An Empirical Study of Oral Cancer Detection Through Histopathological Images

Cancer arising in any mouth area, such as the lips, gums, tongue, inner lining of the cheeks, roof, or floor of the mouth, is called Oral Cancer (OC). Early detection of OC improves treatment outcomes considerably. Early detection lowers the chance of complications and, in the end, saves lives by stopping the cancer from spreading to other parts of the body. Treatment for early-stage cancer is typically less expensive than that for advanced-stage cancer. Image categorization can help detect OC by analyzing images of the mouth cavity, such as those from photographs or medical imaging scans. Algorithms for classifying images can be trained to identify patterns, including alterations in tissue texture, color, or form, that are linked to early-stage OC. Image classification supports healthcare providers by offering objective insights regarding anomalies that might not be immediately evident to the human eye. In this work, we have presented an empirical study on the prediction of OC through histopathological images. First, we discussed the detailed study on OC. We highlighted the various works on OC through Machine Learning (ML) based approaches. Then, we gave the overview of multiple datasets for detecting oral cancer through various ML-based approaches. Then, we performed the experimental analysis via various ML-based algorithms on two publicly available datasets.