Faster sepsis diagnosis with AI and centrifugation method

Sepsis remains one of the most dangerous medical emergencies worldwide. Rapid diagnosis is crucial because every hour of delayed treatment increases mortality risk. A new diagnostic approach from KTH Royal Institute of Technology and Uppsala University, published in Nature Publishing Journal Digital Medicine, could cut the waiting time for confirmation of bloodstream infections from days to just a few hours.

“Diagnosing sepsis is a race against time,” says Henar Marino Miguelez, doctoral student at KTH. “With every hour of delayed treatment of patients in septic shock, survival rates drop by 8 percent.”

In some cases, it can take two to four days before doctors are certain which antibiotic to administer. As a precaution, many clinics begin treatment with broad-spectrum antibiotics. While this can save lives, it also brings risks such as drug toxicity, disruption of healthy gut bacteria, and the spread of antibiotic resistance.

“It takes a hospital two to four days before they are sure which antibiotic to treat a bloodstream infection with,” explains Professor Wouter van der Wijngaart of KTH. “We’re trying to do this in four to six hours.”

The new system combines several advanced techniques:

• Centrifugation: Blood samples are spun on top of a separating agent, causing bacteria to float upwards while blood cells sediment downwards.
• Microfluidic chip: The bacteria-containing layer flows into microscale channels. Tiny traps capture bacteria while excluding blood cells.
• Automated microscopy: Time-lapse images visualize bacterial growth.
• Artificial intelligence: Machine learning software analyzes the images and identifies the bacteria.

The technique was less effective with Staphylococcus aureus, a bacterium that tends to hide within blood clots. Nevertheless, the results show significant potential for faster diagnostics and better-targeted antibiotic therapy.

By enabling earlier identification of pathogens, hospitals can reduce reliance on broad-spectrum antibiotics, tailor treatments more effectively, and potentially improve survival rates for patients with sepsis.

MEDICA-tradefair.com; Source: KTH The Royal Institute of Technology