In the UK, only one in 100 people could identify the main symptoms of diabetes and only one in five would be able to spot a single one. Undiagnosed and untreated, it can lead to serious illness, amputation and even death.
The diagnostics process in the healthcare industry is an important yet complex task. The process starts off with the patient first experiencing a health problem, thereafter deciding whether or not to engage the healthcare system. This is critical given the earlier a health problem is identified, the easier it is for healthcare professionals to treat the patient.
“The challenge, however, is that not many know how to accurately assess their condition. How many then actually go undiagnosed until their health problems worsen?”
Even then, when patients decide to engage the healthcare system, doctors take time to diagnose the problem. The process would include interviewing the patient about his or her clinical history, physically examining the patient, performing diagnostic testing, and referring or consulting with other clinicians, all so as to amass information that may be relevant to understanding a patient’s health problem.
In Singapore, diabetes is common and increasingly so. One in three diabetics were unaware that they had diabetes! It doesn’t help that many would wrongly identify weight gain as a symptom, due to increased awareness of the link between obesity and diabetes.
Fortunately, the advancement in technology and specifically artificial intelligence (AI), enables healthcare professionals to speed up the diagnostic process, mitigating the challenge of misdiagnoses & untreated chronic disease.
Let’s use an example of Luke, a 62 year old Chinese male.
Phase 1: As there is no one medical report that fits all patients, AI can be used to extract Luke’s historical reports which is especially helpful if he might have forgotten what health problems he or his parents might have had in the past.
Phase 2: From a medical knowledge base (reflected on the left of the diagram below), Natural Language Processing (NLP) is used to extract value-adding outcomes. It helps match the symptoms of diabetes and other related diseases with the information extracted from Luke’s medical reports (reflected on the right of the diagram below).
Thus, if there’s a match such as High Blood Pressure as seen below, the AI would be able to pick it up almost instantly.
Here’s a simplified model of what it might look like:
Covid-19 has brought into light the importance of ensuring our healthcare professionals’ time is well utilized and also the importance of an early diagnosis. This has driven many professionals to look to the capabilities of AI to further leverage on its capabilities. We had the opportunity to share 2 potential use cases of TAIGER’s AI capabilities with over 80 healthcare professionals in Singapore at a closed door webinar organized by SingHealth.
If you’re interested to read more about AI in Healthcare, check out this article we wrote previously on how we can overcome the strain on healthcare with AI. Alternatively, you could speak with us to explore how we can help your organisation revolutionise the healthcare industry.