A general fully automated deep-learning method to detect cardiomegaly in chest x-rays
Carregando...
Citações na Scopus
Tipo de produção
conferenceObject
Data de publicação
2021
Título da Revista
ISSN da Revista
Título do Volume
Editora
SPIE-INT SOC OPTICAL ENGINEERING
Citação
MEDICAL IMAGING 2021: COMPUTER-AIDED DIAGNOSIS, v.11597, article ID 115972B, 6p, 2021
Resumo
Cardiomegaly is a medical condition that leads to an increase in cardiac size. It can be manually assessed using the cardiothoracic ratio from chest x-rays (CXRs). However, as that task can be challenging in such limited examinations, we propose the fully automated cardiomegaly detection in CXR. For this, we first trained convolutional networks (ConvNets) to classify the CXR as positive or negative to cardiomegaly and then evaluated the generalization potential of the trained ConvNets on independent cohorts. This work used frontal CXR images from a public dataset for training/testing and another public and one private dataset to test the models' generalization externally. Training and testing were performed using images cropped with a previously developed U-Net model. Experiments were performed with five topologically different ConvNets, data augmentation techniques, and a 50-50 class-weighing strategy to improve performance and reduce the possibility of bias to the majority class. The receiver operating characteristic curve assessed the performance of the models. DenseNet yielded the highest area under the curve (AUC) on testing (0.818) and external validation (0.809) datasets. Moreover, DenseNet obtained the highest sensitivity overall, yielding up to 0.971 on the private dataset with patients from our hospital. Therefore, DenseNet had a statistically higher potential to identify cardiomegaly. The proposed models, especially those trained with DenseNet convolutional core, automatically detected cardiomegaly with high sensitivity. To the best of our knowledge, this was the first work to design a novel general model for classifying specific deep-learning patterns of cardiomegaly in CXRs.
Palavras-chave
Cardiomegaly, radiography, convolutional neural network, medical imaging analysis, deep learning, radiomics
Referências
- Alghamdi SS, 2020, J RADIAT RES APPL SC, V13, P460, DOI 10.1080/16878507.2020.1756187
- Bustos A, 2020, MED IMAGE ANAL, V66, DOI 10.1016/j.media.2020.101797
- Candemir S., 2018, 2018 IEEE LIF, P109
- Cardenas D., 2020, 20 BRAZ S HLTH APPL 20 BRAZ S HLTH APPL, P179
- Chamveha I., 2020, ARXIV200207468
- Demner-Fushman D, 2016, J AM MED INFORM ASSN, V23, P304, DOI 10.1093/jamia/ocv080
- Felker GM, 2000, NEW ENGL J MED, V342, P1077, DOI 10.1056/NEJM200004133421502
- Ferreira JR, 2020, IEEE ENG MED BIO, P1238, DOI 10.1109/EMBC44109.2020.9176517
- Ferreira JR, 2020, INT J COMPUT ASS RAD, V15, P163, DOI 10.1007/s11548-019-02093-y
- Graves CV, 2020, IEEE ENG MED BIO, P1203, DOI 10.1109/EMBC44109.2020.9175256
- Li ZN, 2019, IEEE ACCESS, V7, P37749, DOI 10.1109/ACCESS.2019.2900053
- Litjens G, 2017, MED IMAGE ANAL, V42, P60, DOI 10.1016/j.media.2017.07.005
- Moreno R., 2019, SPIE MED IMAGING, P445
- Pazhitnykh I., 2017, LUNG SEGMENTATION 2D
- Que Q., P612
- Springenberg J. T., 2015, ICLR