A smart tele-cytology point-of-care platform for oral cancer screening

• DOI: 10.1371/journal.pone.0224885
• Publication Date: 2019-11-15T13:29:06Z
• Authors: Sumsum Sunny; Arun Baby; Bonney Lee James; Dev Balaji; Aparna N. V.; Maitreya H. Rana; Praveen Gurpur; Arunan Skandarajah; Michael D’Ambrosio; Ravindra Doddathi...; Sunil Paramel Mohan; Nisheena Raghavan; Uma Kandasarma; Sangeetha N.; Subhasini Raghavan; Naveen Hedne; Felix Koch; Daniel A. Fletcher; Sumithra Selvam; Manohar Kollegal; Praveen Birur N.; Lance Ladic; Amritha Suresh; Hardik J. Pandya; Moni Abraham Kuri...
• Keywords: Point of care

Early detection of oral cancer necessitates a minimally invasive, tissue-specific diagnostic tool that facilitates screening/surveillance. Brush biopsy, though demands skilled cyto-pathologist expertise. In this study, we explored the clinical utility/efficacy tele-cytology system in combination with Artificial Neural Network (ANN) based risk-stratification model for early potentially malignant (OPML)/malignant lesion. A portable, automated tablet-based platform capable digitization cytology slides was evaluated its efficacy OPML/malignant lesions (n = 82) comparison conventional and histology. Then, an image pre-processing algorithm established to segregate cells, ANN trained images 11,981) developed. The specificity, sensitivity accuracy platform/ stratification were computed, agreement examined using Kappa statistics. platform, Cellscope, showed overall 84–86% no difference between (kappa, 0.67–0.72). However, OPML could be detected low (18%) accordance limitations cytology. integration processing development ANN-based risk improved (93%) high grade (73%), thereby increasing by 30%. Tele-cytology integrated model, novel strategy can invaluable Point-of-Care (PoC) detection/screening cancer. This study hence establishes applicability accurate, remote diagnosis use analysis improving efficacy.