Artificial intelligence in tongue diagnosis: Using deep convolutional neural network for recognizing unhealthy tongue with tooth-mark.

Tongue diagnosis plays a pivotal role in traditional Chinese medicine (TCM) for thousands of years. As one of the most important tongue characteristics, tooth-marked tongue is related to spleen deficiency and can greatly contribute to the symptoms differentiation and treatment selection. Yet, the tooth-marked tongue recognition for TCM practitioners is subjective and challenging. Most of the previous studies have concentrated on subjectively selected features of the tooth-marked region and gained accuracy under 80%. In the present study, we proposed an artificial intelligence framework using deep convolutional neural network (CNN) for the recognition of tooth-marked tongue. First, we constructed relatively large datasets with 1548 tongue images captured by different equipments. Then, we used ResNet34 CNN architecture to extract features and perform classifications. The overall accuracy of the models was over 90%. Interestingly, the models can be successfully generalized to images captured by other devices with different illuminations. The good effectiveness and generalization of our framework may provide objective and convenient computer-aided tongue diagnostic method on tracking disease progression and evaluating pharmacological effect from a informatics perspective.

Lactate dehydrogenase is a prognostic indicator in patients with hepatocellular carcinoma treated by sorafenib: results from the real life practice in HBV endemic area.

PURPOSE: Lactate dehydrogenase (LDH), which was an indirect marker of hypoxia, was a potentially prognostic factor in several malignancies. There is a lack of evidence about the prognostic value of serum LDH level in patients with hepatocellular carcinoma (HCC) receiving sorafenib treatment from hepatitis B virus endemic areas. MATERIALS AND METHODS: A total of 119 HBV-related HCC patients treated by sorafenib from a Chinese center were included into the study. They were categorized into 2 groups according to the cut-off value of pre-treatment LDH, which was determined by the time dependent receiver operating characteristics (ROC) curve for the overall survival. The prognostic value of LDH was evaluated. The relationships between LDH and other clinicopathological factors were also assessed. RESULTS: The cut-off value was 221 U/L. With a median follow up of 15 (range, 3-73) months, 91 patients reached the endpoint. Multivariate analysis proved that pre-treatment serum LDH level was an independent prognostic factor for both overall survival (OS) and progression-free survival (PFS). For patients whose pre-treatment LDH ≥ 221 U/L, increased LDH value after 3 months of sorafenib treatment predicted inferior OS and PFS. And patients with elevated pre-treatment LDH level predisposed to be featured with lower serum albumin, presence of macroscopic vascular invasion, advanced Child-Pugh class, advanced T category, higher AFP, and higher serum total bilirubin. CONCLUSIONS: Serum LDH level was a potentially prognostic factor in HCC patients treated by sorafenib in HBV endemic area. More relevant studies with reasonable study design are needed to further strengthen its prognostic value.