Swedish AI healthcare technology could halve radiologists’ workload in breast cancer detection
In a significant development emerging from Sweden, AI healthcare technology could potentially cut radiologists’ workload nearly in half, specifically concerning the detection of breast cancer indications within routine scans. Though excited by these initial findings, the researchers on this project emphasised the necessity for further examination before AI can be utilised widely in screening for breast cancer.
Often, it is advanced chatbots such as ChatGPT that spark discussions regarding potential future usage of AI. However, the technology has already proved successful in analysing medical scans. As a global shortage of radiologists persists, AI could potentially expedite and enhance the accuracy of routine scan analysis.
The implications could be particularly significant in the fight against breast cancer. According to figures from the World Health Organisation, more than 2.3 million women were diagnosed with the disease in 2020, which resulted in 685,000 fatalities.
Timely detection through regular screening can play a crucial role in identifying the early signs of cancer. In Europe, it is typically recommended that women aged between 50 and 69 undergo a mammogram every two years. These scans are normally examined by a pair of radiologists.
The research conducted in Sweden was based on the mammograms of 80,000 women at four locations in the south-west of the country from April until July of the past year.
The scans were divided at random for analysis between an AI support system and a control group comprising of two human radiologists. The AI program read the scans and assigned a cancer risk score from one to ten, which was subsequently verified by a radiologist.
Impressively, the AI-aided system was able to identify 20% more instances of cancer, which came to an extra detection for every thousand women screened. Both the AI system and human radiologists were found to have the same rate of false positives – instances where a mammogram initially appears worrisome but is later cleared – at 1.5%.
These results represent a momentous foresight into the utilisation of AI within healthcare and marks a substantial stride in improving cancer detection and treatment procedures.
In related news, New TB treatment plan unveiled by Philippines’ Health Department to curb rising cases.
