Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis.
<h4>Introduction</h4>Tuberculous meningitis (TBM) leads to high mortality, especially amongst individuals with HIV. Predicting the incidence of disease-related complications is challenging, for which purpose the value of brain magnetic resonance imaging (MRI) has not been well investigat...
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2025-01-01
|
| Series: | PLoS ONE |
| Online Access: | https://doi.org/10.1371/journal.pone.0321655 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839655811817668608 |
|---|---|
| author | Trinh Huu Khanh Dong Liane S Canas Joseph Donovan Daniel Beasley Nguyen Thuy Thuong-Thuong Nguyen Hoan Phu Nguyen Thi Ha Sebastien Ourselin Reza Razavi Guy E Thwaites Marc Modat |
| author_facet | Trinh Huu Khanh Dong Liane S Canas Joseph Donovan Daniel Beasley Nguyen Thuy Thuong-Thuong Nguyen Hoan Phu Nguyen Thi Ha Sebastien Ourselin Reza Razavi Guy E Thwaites Marc Modat |
| author_sort | Trinh Huu Khanh Dong |
| collection | DOAJ |
| description | <h4>Introduction</h4>Tuberculous meningitis (TBM) leads to high mortality, especially amongst individuals with HIV. Predicting the incidence of disease-related complications is challenging, for which purpose the value of brain magnetic resonance imaging (MRI) has not been well investigated. We used a convolutional neural network (CNN) to explore the complementary contribution of brain MRI to the conventional prognostic determinants.<h4>Methods</h4>We pooled data from two randomised control trials of HIV-positive and HIV-negative adults with clinical TBM in Vietnam to predict the occurrence of death or new neurological complications in the first two months after the subject's first MRI session. We developed and compared three models: a logistic regression with clinical, demographic and laboratory data as reference, a CNN that utilised only T1-weighted MRI volumes, and a model that fused all available information. All models were fine-tuned using two repetitions of 5-fold cross-validation. The final evaluation was based on a random 70/30 training/test split, stratified by the outcome and HIV status. Based on the selected model, we explored the interpretability maps derived from the models.<h4>Results</h4>215 patients were included, with an event prevalence of 22.3%. On the test set our non-imaging model had higher AUC (71.2% [Formula: see text] 1.1%) than the imaging-only model (67.3% [Formula: see text] 2.6%). The fused model was superior to both, with an average AUC = 77.3% [Formula: see text] 4.0% in the test set. The non-imaging variables were more informative in the HIV-positive group, while the imaging features were more predictive in the HIV-negative group. All three models performed better in the HIV-negative cohort. The interpretability maps show the model's focus on the lateral fissures, the corpus callosum, the midbrain, and peri-ventricular tissues.<h4>Conclusion</h4>Imaging information can provide added value to predict unwanted outcomes of TBM. However, to confirm this finding, a larger dataset is needed. |
| format | Article |
| id | doaj-art-d65d7f39c31043b9a44b9c87d6198a21 |
| institution | Matheson Library |
| issn | 1932-6203 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS ONE |
| spelling | doaj-art-d65d7f39c31043b9a44b9c87d6198a212025-06-25T05:31:05ZengPublic Library of Science (PLoS)PLoS ONE1932-62032025-01-01205e032165510.1371/journal.pone.0321655Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis.Trinh Huu Khanh DongLiane S CanasJoseph DonovanDaniel BeasleyNguyen Thuy Thuong-ThuongNguyen Hoan PhuNguyen Thi HaSebastien OurselinReza RazaviGuy E ThwaitesMarc Modat<h4>Introduction</h4>Tuberculous meningitis (TBM) leads to high mortality, especially amongst individuals with HIV. Predicting the incidence of disease-related complications is challenging, for which purpose the value of brain magnetic resonance imaging (MRI) has not been well investigated. We used a convolutional neural network (CNN) to explore the complementary contribution of brain MRI to the conventional prognostic determinants.<h4>Methods</h4>We pooled data from two randomised control trials of HIV-positive and HIV-negative adults with clinical TBM in Vietnam to predict the occurrence of death or new neurological complications in the first two months after the subject's first MRI session. We developed and compared three models: a logistic regression with clinical, demographic and laboratory data as reference, a CNN that utilised only T1-weighted MRI volumes, and a model that fused all available information. All models were fine-tuned using two repetitions of 5-fold cross-validation. The final evaluation was based on a random 70/30 training/test split, stratified by the outcome and HIV status. Based on the selected model, we explored the interpretability maps derived from the models.<h4>Results</h4>215 patients were included, with an event prevalence of 22.3%. On the test set our non-imaging model had higher AUC (71.2% [Formula: see text] 1.1%) than the imaging-only model (67.3% [Formula: see text] 2.6%). The fused model was superior to both, with an average AUC = 77.3% [Formula: see text] 4.0% in the test set. The non-imaging variables were more informative in the HIV-positive group, while the imaging features were more predictive in the HIV-negative group. All three models performed better in the HIV-negative cohort. The interpretability maps show the model's focus on the lateral fissures, the corpus callosum, the midbrain, and peri-ventricular tissues.<h4>Conclusion</h4>Imaging information can provide added value to predict unwanted outcomes of TBM. However, to confirm this finding, a larger dataset is needed.https://doi.org/10.1371/journal.pone.0321655 |
| spellingShingle | Trinh Huu Khanh Dong Liane S Canas Joseph Donovan Daniel Beasley Nguyen Thuy Thuong-Thuong Nguyen Hoan Phu Nguyen Thi Ha Sebastien Ourselin Reza Razavi Guy E Thwaites Marc Modat Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis. PLoS ONE |
| title | Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis. |
| title_full | Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis. |
| title_fullStr | Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis. |
| title_full_unstemmed | Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis. |
| title_short | Convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis. |
| title_sort | convolutional neural network using magnetic resonance brain imaging to predict outcome from tuberculosis meningitis |
| url | https://doi.org/10.1371/journal.pone.0321655 |
| work_keys_str_mv | AT trinhhuukhanhdong convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT lianescanas convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT josephdonovan convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT danielbeasley convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT nguyenthuythuongthuong convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT nguyenhoanphu convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT nguyenthiha convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT sebastienourselin convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT rezarazavi convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT guyethwaites convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis AT marcmodat convolutionalneuralnetworkusingmagneticresonancebrainimagingtopredictoutcomefromtuberculosismeningitis |