Deep Multimodal MRI Fusion Model for Brain Tumor Grading

뇌 종양 등급 분류를 위한 심층 멀티모달 MRI 통합 모델

  • Published : 2022.05.26

Abstract

Glioma is a type of brain tumor that occurs in glial cells and is classified into two types: high hrade hlioma with a poor prognosis and low grade glioma. Magnetic resonance imaging (MRI) as a non-invasive method is widely used in glioma diagnosis research. Studies to obtain complementary information by combining multiple modalities to overcome the incomplete information limitation of single modality are being conducted. In this study, we developed a 3D CNN-based model that applied input-level fusion to MRI of four modalities (T1, T1Gd, T2, T2-FLAIR). The trained model showed classification performance of 0.8926 accuracy, 0.9688 sensitivity, 0.6400 specificity, and 0.9467 AUC on the validation data. Through this, it was confirmed that the grade of glioma was effectively classified by learning the internal relationship between various modalities.

신경교종(glioma)은 신경교세포에서 발생하는 뇌 종양으로 low grade glioma와 예후가 나쁜 high grade glioma로 분류된다. 자기공명영상(magnetic Resonance Imaging, MRI)은 비침습적 수단으로 이를 이용한 신경교종 진단에 대한 연구가 활발히 진행되고 있다. 또한, 단일 modality의 정보 한계를 극복하기 위해 다중 modality를 조합하여 상호 보완적인 정보를 얻는 연구도 진행되고 있다. 본 논문은 네가지 modality(T1, T1Gd, T2, T2-FLAIR)의 MRI 영상에 입력단 fusion을 적용한 3D CNN 기반의 모델을 제안한다. 학습된 모델은 검증 데이터에 대해 정확도 0.8926, 민감도 0.9688, 특이도 0.6400, AUC 0.9467의 분류 성능을 보였다. 이를 통해 여러 modality 간의 상호관계를 학습하여 신경교종의 등급을 효과적으로 분류함을 확인하였다.

Keywords

Acknowledgement

This study was supported by National Research Foundation (NRF-2020M3E5D2A01084892), Institute for Basic Science (IBS-R015-D1), Ministry of Science and ICT (IITP-2020-2018-0-01798), AI Graduate School Support Program (2019-0-00421), ICT Creative Consilience program (IITP-2020-0-01821), and the Artificial Intelligence Innovation Hub (2021-0-02068).