Artificial intelligence in women’s healthcare: A glimpse of the future

Abstract

INTRODUCTION Artificial intelligence (AI) refers to a computer program’s capability to carry out tasks typically associated with human intelligence, including reasoning, learning, adaptation, sensory perception, and interaction.[1] AI has already become integral to daily life, powering technologies such as facial recognition, speech recognition in virtual assistants (e.g., Amazon Alexa, Apple’s Siri, Google Assistant, and Microsoft Cortana), and self-driving vehicles. In a recent interview, billionaire entrepreneur Elon Musk, who leads Tesla, X, and SpaceX, stated, I guess that we’ll have AI that is smarter than any one human probably around the end of next year.“ AI also drives innovation in healthcare, contributing to drug development, clinical decision-making, and quality assurance in radiology. The volume of AI-related research is growing; for instance, at the 29th World Congress of the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) in 2019, 14 abstracts specifically referenced AI, compared to a total of 13 abstracts across the previous six ISUOG World Congresses (2013–2018). In 1950, Alan Turing proposed a test, now known as the “Turing test,” to determine whether a machine exhibits intelligent behavior indistinguishable from a human’s. If an evaluator cannot differentiate between the two, the machine must have passed the test. The term ‘AI’ was later introduced by John McCarthy in 1956 during the Dartmouth Conference. Key milestones in AI history include IBM’s Deep Blue defeating world chess champion Garry Kasparov in 1997 and AlphaGo overcoming Lee Sedol, one of the top players of the ancient Chinese game Go, in 2016.[1] APPLICATION OF ARTIFICIAL INTELLIGENCE IN GENERAL MEDICINE In clinical practice, AI technologies promise to revolutionize healthcare by extracting valuable insights from the vast digital data generated during medical care. Traditionally, evidence-based medicine has relied on statistical methods to identify patterns and represent them as mathematical equations. However, through machine learning (ML), AI introduces advanced techniques that reveal intricate relationships within data—ones that cannot be easily expressed in a simple equation. This enables ML systems to tackle complex problem-solving like clinicians, carefully analyzing evidence to arrive at well-reasoned conclusions.[2] Below are some examples showcasing AI’s exceptional performance across various medical specialties. An AI-driven smartphone app now capably handles triaging 1.2 million people in North London to Accident & Emergency (A&E).[3] A recent study shows that AI correctly diagnoses pulmonary TB with a sensitivity of 95% and specificity of 100%.[4] Researchers demonstrate that AI had superior sensitivity and specificity than dermatologists when classifying previously unseen photographs of biopsy-validated lesions.[5] As of May 2020, more than 50 deep learning (DL)-based imaging applications have been approved by the USA Food and Drug Administration, spanning most imaging modalities, including X-ray, computerized tomography, magnetic resonance imaging, retinal optical coherence tomography, and ultrasound. Applications include the identification of cerebrovascular accidents, diabetic retinopathy, skeletal fractures, cancer, pulmonary embolism, and pneumothorax.[2] There is an ongoing clinical trial using AI to calculate target zones for head and neck radiotherapy more accurately and quickly than a human being.[2] HOW DOES ARTIFICIAL INTELLIGENCE INFLUENCE OBSTETRICS AND GYNECOLOGY? Direct-to-consumer maternal health (mHealth) applications have the potential to increase engagement and empower pregnant people in their healthcare. Three studies arose in Kenya, focusing on the following areas: askNivi, a free sexual and reproductive health information service where users seek factual information, followed by requests for advice and reporting symptoms. Tess, a prototype mHealth text messaging system for perinatal depression. Predicting a woman’s fertile window through data received by a wearable bracelet achieved 90% accuracy.[6] AI assessment of embryo images or videos can improve ART outcomes. Applications guide the identification of embryos from the culture medium during early human in vitro development; raw time-lapse videos/images of embryos are sent to “in vitro fertilization” electronic health record (EHR) data.[6] A semi-automated learning-based framework approach was reported to improve gestational age prediction, with an accuracy of ±6.1 days, using DF on structural brain image and clinical data.[7] The fetal heart rate (FHR) signal is more complicated to identify than the adult signal. Therefore, there has been an effort to improve methods to detect, sample and quantify the FHR signal accurately. Various studies applied ML methods to classify cardiotocography signals and determine the fetal state.[6] In cases of preterm birth, cervix properties were observed using AI methods to determine material properties and cervical length (CL). The perinatal outcome was predicted with a DL model interpreting amniotic fluid metabolomics and proteomics in asymptomatic pregnant women with short CL.[6] For the mode of delivery, the Adana System applied artificial neural network to classify between cesarean section and vaginal delivery, including input variables of maternal characteristics and labor information.[8] Predicting the delivery route can inform care, allow appropriate allocation of resources, and improve pregnancy outcomes. ARTIFICIAL INTELLIGENCE: BENEFITS AND CHALLENGES One of the key advantages of AI is its superior reproducibility compared to humans. AI maintains absolute consistency over time, unlike clinicians, whose performance may vary due to experience, fatigue, or distractions. Additionally, AI systems have a significantly higher processing capacity—while a radiographer may analyze 50–100 scans daily, AI can theoretically interpret thousands.[1] AI also enhances efficiency by extracting valuable insights from a patient’s electronic records. Initially, this improves workflow and saves time, but with rigorous validation, AI could potentially play a direct role in patient management. Moreover, AI can serve large populations, particularly beneficial in areas with limited medical expertise. These systems continuously learn from each case and can process millions of cases within minutes, leading to highly efficient results. However, can AI replace human doctors? Despite its capabilities, AI lacks essential human qualities such as empathy and compassion, which are fundamental in patient care. Patients must feel that human physicians lead their consultations. Ethical concerns also arise with AI’s role in medicine. Should we trust AI to screen for diseases, prioritize treatment, diagnose, and discharge patients? Would we allow an AI system to determine which patient gets the last available intensive care unit bed? Additionally, data privacy remains a major concern. AI development requires vast amounts of patient data, raising questions about security and confidentiality. While AI presents significant advancements in healthcare, addressing ethical challenges and ensuring patient trust remain critical considerations. There is a concerted effort to incorporate AI into clinical practice, presenting an opportunity to introduce a “third participant” in patient care—one that can actively contribute to healthcare delivery.[1] However, stronger interdisciplinary collaboration between AI developers and healthcare professionals is essential for this potential to be fully realized. To facilitate the seamless integration of AI, medical professional organizations should begin assessing its impact, encourage physicians to document and share their experiences with AI technologies, and establish relevant guidelines or committees to address AI-related considerations.[1] From a financial perspective, AI can potentially reduce workforce requirements in labor-intensive fields such as healthcare. A well-known example from science fiction is the depiction of childbirth assisted by a “Midwife Droid” in Star Wars: Revenge of the Sith, a robot equipped with AI. While this remains fictional for now, rapid advancements in AI could soon make such scenarios a reality. The application of AI in obstetrics holds great promise. It can potentially enhance the quality of maternal and neonatal care, particularly in remote areas, thereby reducing maternal and perinatal morbidity and mortality. The future of AI in healthcare appears highly promising, offering innovative solutions to improve patient outcomes and accessibility. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.

Ähnliche Arbeiten