New NUHS AI can spot high blood calcium levels
The National University Health System has developed a new AI tool that can detect cases of hypercalcemia in real time.
WHAT IT DOES
The AI, called CalSense, screens blood calcium level reports across the organization to flag cases of hypercalcemia, which are then displayed in real time via a live dashboard.
It also checks other research test results to “more accurately” identify the underlying causes of patients’ hypercalcemia.
The AI is hosted within the NUHS Endeavor AI platform, which also displays other medical information from patients across the organisation.
WHY IT MATTERS
Hypercalcemia is a result of overactive parathyroid glands, which regulate blood calcium levels, according to NUHS. The condition can be caused by cancer, certain medical conditions, some medications, and excessive consumption of calcium and vitamin D supplements.
“CalSense allows us to automate a process that previously relied on doctors to review blood test results when patients return for clinical assessment,” said NUHS Associate Professor Ngiam Kee Yuan.
“It identifies cases requiring rapid interventions in our network of hospitals, national specialty centers and outpatient clinics in real time, reducing doctors’ administrative work and accelerating medical care.”
CalSense could also potentially shorten the period of diagnosis and treatment of hypercalcemia, which takes an average of four years, A/Prof Ngiam said, citing global studies.
“If patients do not receive timely treatment, they may experience complications from hypercalcemia, including osteoporosis and fractures, kidney stones, kidney failure, mental disorders and even possible sudden cardiac death,” added NUHS Assistant CTO Dr. James Lee.
NUHS said CalSense can identify 1,600 cases of abnormal blood calcium levels in 26,000 tests in a trial carried out from January to July this year.
THE BIG TREND
NUHS is also trying out another AI project on Endeavor AI. Called Pathfinder, the AI, provides real-time insights into bed occupancy rates, wait times, and predictive insights into daily attendance at National University Hospital ED.
Another AI in the NUHS pipeline will support the X-ray assessment of scoliosis by automatically measuring the degree of scoliosis.