The first three PhD graduates of Mohamed bin Zayed University of Artificial Intelligence are working to use AI more efficiently in health care and data analysis.
Numan Saeed, William de Vazelhes and Hilal Mohammad Hilal Al Quabeh received their doctoral degrees in machine learning on Thursday, along with 98 master's degree graduates at the commencement ceremony in Abu Dhabi.
MBZUAI, which officially opened in 2021, aims to strengthen the UAE’s position as a global destination for AI-based research and innovation.
Mr Saeed, 32, from Pakistan, is confident of using AI to enhance the diagnosis and prognosis of head and neck cancer, which is the seventh-most common cancer globally.
PhD graduate of MBZUAI
His model can interpret PET and CT scans, decipher doctors’ notes using a natural language processing (NLP) solution and factor in variables such as a patient’s age, gender, weight, previous treatment and whether they smoke or drink alcohol.
“These cancers can appear in various locations within the head and neck, posing difficulties in early detection due to differing symptoms," said Mr Saeed.
“The challenges lie not only in its detection but also in accurately pinpointing its location, which our AI models aim to address."
Mr Saeed, who decided to quit his job in the aviation industry to pursue his passion, is confident AI will "significantly alleviate the burden on healthcare systems, and in early cancer detection".
His goal was "to come up with more efficient, sustainable and cost-effective solutions, using AI to help people get treatment, more specifically in remote places".
Advancing cancer treatment
Mr Saeed's next step is to trial his model at hospitals, after which he would consider establishing a start-up.
While acknowledging that AI is "quite disruptive", he believes it "is going to change everything".
"We must embrace it and utilise it. If we look at health care, it still has quite a long way to go, probably because we are quite cautious.
“But if we come up with models that are reliable, it can benefit people not only in remote areas but in any place."
Mr Saeed believes the impact of his research, combined with technology such as portable ultrasound scanners, could be useful in countries where oncological services are limited and there is a shortage of specialist clinicians.
He is currently working as a postdoctoral graduate at MBZUAI.
Making AI more efficient
Both Mr de Vazelhes and Mr Al Quabeh have focused on using computing resources to make AI work more efficiently.
Mr de Vazelhes, 30, the first French citizen to graduate from MBZUAI, has focused on optimisation algorithms, which is the mechanism through which machine-learning models learn from training data. It looks at a special type of algorithm called hard thresholding that enables AI to simplify and focus on the most critical information, making data easier to analyse and work with.
This research will prove vital as AI becomes more ubiquitous across industries. This is because the energy demands of AI models in their current form are not sustainable due to the intense demand for computing power and energy.
Mr de Vazelhes, who had earlier worked as a research engineer, is confident his PhD "will allow [him] to collaborate as a researcher in companies or government entities", using his model for medical imaging, and text and image analysis.
Jordanian citizen Mr AlQuabeh, 31, also concentrated on the need to improve AI efficiency in his research because "if not addressed, it could have a detrimental effect on sectors including health care, transport and logistics, and agriculture – particularly in remote areas – in the future".
His model looked at how machines can be taught to "enhance critical problem-solving and spot nuanced details that might be ordinarily missed, even with limited information and resources".
Mr Al Quabeh used the spiking neural network algorithm that uses low energy compared to normal algorithms, making it beneficial for applications with limited power, like drones, mobile phones or agriculture devices.
He is keen to continue working on theoretical machine-learning challenges and conduct research at MBZUAI.
