{"title":"Binary Classification of Pulmonary Nodules using Long Short-Term Memory (LSTM)","authors":"Smridhi Gupta, Arushi Garg, Vidhi Bishnoi, Nidhi Goel","doi":"10.1109/AIST55798.2022.10065049","DOIUrl":null,"url":null,"abstract":"Lung cancer is a prominent reason for deaths all over the globe. A large number of cases have been detected in developed as well as developing nations. It is evident that the probability of survival in the patients is higher only if detected at its nascent stages. Thus, systems employing Computer-Aided Detection (CAD) deliver a faster diagnosis and hence can probably save lives. In the present paper, a classification model for lung nodules that uses Computed Tomography (CT) scans which classifies the given nodule into benign and malignant based on Long Short Term Memory (LSTM) is proposed. The architecture analyzes the images of the nodules extracted from LIDC/ IDRI and Luna-16 datasets. The nodule extraction is executed using the python package pylidc and LSTM is implemented using PyTorch. The highest achieved accuracy using the proposed architecture is 86.98%.","PeriodicalId":360351,"journal":{"name":"2022 4th International Conference on Artificial Intelligence and Speech Technology (AIST)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 4th International Conference on Artificial Intelligence and Speech Technology (AIST)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/AIST55798.2022.10065049","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
Lung cancer is a prominent reason for deaths all over the globe. A large number of cases have been detected in developed as well as developing nations. It is evident that the probability of survival in the patients is higher only if detected at its nascent stages. Thus, systems employing Computer-Aided Detection (CAD) deliver a faster diagnosis and hence can probably save lives. In the present paper, a classification model for lung nodules that uses Computed Tomography (CT) scans which classifies the given nodule into benign and malignant based on Long Short Term Memory (LSTM) is proposed. The architecture analyzes the images of the nodules extracted from LIDC/ IDRI and Luna-16 datasets. The nodule extraction is executed using the python package pylidc and LSTM is implemented using PyTorch. The highest achieved accuracy using the proposed architecture is 86.98%.