Artificial Intelligence Revolutionizes Polycystic Ovary Syndrome Diagnosis

News

Artificial Intelligence Revolutionizes Polycystic Ovary Syndrome Diagnosis

Artificial Intelligence Revolutionizes Polycystic Ovary Syndrome Diagnosis.

According to a recent study by the National Institutes of Health (NIH) in the United States, artificial intelligence (AI) and machine learning (ML) have proven effective in detecting and diagnosing Polycystic Ovary Syndrome (PCOS), the most common hormonal disorder in women, typically occurring in females aged 15 to 45.

Researchers systematically reviewed published scientific studies that utilized AI/ML analysis of data for diagnosing and categorizing PCOS. They found that AI/ML-based programs could successfully detect PCOS.

Artificial Intelligence Revolutionizes Polycystic Ovary Syndrome Diagnosis

Image source: NIH

“In light of the significant burden of underdiagnosis and misdiagnosis of PCOS in the community and its potential serious consequences, we aimed to determine the utility of AI/ML in identifying patients at risk of PCOS,” said Dr. Janet Hall, a co-author of the study and a Senior Researcher in Endocrinology at the National Institute of Environmental Health Sciences (NIEHS), a subsidiary of the NIH. “The effectiveness of artificial intelligence and machine learning in detecting PCOS is even more impressive than we had imagined.”

Polycystic Ovary Syndrome occurs when the ovaries do not function properly and is often accompanied by elevated levels of testosterone. This condition can lead to irregular periods, acne, excessive facial hair growth, or hair loss on the scalp. Women with PCOS also face an increased risk of type 2 diabetes, sleep issues, mental health challenges, cardiovascular problems, and other reproductive system disorders such as uterine cancer and infertility.

“Given the overlap of PCOS with other conditions, diagnosing PCOS can be challenging,” noted Dr. Skand Shekhar, a senior author of the study and an Assistant Research Physician in Endocrinology at the National Institutes of Health. “These data reflect that incorporating AI/ML into electronic health records and other clinical settings has untapped potential to improve the diagnosis and care of women with PCOS.”

The study’s authors suggest combining large-scale, population-based studies with electronic health data sets and analyzing common laboratory tests to identify sensitive diagnostic biomarkers that contribute to PCOS diagnosis.

The diagnosis of PCOS is based on standardized criteria that have evolved over the years and are widely accepted, typically including clinical features such as acne, excessive hair growth, and irregular menstruation, as well as laboratory (e.g., elevated blood testosterone) and radiological findings (e.g., ultrasound revealing multiple small cysts and increased ovarian volume). However, some features of PCOS may coexist with other conditions such as obesity, diabetes, and cardiovascular metabolic disorders, often leading to oversight.

Artificial intelligence refers to the use of computer-based systems or tools to mimic human intelligence and assist in decision-making or prediction. ML is a branch of AI that focuses on learning from past events and applying that knowledge to future decisions. AI can handle vast and diverse data, such as that obtained from electronic health records, making it an ideal auxiliary tool for diagnosing challenging conditions like PCOS.

Researchers conducted a systematic review of all peer-reviewed studies published in the past 25 years (1997-2022) that used AI/ML for detecting PCOS. With the assistance of an experienced librarian at the NIH, they identified potentially eligible studies. They screened a total of 135 studies and included 31 in their analysis. All studies were observational and assessed the application of AI/ML techniques in patient diagnosis. Approximately half of the studies included ultrasound images, and the average age of study participants was 29 years.

In the 10 studies that diagnosed PCOS using standardized diagnostic criteria, the detection accuracy ranged between 80% and 90%.

“In various diagnostic and classification models, AI/ML excelled in detecting PCOS, which is the most significant conclusion from our research,” stated Shekhar.

The authors pointed out that AI/ML-based initiatives have the potential to significantly enhance our ability to detect PCOS in women at an early stage, thereby saving associated costs and alleviating the burden of PCOS on patients and healthcare systems. Subsequent research with robust validation and testing practices will seamlessly integrate AI/ML with chronic health conditions.