Australian women are afraid that AI is used in breast cancer research – modern studies

Artificial intelligence (AI) is becoming more and more essential in many aspects of society, including healthcare. For example, it is already used for robotic surgery and virtual support of mental health.

In recent years, scientists have developed AI algorithms that can analyze mammograms in terms of breast cancer symptoms. These algorithms can be as good as or Better to find cancer than human radiologists and Save the money of the healthcare system.

At the same time, evidence of the accuracy of artificial intelligence in breast cancer screening it still appears. We must ensure that the benefits exceed the risk such as excessive diagnosis. Compact cancers are detected here that would not cause damage, which caused unnecessary treatment.

IN New studyMy colleagues and I wanted to understand how Australian women – which would be affected if AI were introduced into the breast screening in the future – rely on technology.

AI tests and breast cancer

Breast cancer screening programs reduce the number of women who die by finding cancer early.

In Australia, as in many countries around the world, two specially trained healthcare workers, usually radiologists, Review every mammographic test For signs of cancer. If two radiologists do not agree, the third consult.

This double approach to reading improves cancer detection indicators without reminding too many women for Further testing unnecessarily. However, this is intense resources. And now it is A shortage of radiologists all over the world.

AI was examined to support radiologists, replace a radiologist or as a binder tool to identify mammography at the highest risk so that they can view them through a radiologist. However, there is no consensus yet how to best implement artificial intelligence During breast cancer screening.

Our study

The success of cancer screening programs depends on the high participation indicator. While people are generally Receptive on AIIn previous studies, many reported that they do not want to trust AI With their health care.

There are fears to introduce artificial intelligence into screening programs towards breast cancer may threaten the indicators of participation in screenings if people do not trust artificial intelligence.

We He asked 802 women If and how they thought AI should be implemented during breast cancer screening. Our attempt was generally representative of the women’s population in Australia qualifying for screening.

We measured how their preferences influenced factors such as:

• How artificial intelligence was used (regardless of whether he supplemented radiologists, replaced one or both radiologists or used for segregation)

• How precise was the AI ​​algorithm

• which was the owner of the AI ​​algorithm (for example, the Australian Health Department, Australian company or international company)

• As a representative algorithm was the Australian population (for example, the algorithm may not work so well for people from some ethnic groups)

• How was privacy managed

• How long patients had to wait for the results of their mammogram.

We used the answers to assess which factors were the most essential and how the introduction of AI can affect participation in breast cancer screening.

Before the survey, we provided participants with information about artificial intelligence and how it can be used in breast cancer screening. The information we provided could change the beliefs and preferences of participants regarding the operate of AI in this context in relation to the general population. This may be a limitation of our study.

What we found

In general, we saw mixed reactions to the introduction of AI for breast cancer screening. About 40% of respondents were open to the operate of AI, provided that it was more precise than human radiologists. However, 42% were definitely against the operate of AI, and 18% had reservations.

In general, participants wanted artificial intelligence to be precise, Australian, representative for Australian women and faster than human radiologists before implementation.

In particular, up to 22% of respondents said that they could less often participate in breast cancer screening if AI was implemented in a way that made them uncomfortable.

It is possible that the attitudes towards AI may differ in contexts of different social values ​​or existing screening practices in Australia. But our findings were fundamentally consistent with what we see in other countries.

Women around the world generally sensitive to the benefits of AI During breast cancer screening. But they feel difficult that AI should Supplement or supporting cliniciansinstead Replace them.

We have to continue carefully

AI promises to improve the effectiveness and efficiency of breast cancer screening in the future.

After saying, these benefits may be balanced if the participation in a screening test falls. This is particularly disturbing in Australia, where the indicators of participation on the breast screen are already relatively low (less than 50%).

The implementation of artificial intelligence without solving social problems regarding accuracy, property, privacy and implementation model may undermine confidence in breast cancer screening programs.

Decision -makers should carefully consider the community concern regarding the implementation of AI technology in healthcare before continuing. And participants of screening for breast cancer will need reliable information to understand AI’s risk and benefits in the field of screening services.

If this is not done correctly, and the participation in the screening falls lower, it can lead to the diagnosis of more breast cancers, and thus more arduous to treat.