Modified MobileNet for Patient Survival Prediction

Akbar, Agus Subhan; Fatichah, Chastine; Suciati, Nanik

doi:10.1007/978-3-030-72087-2_33

Cited by 9 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 2 shows the comparisons of the SpearmanR and Accuracy performances with other published methods on the BraTS 2020 validation dataset, while Table 2 presents a comparison of the SpearmanR performances with other published methods on the BraTS 2019 and BraTS 2018 validation datasets. From the experimental results, the FLAIR modality showed the highest correlation with the survival prediction output with a correlation SpearmanR of Method SpearmanR Accuracy [24] 0.123 0.345 [26] 0.134 0.483 [27] 0.134 0.517 [28] 0.217 0.517 [29] 0.228 0.414 [30] 0.249 0.517 [25] 0.253 0.414 [31] 0.280 0.450 Ours 0.459 0.517 5 A and Figure 5 B present comparisons of SpearmanR and Accuracy performances between using our FLAIR modality and other modalities, combination approaches on the BraTS 2020 validation dataset. In the segmentation phase, our segmented results achieved a dice score of 0.89845 in the whole tumor, 0.77734 in the tumor core, and 0.78957 in the enhancing tumor.…”

Section: B Dataset and Implementation Detailsmentioning

confidence: 99%

“…A method for forecasting patient survival, which combines MobileNet with a linear survival prediction model (SPM), is outlined in [24]. Different versions of MobileNet are assessed to identify the most effective one, including adapting Mo-bileNet V1 with either frozen or unfrozen layers and modifying MobileNet V2 with either frozen or unfrozen layers connected to SPM.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Prediction of Survival of Glioblastoma Patients Using Local Spatial Relationships and Global Structure Awareness in FLAIR MRI Brain Images

et al. 2023

View full text Add to dashboard Cite

This article introduces a framework for predicting the survival of brain tumor patients by analyzing magnetic resonance images. The prediction of brain tumor survival is challenging due to the limited size of available datasets. To overcome the issue of overfitting, we propose a self-supervised learning method that involves identifying image patches from the same or different images. By recognizing intra-and inter-image differences, the network can learn the relationships between local spatial windows in the same image and across different images. In addition to analyzing local information, we also incorporate a global structure awareness network to capture global information from the entire image. Our proposed method shows a strong correlation between local spatial relationships and survivor class prediction in FLAIR MRI brain images. We evaluate our method using the BraTS 2020 validation dataset and observe that our method outperforms others in accuracy and SpearmanR correlation metrics.

show abstract

Section: B Dataset and Implementation Detailsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Prediction of Survival of Glioblastoma Patients Using Local Spatial Relationships and Global Structure Awareness in FLAIR MRI Brain Images

et al. 2023

View full text Add to dashboard Cite

show abstract

“…Multi-sequence MRIs can comprehensively express the information of tumor lesions by imaging targets with different parameters, in which the effective information should be complementary. Akbar et al (2020) trained a neural network to predict the overall survival for a cohort of 95 GBM patients, based on the imaging features extracted from T1ce, FLAIR and T2 sequence of MRIs, and the mean square error (MSE) was 78 374.17. Lao et al (2017) constructed a survival prediction model using combined available clinical features, 1403 radiomic features and 98 304 deep features based on 4 sequences of MRIs for 75 GBM patients, and the C-index on the independent verification set was 0.71.…”

Section: Introductionmentioning

confidence: 99%

A subregion-based survival prediction framework for GBM via multi-sequence MRI space optimization and clustering-based feature bundling and construction

et al. 2023

View full text Add to dashboard Cite

Aiming at accurate survival prediction of Glioblastoma (GBM) patients following radiation therapy, we developed a subregion-based survival prediction framework via a novel feature construction method on multi-sequence MRIs. The proposed method consists of two main steps: 1) a feature space optimization algorithm to determine the most appropriate matching relation derived between multi-sequence MRIs and tumor subregions, for using multimodal image data more reasonable; 2) a clustering-based feature bundling and construction algorithm to compress the high-dimensional extracted radiomic features and construct a smaller but effective set of features, for accurate prediction model construction. For each tumor subregion, a total of 680 radiomic features were extracted from one MRI sequence using Pyradiomics. Additional 71 geometric features and clinical information were collected resulting in an extreme high-dimensional feature space of 8,231 to train and evaluate the survival prediction at 1-year, and the more challenging overall survival prediction. The framework was developed based on 98 GBM patients from the BraTS 2020 dataset under 5-fold cross-validation, and tested on an external cohort of 19 GBM patients randomly selected from the same dataset. Finally, we identified the best matching relationship between each subregion and its corresponding MRI sequence, a subset of 235 features (out of 8,231 features) were generated by the proposed feature bundling and construction framework. The subregion-based survival prediction framework achieved AUCs of 0.998 and 0.983 on the training and independent test cohort respectively for 1-year survival prediction, compared to AUCs of 0.940 and 0.923 for survival prediction using the 8,231 initial extracted features for training and validation cohorts respectively. Finally, we further constructed an effective stacking structure ensemble regressor to predict the overall survival with the C-index of 0.872. The proposed subregion-based survival prediction framework allow us to better stratified patients towards personalized treatment of GBM.

show abstract

“…Although deep learning methods have achieved state-of-the-art results in numerous applications in clinical and translational imaging, their efficacy in the OS classification task of brain gliomas is yet to be established. Numerous studies have shown that, in comparison to classification models trained with handcrafted (radiomic) features, deep models reported poor predictive performance on BraTS validation and challenge cohorts ( Suter et al, 2018 ; Guo et al, 2019 ; Starke et al, 2019 ; Akbar et al, 2020 ). For instance, Akbar et al (2020) extracted deep features from 2D multi-parametric MRI scans by employing the modified versions of MobileNet V1 ( Howard et al, 2017 ) and MobileNet V2 ( Sandler et al, 2018 ) architectures.…”

Section: Introductionmentioning

confidence: 99%

“…Numerous studies have shown that, in comparison to classification models trained with handcrafted (radiomic) features, deep models reported poor predictive performance on BraTS validation and challenge cohorts ( Suter et al, 2018 ; Guo et al, 2019 ; Starke et al, 2019 ; Akbar et al, 2020 ). For instance, Akbar et al (2020) extracted deep features from 2D multi-parametric MRI scans by employing the modified versions of MobileNet V1 ( Howard et al, 2017 ) and MobileNet V2 ( Sandler et al, 2018 ) architectures. Deep features, augmented with a clinical feature (Age in years), were subsequently fed to a deep learning prediction module called the survival prediction model (SPM).…”

Section: Introductionmentioning

confidence: 99%

Overall Survival Prediction of Glioma Patients With Multiregional Radiomics

et al. 2022

View full text Add to dashboard Cite

Radiomics-guided prediction of overall survival (OS) in brain gliomas is seen as a significant problem in Neuro-oncology. The ultimate goal is to develop a robust MRI-based approach (i.e., a radiomics model) that can accurately classify a novel subject as a short-term survivor, a medium-term survivor, or a long-term survivor. The BraTS 2020 challenge provides radiological imaging and clinical data (178 subjects) to develop and validate radiomics-based methods for OS classification in brain gliomas. In this study, we empirically evaluated the efficacy of four multiregional radiomic models, for OS classification, and quantified the robustness of predictions to variations in automatic segmentation of brain tumor volume. More specifically, we evaluated four radiomic models, namely, the Whole Tumor (WT) radiomics model, the 3-subregions radiomics model, the 6-subregions radiomics model, and the 21-subregions radiomics model. The 3-subregions radiomics model is based on a physiological segmentation of whole tumor volume (WT) into three non-overlapping subregions. The 6-subregions and 21-subregions radiomic models are based on an anatomical segmentation of the brain tumor into 6 and 21 anatomical regions, respectively. Moreover, we employed six segmentation schemes – five CNNs and one STAPLE-fusion method – to quantify the robustness of radiomic models. Our experiments revealed that the 3-subregions radiomics model had the best predictive performance (mean AUC = 0.73) but poor robustness (RSD = 1.99) and the 6-subregions and 21-subregions radiomics models were more robust (RSD 1.39) with lower predictive performance (mean AUC 0.71). The poor robustness of the 3-subregions radiomics model was associated with highly variable and inferior segmentation of tumor core and active tumor subregions as quantified by the Hausdorff distance metric (4.4−6.5mm) across six segmentation schemes. Failure analysis revealed that the WT radiomics model, the 6-subregions radiomics model, and the 21-subregions radiomics model failed for the same subjects which is attributed to the common requirement of accurate segmentation of the WT volume. Moreover, short-term survivors were largely misclassified by the radiomic models and had large segmentation errors (average Hausdorff distance of 7.09mm). Lastly, we concluded that while STAPLE-fusion can reduce segmentation errors, it is not a solution to learning accurate and robust radiomic models.

show abstract

Modified MobileNet for Patient Survival Prediction

Cited by 9 publications

References 14 publications

Prediction of Survival of Glioblastoma Patients Using Local Spatial Relationships and Global Structure Awareness in FLAIR MRI Brain Images

Prediction of Survival of Glioblastoma Patients Using Local Spatial Relationships and Global Structure Awareness in FLAIR MRI Brain Images

A subregion-based survival prediction framework for GBM via multi-sequence MRI space optimization and clustering-based feature bundling and construction

Overall Survival Prediction of Glioma Patients With Multiregional Radiomics

Contact Info

Product

Resources

About