BLOG POST
TECH UP THOUGHTS BY MEG HATTON, TECH UP FOR WOMEN TEAM
As artificial intelligence becomes intertwined in all aspects of life, both lovers and skeptics of this new technology have expressed their views. Artificial intelligence has the capability to reinvent the way we accomplish daily tasks, powering deep learning algorithms that can make predictions based on certain data. In the healthcare industry, new technologies have revolutionized the way doctors diagnose and treat patients, and at Tech Up For Women, we are exploring how artificial intelligence is affecting women’s health
In the fight against breast cancer, new technology has become crucial. At MIT, researchers developed a deep learning algorithm that assesses breast density from mammograms. Additionally, with help from medical centers and universities, Google designed an algorithm to read mammograms with a lower rate of false negatives and false positives than when read by a radiologist. Algorithms to identify precancerous changes in womens’ cervix have also been developed. With the help of artificial intelligence, women can have access to better treatments, and early diagnosis of cancers.
Along with these positive impacts, there exists the possibility that artificial intelligence will perpetuate age-old biases in the medical fields. When it comes to health care, women, people of color, and other underrepresented groups have been the victim of many biases. Though a plethora of health research may exist, it must be noted that, historically, white males were the subject of this research. In fact, women of childbearing age were actively excluded from most clinical trials until 1993. As new technologies are trained and used on existing data, algorithms can “learn” biases that are based on this data, and make predictions that harm underrepresented groups.
Though artificial intelligence is not “racist”, nor does it truly “think” like humans, systems can inherit biases through the data they are fed. Additionally, with women representing just 12% of machine learning researchers, the potential for implicit biases is concerning. To prevent negative effects, and ensure artificial intelligence is achieving its true potential, these biases must be recognized and mitigated.
Artificial intelligence has the capability to transform women’s health, and its impact is already being seen, with breakthrough technologies that can give more accurate data than doctors. However, with AI, the historical biases in the healthcare industry against women, people of color, and other underrepresented groups have the potential to stick around and impact doctors’ decisions. Thus, developers must move forward with caution, and with intention to negate their effects. To learn more about this topic, register for our IFA Berlin conference where Nickey Keay, Chief Medical Officer at Forth, will be presenting a talk on pioneering women’s health through AI. https://techupforwomen.com/ifa-2021-agenda/
Sources:
