Artificial intelligence (AI) in the medical field is gaining much traction in clinical practices. With recent advances through deep learning and machine learning, AI technology has been commercialized in a number of medical disciplines due to the development of studies in big data and artificial neural networks used to analyze medical images.
Obstetrics and gynecology are the medical fields concerning pregnancy, delivery care, and female reproductive health. This division has not integrated AI into many of its practices due to ethical concerns and a lack of research regarding its implications. But considering its potential benefits in creating treatment plans and interpreting fetal measurements, perhaps the field ought to adopt it.
Current and future AI applications of AI in obstetrics
AI has become very prevalent in the field of diagnostic methods and could be used for such purposes in obstetrics and gynecology.
For example, in obstetrics, cardiotocography (CTG) — developed in the 1960s — is the chief means of assessing fetal health, through measurements of uterine contractions and the fetus’ heart rate in utero. When interpreting measurements, different clinicians can come to different conclusions about the same number. Therefore, implementing AI into CTG practices may prove highly effective because it could mean avoiding poor communication, fatigue and distraction, cognitive overload, or bias.
A second area in obstetrics where AI might be fruitfully applied is in the analysis of ultrasounds: non-invasive checkup routines for prenatal diagnosis. Screening ultrasounds manually, as is the common practice, is slow and susceptible to human error. Consequently, if we pair these screenings with AI, the potential improvements in accuracy and standardization might be valuable in years to come.
In the field of gynecology, AI application has also been rather slow. However, in recent years, researchers have used AI and machine learning to develop new means of managing certain gynecological conditions, predicting disease progression, and navigating treatment decisions and protocols. For example, AI has recently been employed to analyze images of the pelvic region and recognize the presence of endometrial tissue, a potential signifier of endometriosis. Further, researchers use AI to evaluate imaging data and predict the growth and behaviour patterns of fibroids — benign tumours that typically grow in the uterus. It provides aid in developing personalized treatment plans.
Ethics of AI in healthcare
Ethical concerns have, naturally, prevented doctors’ complete acceptance and integration of AI in obstetrics and gynecology.
As is the case for many areas of AI, many patients have concerns about replacing human expertise — particularly when babies are involved. Researchers should explore whether such technologies have true value in patients’ diagnoses and treatment plans.
In obstetrics, people hesitate to let computer systems make decisions for treatment plans because of how personal the care is. Some also fear that pregnancy, a generally natural process, may turn into a highly medicalized one, which in turn may have physical and psychological implications for pregnant individuals.
It is important to note that research about women’s health has historically lagged behind that about men, leading to the underrepresentation of female subjects in clinical trials; a lack of sex-specific data on certain medications and treatments; and limited attention to conditions that most often happen in women, such as endometriosis — a disease in which uterine tissues grow outside of the uterus — and postpartum depression. Recognizing this disparity may aid in understanding why obstetrics and gynecology have been much slower in adopting and integrating AI technology in medical procedures.
AI can help improve diagnosis, treatment strategy, and clinical outcomes in obstetrics and gynecology. It is not unfeasible to consider the vast improvements that can be made to medical procedures to facilitate pregnancy management and public health. Still, it is important to recognize that AI technology should not substitute medical staff but rather play the role of an assistant in clinical practices.
