Artificial intelligence and data experts improving healthcare systems in Vietnam

Researchers from RMIT Vietnam and the Oxford University Clinical Research Unit (OUCRU) have joined forces to tackle pressing challenges facing Vietnam’s healthcare sector.

SDGs

Artificial intelligence arrhythmia OUCR Vietnam healthcare dengue digital transformation

Researchers from RMIT Vietnam and the Oxford University Clinical Research Unit (OUCRU) have joined forces to tackle pressing challenges facing Vietnam’s healthcare sector.

Computer and data scientists are collaborating closely with local medical professionals across several projects. Together, they are developing artificial intelligence and machine learning solutions for issues including digitalising medical records and forecasting dengue outbreaks.

Lead researcher, Associate Professor Minh Ngoc Dinh said the strong partnership with OUCRU was particularly valuable because they have excellent links with local hospitals.

The partnership and direct access to medical staff who can tell us about their challenges ensures our work is truly impactful on the ground and making a positive difference for Vietnam’s health sector.

Other valuable partnerships include research with Tam Duc Heart Hospital in Vietnam to improve arrhythmia screening.

Minh Ngoc Dinh standing in front of a blue table with a robot sitting on it.

Associate Professor Minh Ngoc Dinh, Program Manager, Master of Artificial Intelligence at RMIT University, Vietnam

Digitalising medical records via machine learning technology

Vietnam is a developing country with a population of around one million and is just starting to focus on digitalising important physical documents such as medical records.

Tools that are currently available cannot recognise and transcribe Vietnamese handwritten text and challenges arise because the language comprises more character classes, complex vocals and tonal symbols than languages such as English.

However, Dinh said local doctors and nurses have emphasised the urgent need for digital systems to track their work and patient histories.

Digitalising old physical records will allow for more efficient and precise care, especially in emergency units and will also help to manage and prepare for future health emergencies.

Digital transformation tools

Dinh and colleagues from OUCR and the Hospital for Tropical Diseases in Ho Chi Minh City are working to develop an end-to-end-solution that transforms hard copy and handwritten documents into electronic versions.

Dinh said their approach entailed developing a prediction and machine learning pipeline that would take an image of a hard copy medical document and transcribe it into the digital version.

“We do so by performing segmentation to word-level and then leveraging an artificial neural network consisting of both convolutional neural network (CNN) and a long short-term memory recurrent neural network (LSTM) to propagate the sequence information,” said Dinh.

The team is continuing to build on this research, with the most recent results proving extremely promising. Using a dataset of 8000 scans of handwriting from 30 doctors and nurses, the Character Error Rate for the AI model (CER) was only 2%.

Impacts

Enhancing healthcare quality: The ability to share patient records between departments reduces unnecessary tests, streamlines treatment, and ultimately optimises patient care.

Generating a rich dataset: The dataset created through this system opens doors to a range of other potential medical machine-learning solutions. Healthcare collaborators are planning to use this data to develop expert diagnostic systems, to improve treatment processes and minimise errors in their healthcare practices.

Researchers from RMIT and the Oxford University Clinical Research Unit who are working together to tackle challenges facing Vietnam’s healthcare sector.

Digital tool for predicting and monitoring dengue outbreaks

A serious global health issue

Dengue is a mosquito-borne infection that is particularly prevalent in Vietnam and other tropical climates. The disease can cause serious symptoms and can become life-threatening without adequate treatment.

A dengue surveillance tool is urgently needed to help predict, monitor and manage Dengue outbreaks and provide publicly available and geographically specific information about their extent and location.

Dengue Advanced Readiness Tools (DART)

An RMIT research team worked with researchers from OUCRU on a pilot project to help develop the Dengue Advanced Readiness Tools (DART).

Mr Tuyen Huynh worked on the capstone project during his final year completing a Bachelor of Information Technology at RMIT Vietnam. He said the tool developed during the pilot project provides a semi-automated data pipeline to collect and structure data from various sources including geographical, climate and socio-economic factors.

“The ultimate goal is to produce a user-friendly, interactive web application that provides the most up-to-date status of potential or current dengue outbreaks,” he said.

“The ability to make more accurate forecasts and have the information publicly accessible will help the community better prepare for potential outbreaks.

“The system's scalability and adaptability to different data sources will also provide deeper insights about the impact that elements such as weather and climate have on dengue,” he said.

Huynh is focusing on the development of a time-series model that aims to predict the potential number of dengue cases in Ho Chi Minh City through analysing historical dengue incidence reports. The prediction will later be aligned with spatial aspects to present the comprehensive outline of any potential dengue outbreak.

Next steps

Huynh is continuing to work on the tool as a full-time PhD candidate at OUCRU.

Research partnership with Tam Duc Hospital

Artificial Intelligence to assist arrhythmia screening

Electrocardiograms (ECGs) play a pivotal role in assessing cardiovascular health by recording electrical activity of the heart. Given cardiovascular disease is a significant global health concern, there is a growing emphasis on utilising AI and modelling to predict heart health and related issues including arrhythmia.

Current models need refining

Dinh said that despite notable progress in harnessing ECG data to identify critical features for predicting and monitoring heart health in recent years, challenges persist.

“Not all machine-learning and AI applications are approved clinically to be used in healthcare as they must comply with stringent regulatory guidelines with the highest level of evidence,” he said.

RMIT study with Tam Duc Heart Hospital

The new machine-learning model being developed by RMIT Vietnam researchers has shown improved performance for helping non-cardiologists with arrhythmia screening.

The study demonstrated a significant improvement in Arrhythmias machine learning classifiers, the algorithms that are used to categorise data into distinct groups or classes.

It used a method comprising a fixed set of statistical features to extract information from the ECG data, along with categorising the dataset before using it to pre-train the machine learning application. With no deep learning involved, the models outperformed non-physicians in ECG screening, including medical students and highly trained nurses.

Next steps

The research team is continuing to work with Tam Duc Heart Hospital in Vietnam to further develop the AI model, with the aim of providing accurate diagnosis and explanations.

The project team is also working to create a rich ECG-based clinical dataset from Tam Duc Heart Hospital that can be used for further studies in this area.

Key contact

Associate Professor Minh Ngoc Dinh
Acting Deputy Dean, Research and Innovation
Program Manager, Master of Artificial Intelligence
School of Science, Engineering and Technology
RMIT University, Vietnam

Continue exploring

Read related RMIT impact stories

Understanding and tackling Spotty Liver Disease in poultry

RMIT University researchers identified the bacterial species that causes Spotty Liver Disease in chickens and developed diagnostic tools for the serious issue that is affecting global egg production industries.

Coffee concrete trial turns organic waste into a valuable resource

RMIT researchers collaborated with Macedon Ranges Shire Council to conduct a world-first coffee concrete footpath trial.

Improving gender representation in Australia’s cyber security workforce

RMIT research has provided greater insights into gender-related workforce challenges faced by the Australian cyber security sector and identified ways to expand and diversify the talent pool.

Improving understanding of cryptocurrency regulation, crime and sentencing

RMIT researchers found that using cryptocurrency to commit a crime leads to a harsher sentence than if a similar crime was committed using regular currency.

See more stories +

Connect with RMIT Research

If you're not sure how you can best work with us, our team can explain what's possible and put you in touch with the right person.

Acknowledgement of Country

RMIT University acknowledges the people of the Woi wurrung and Boon wurrung language groups of the eastern Kulin Nation on whose unceded lands we conduct the business of the University. RMIT University respectfully acknowledges their Ancestors and Elders, past and present. RMIT also acknowledges the Traditional Custodians and their Ancestors of the lands and waters across Australia where we conduct our business - Artwork 'Sentient' by Hollie Johnson, Gunaikurnai and Monero Ngarigo.

More information

Acknowledgement of Country

More information