Importance of complementary data to histopathological image analysis of oral leukoplakia and carcinoma using deep neural networks  

作　　者：Leandro Muniz de Lima Maria Clara Falcão Ribeiro de Assis Júlia Pessini Soares Tânia Regina Grão-Velloso Liliana Aparecida Pimenta de Barros Danielle Resende Camisasca Renato Antonio Krohling 

机构地区：[1]Nature-inspired Computing Lab,Federal University of Espirito Santo,Vitoria,Brazil [2]Graduate Program in Computer Science,Federal University of Espirito Santo,Vitoria,Brazil [3]School of Dentistry,Departamento de Clinica Odontologica,Federal University of Espirito Santo,Vitoria,Brazil [4]Graduate Program in Clinical Dentistry,Federal University of Espirito Santo,Vitoria,Brazil

出　　处：《Intelligent Medicine》2023年第4期258-266,共9页智慧医学（英文）

基　　金：The research were supported by Grant Nos.304688/2021-5 and21/2022.

摘　　要：Background Oral cancer is one of the most common types of cancer in men causing mortality if not diagnosedearly.In recent years,computer-aided diagnosis(CAD)using artificial intelligence techniques,in particular,deepneural networks have been investigated and several approaches have been proposed to deal with the automateddetection of various pathologies using digital images.Recent studies indicate that the fusion of images with thepatient’s clinical information is important for the final clinical diagnosis.As such dataset does not yet exist fororal cancer,as far as the authors are aware,a new dataset was collected consisting of histopathological images,demographic and clinical data.This study evaluated the importance of complementary data to histopathologicalimage analysis of oral leukoplakia and carcinoma for CAD.Methods A new dataset(NDB-UFES)was collected from 2011 to 2021 consisting of histopathological imagesand information.The 237 samples were curated and analyzed by oral pathologists generating the gold standardfor classification.State-of-the-art image fusion architectures and complementary data(Concatenation,MutualAttention,MetaBlock and MetaNet)using the latest deep learning backbones were investigated for 4 distincttasks to identify oral squamous cell carcinoma,leukoplakia with dysplasia and leukoplakia without dysplasia.We evaluate them using balanced accuracy,precision,recall and area under the ROC curve metrics.Results Experimental results indicate that the best models present balanced accuracy of 83.24%using images,demographic and clinical information with MetaBlock fusion and ResNetV2 backbone.It represents an improvement in performance of 30.68%(19.54 pp)in the task to differentiate samples diagnosed with oral squamous cellcarcinoma and leukoplakia with or without dysplasia.Conclusion This study indicates that cured demographic and clinical data may positively influence the performance of artificial intelligence models in automated classification of oral cancer.

关 键 词：LEUKOPLAKIA Squamous carcinoma PATHOLOGY Deep learning Transfer learning 

分 类 号：R78[医药卫生—口腔医学]