A novel, one stop, affordable, point of care and artificial intelligence supported system of screening, triage and treatment selection for cervical cancer and precancer in the LMICs - EASTER Application identifier 1161104 Summary statement Artificial intelligence (AI) has penetrated many medical disciplines, AI is fast gaining reputation as offering highly promising solutions for cervical cancer prevention, in particular the AI-supported image capture and prediction of cervical abnormality or Automated Visual Examination (AVE). Despite the availability of effective screening, triage and treatment methods for cervical pre-cancer, cervical cancer elimination is still elusive in low- and middle-income countries (LMICs) where the great majority of disease prevails. Even in the USA there is a widening gap in successful prevention with recognized shortage of skilled providers of colposcopy and precancer management. We propose to develop and evaluate the performance characteristics of a novel AI system to both screen and triage women as well as guide best practice treatment method selection. Our proposal, being implemented in Zimbabwe (a country among the highest cervical cancer burden in the world) and in the USA, advances AI support to analyse infrared spectroscopic signals derived from urine samples of screen-eligible women for the presence of high-risk Human Papillomavirus (hr-HPV). For HPV-positive women the AI will interpret a set of cervical images captured with a high-quality devoted camera to detect high grade cervical precancer and cancer and to determine the type of transformation zone (TZ), which is crucial in determining the optimum treatment choice (excision or ablation). The prototype device for both spectroscopic analysis of urine and image capture along with the AI algorithms have already been developed by us. The technologies have been further improved in phase 1 (initial 2 years of the NIH supported project) and will be validated in phase 2 (subsequent 3 years). During phase 1, we analysed with spectroscopy urine samples collected from 1100 women at screening clinics in Zimbabwe for the presence of hr-HPV and used the signals generated to improve the AI algorithm. In this phase we have also assessed the concordance between spectroscopy and a validated HPV test to detect hr-HPV in urine samples. The cervical image recognition device and the built-in AI algorithm were further improved during phase 1 by collecting more images from 1100 women undergoing hrHPV screening in Zimbabwe. AI is being trained to interpret the cervical images to determine the TZ type. In phase 2, total 2700 women will be screened in Zimbabwe with both our AI-supported spectroscopic analysis of urine and a validated HPV test to evaluate and compare their sensitivity and specificity to detect histology-proved high grade cervical precancers and cancers. The sensitivity and specificity of AI-supported detection of cervical abnormality in captured cervical images will be evaluated to triage the HPV screen positive women in Zimbabwe. The accuracy of AI to determine TZ type will be compared with expert opinion. Additionally, we are recruiting 250 HPV-positive women at two colposcopy clinics in Cincinnati, USA. Images and data collected from the USA population will help us calibrate our model to be used in the USA. During the field validation phase (phase 2), we will also conduct a cost analysis and compare cost of our approach to current standard practice. The International Agency for Research on Cancer has partnered with Neo Sense Vector (NSV), Delaware, USA (industry), The Engineering Department at University of Lancashire, UK, The University of Zimbabwe, College of Health Sciences, Zimbabwe and the University of Cincinnati, USA. Our study focuses on high-tech low-cost innovation that could seriously contribute to the global elimination of cervical cancer. With more than 16,000 deaths from cervical cancer reported every year in Zimbabwe and more than 5000 deaths in the USA, this project has strong relevance for both the countries. As HPV-based screening expands in the USA, the demand for colposcopy services is expected to increas
