top of page

MED+TECH: Using Artificial Intelligence to Address Pressing Needs in the Emergency Room

Part 1

Medicine is complex and ever evolving. From the pressure experienced at a high rate in the ERs to the precision required for a Surgical procedure. Practicing medicine has proven to be the epitome of stellar human performance. Medical professionals and institutions are always faced with unprecedented health challenges that can be as subtle as an undetected rare disease to a full-blown pandemic, such as the infamous Covid-19 Pandemic. Either scenario, all hands-on-deck is the default expectation from all professionals engaged in Medicine. Here, supporting tools can be useful and life-saving at most times. AI, being a pinnacle of technology to ever be invented by humans, has so much potential in saving time, cost, and energy in supporting Medical professionals in performing time-sensitive, precision-driven, and cognitive-heavy medical procedures...and this is just the tip of the iceberg in terms of integrating AI to Medicine.

In this blog post, I will focus on the Emergency Department due to one important variable...TIME.

The unpredictability and diversity of incoming patient cases coupled with the pressure of time to decide as to what ought to be done; so to optimize the survival rate of the patient at ANY given time is truly unique. This challenge itself is inviting a plethora of potential algorithmic models that can aid the Emergency Department in managing incoming patients better, managing ongoing patient cases upon escalation ⬆ or deescalation ⬇, and also optimizing the rate of patients that are to be transferred to different departments at an optimized time. When I was the Chief of Department of Biomedical Engineering at a Multi-Specialty Tertiary Teaching Hospital, I understood that it is necessary to support the frontline staff (Full-time Professionals, Doctor Interns and Residents) by eliminating technologies that are not usable, not easy to manipulate or are too technical for ANY professional's level of utilization. I focused in making sure that the frontline technologies that are (over)used by different professionals to be upgraded with intelligent devices that are not overwhelming in training and has low learning curve. This way, the AI-enabled medical devices that were in circulation in the Emergency Department were aspired to save several minutes in any patient survival rate, hence, contributing to their overall potential of life continuity.

Now back to AI as a tool itself, the reason why these AI-Enabled medical devices were capable of optimizing time is due to the fact that they were armored by a Clinically-sound Decision Support Systems (CDSS) that enabled the physicians to diagnose faster and decide accurately what the next steps must be at a time-sensitive episode.

Here is a concrete example of such AI-enabled device that helped in the optimization problem above:

A mobile ECG Machine.

Artificially-enhanced electrocardiography circulating in an Emergency Department proved to be a useful investment in diagnosing symptoms of potential heart problem at a faster rate, i.e. chest pain, heart palpitations, dizziness and shortness of breath. ECG can detect arrhythmias – irregular heart beats, coronary disease, heart attacks, cardiomyopathy... implying a potential cardiac disorder under wraps. The AI CDSS model embedded in such a device helped in supporting the physicians to accurately diagnosing and transferring the patient to the designated specialty department (whether in-house or another healthcare system) – and at times, the exact specialist on call. How efficient that is!

From the diagram above, you can see that a deep learning approach : Convolutional Neural Network (CNN) is recommended in detecting arrhythmia from ECG data. The reason for this is because CNN is able to classify ECG signals directly WITHOUT human interference – which is exactly what is needed when TIME is of the essence. More on this on later posts, I promise.😉


During my tenure, I chose Nihon Kohden CardiaFax as the the AI-enabled ECG machine for ER operations and here is why:

AI-Enabled Analysis:

  • Advanced 12-lead ECG analysis: it provides simultaneous 12-lead ECG acquisition of up to 24 seconds, plus analysis with approximately 200 findings and five judgment categories. It can also find typical waveforms of Brugada-type ECG.

  • Breakthrough synECi18 technology: - It provides 18-lead ECG information from a standard 12-lead ECG through synthesis of the additional leads V3R to V5R and V7 to V9, to help identify invisible ISCHEMIA. - Diagnostic inaccuracy may cause harmful delays, especially if presentation is not typical or initial 12-lead ECG is negative. Timely ischemia detection with synECi18 may preven myocardial damage or shorten time to percutaneous coronary intervention (PCI) indication. - Particularly in emergencies, synECi18 is regarded as a useful triage tool to enhance outcomes through early recognition and stratification. With the same workload and cost as for standard ECG procedures, patient safety is optimized and time to reperfusion may be reduced.

  • High-performance compliance with IEC 60601-2-25:2011: It complies with the IEC60601-2-25:2011 standard, which indicates high accuracy of signal processing (AC filtering), ECG measurement, and ECG analysis. It therefore supports accurate ECG diagnosis.

P.S NOT sponsored by Nihon Kohden

Like what you read? Stay tuned for the next MED+TECH Posts! Join my exclusive list:

Author: Kat Usop, MSHI



Next MED+TECH: AI-Enabled Electrocardiogram – Outpatient / ER Frontline

bottom of page