Histopathological Diagnostic Support Technology Using Higher-Order Local Autocorrelation Features

This paper proposes a technology for histopathological diagnostic support that utilizes the correlation-based features of histopathological tissues. In histopathological diagnosis, a clinical pathologist conducts a diagnosis of normal tissues and cancerous tissues. However, recently, the shortage of clinical pathologists is posing increasing burdens to meet the demands for such diagnoses, and this is causing serious social problems. In order to overcome this problem, we propose a technology of histopathological diagnostic support that uses higher-order local autocorrelation (HLAC) features. The proposed method can automatically screen tissue that is believed to be normal tissue to detect cancerous tissue as well as tissue that is suspected of being cancerous to detect abnormalities. Consequently, we can reduce the burden on clinical pathologists, allowing them to concentrate on diagnosing cancer.