Rob Brisk [view email]
[v1] Mon, 11 Mar 2019 16:33:10 GMT (546kb)
[v2] Wed, 13 Mar 2019 09:16:26 GMT (550kb)
[v3] Wed, 20 Mar 2019 13:19:45 GMT (550kb)
[v4] Mon, 18 Nov 2019 10:15:13 GMT (0kb,I)

Link back to: arXiv, form interface, contact.