An international team of scientists promises to revolutionise the diagnosis of cervical cancer with an innovative approach to cervical cytology, better known as the Pap smear.

The study, now published in the journal Nature, with the participation of Fernando Schmitt, director of the RISE-Health Research Unit and professor at the Faculty of Medicine of the University of Porto (FMUP), demonstrated the advantages of a new form of automated analysis of cervical cell samples using artificial intelligence, compared to the traditional method of clinical cytology. The aim is to move earlier towards life-saving treatments.

‘The use of artificial intelligence in the clinic allows us to evaluate cellular characteristics and classify them as normal or abnormal,’ explains Fernando Schmitt (RISE-Health/FMUP), internationally recognised as one of the world’s leading experts in the field of cytopathology.

According to the RISE-Health expert, the automation of this screening will speed up the diagnosis of cervical cancer, a disease caused mainly by infection with Human Papillomavirus (HPV), which is sexually transmitted and accounts for 10% of cancers in women.

Currently, the cells collected are evaluated under a microscope by a professional. However, the process has some disadvantages, such as the subjectivity of interpretation and the variability of results.

In order to address this problem, this new artificial intelligence system applied to traditional cytology is the first to be able to screen abnormal cells completely autonomously, enabling a faster, more accurate and more objective diagnosis.

‘The automation of cytopathology can also detect early lesions, speeding up and improving cancer diagnosis,’ explains the director of the RISE-Health Research Unit.

The new method scans the cells and reconstructs a 3D image in real time, allowing their characteristics to be better seen. The platform then uses advanced algorithms to group similar profiles and identify abnormal cells with greater accuracy, reducing the risk of human error.

This AI-based approach, already tested and validated with thousands of samples from real patients, can now help professionals and pathology laboratories by providing a ‘visual map’ of cell classification, which will be an added value advantage over the conventional method.

The authors of the study share their enthusiasm and desire to implement this new diagnostic method in clinical practice in the near future, with predictable gains for patients, health systems and the quality of life of populations on a global scale.

It is hoped that this technology will be accessible in several countries, becoming an important tool in the approach to cervical cancer, which continues to affect women around the world. Warning signs include abnormal vaginal bleeding, increased vaginal discharge, pelvic pain and pain during sexual intercourse.

Entitled ‘Clinical-grade autonomous cytopathology through whole-slide edge tomography’, this research also involved leading scientists, hospitals and companies from Japan, China and the US.