The potential for clinical application of automatic quantification of olfactory bulb volume in MRI scans using convolutional neural networks
The olfactory bulbs (OBs) play a key role in olfactory processing; their volume is important for diagnosis, prognosis and treatment of patients with olfactory loss. Until now, measurements of OB volumes have been limited to quantification of manually segmented OBs, which is a cumbersome task and mak...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-01-01
|
| Series: | NeuroImage: Clinical |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2213158223001006 |
| _version_ | 1797803941387304960 |
|---|---|
| author | Elbrich M. Postma Julia M.H. Noothout Wilbert M. Boek Akshita Joshi Theresa Herrmann Thomas Hummel Paul A.M. Smeets Ivana Išgum Sanne Boesveldt |
| author_facet | Elbrich M. Postma Julia M.H. Noothout Wilbert M. Boek Akshita Joshi Theresa Herrmann Thomas Hummel Paul A.M. Smeets Ivana Išgum Sanne Boesveldt |
| author_sort | Elbrich M. Postma |
| collection | DOAJ |
| description | The olfactory bulbs (OBs) play a key role in olfactory processing; their volume is important for diagnosis, prognosis and treatment of patients with olfactory loss. Until now, measurements of OB volumes have been limited to quantification of manually segmented OBs, which is a cumbersome task and makes evaluation of OB volumes in large scale clinical studies infeasible. Hence, the aim of this study was to evaluate the potential of our previously developed automatic OB segmentation method for application in clinical practice and to relate the results to clinical outcome measures.To evaluate utilization potential of the automatic segmentation method, three data sets containing MR scans of patients with olfactory loss were included. Dataset 1 (N = 66) and 3 (N = 181) were collected at the Smell and Taste Center in Ede (NL) on a 3 T scanner; dataset 2 (N = 42) was collected at the Smell and Taste Clinic in Dresden (DE) on a 1.5 T scanner. To define the reference standard, manual annotation of the OBs was performed in Dataset 1 and 2. OBs were segmented with a method that employs two consecutive convolutional neural networks (CNNs) that the first localize the OBs in an MRI scan and subsequently segment them.In Dataset 1 and 2, the method accurately segmented the OBs, resulting in a Dice coefficient above 0.7 and average symmetrical surface distance below 0.3 mm. Volumes determined from manual and automatic segmentations showed a strong correlation (Dataset 1: r = 0.79, p < 0.001; Dataset 2: r = 0.72, p = 0.004). In addition, the method was able to recognize the absence of an OB. In Dataset 3, OB volumes computed from automatic segmentations obtained with our method were related to clinical outcome measures, i.e. duration and etiology of olfactory loss, and olfactory ability. We found that OB volume was significantly related to age of the patient, duration and etiology of olfactory loss, and olfactory ability (F(5, 172) = 11.348, p < 0.001, R2 = 0.248).In conclusion, the results demonstrate that automatic segmentation of the OBs and subsequent computation of their volumes in MRI scans can be performed accurately and can be applied in clinical and research population studies. Automatic evaluation may lead to more insight in the role of OB volume in diagnosis, prognosis and treatment of olfactory loss. |
| first_indexed | 2024-03-13T05:28:36Z |
| format | Article |
| id | doaj.art-9a5e0201f42d42399535511e63d4a0d7 |
| institution | Directory Open Access Journal |
| issn | 2213-1582 |
| language | English |
| last_indexed | 2024-03-13T05:28:36Z |
| publishDate | 2023-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | NeuroImage: Clinical |
| spelling | doaj.art-9a5e0201f42d42399535511e63d4a0d72023-06-15T04:55:54ZengElsevierNeuroImage: Clinical2213-15822023-01-0138103411The potential for clinical application of automatic quantification of olfactory bulb volume in MRI scans using convolutional neural networksElbrich M. Postma0Julia M.H. Noothout1Wilbert M. Boek2Akshita Joshi3Theresa Herrmann4Thomas Hummel5Paul A.M. Smeets6Ivana Išgum7Sanne Boesveldt8Division of Human Nutrition and Health, Wageningen University & Research, Wageningen, The Netherlands; Department of Otorhinolaryngology, Hospital Gelderse Vallei, Ede, The Netherlands; Corresponding author at: Division of Human Nutrition and Health, Wageningen University & Research, PO Box 17, 6700 AA Wageningen, The Netherlands.Image Sciences Institute, University Medical Center Utrecht, Utrecht, The Netherlands; Department of Biomedical Engineering and Physics, Amsterdam UMC – location AMC, University of Amsterdam, Amsterdam, The NetherlandsDepartment of Otorhinolaryngology, Hospital Gelderse Vallei, Ede, The NetherlandsSmell and Taste Clinic, Department of Otorhinolaryngology, TU Dresden, Dresden, GermanySmell and Taste Clinic, Department of Otorhinolaryngology, TU Dresden, Dresden, GermanySmell and Taste Clinic, Department of Otorhinolaryngology, TU Dresden, Dresden, GermanyDivision of Human Nutrition and Health, Wageningen University & Research, Wageningen, The NetherlandsImage Sciences Institute, University Medical Center Utrecht, Utrecht, The Netherlands; Department of Biomedical Engineering and Physics, Amsterdam UMC – location AMC, University of Amsterdam, Amsterdam, The Netherlands; Department of Radiology and Nuclear Medicine, Amsterdam UMC – location AMC, University of Amsterdam, Amsterdam, The NetherlandsDivision of Human Nutrition and Health, Wageningen University & Research, Wageningen, The NetherlandsThe olfactory bulbs (OBs) play a key role in olfactory processing; their volume is important for diagnosis, prognosis and treatment of patients with olfactory loss. Until now, measurements of OB volumes have been limited to quantification of manually segmented OBs, which is a cumbersome task and makes evaluation of OB volumes in large scale clinical studies infeasible. Hence, the aim of this study was to evaluate the potential of our previously developed automatic OB segmentation method for application in clinical practice and to relate the results to clinical outcome measures.To evaluate utilization potential of the automatic segmentation method, three data sets containing MR scans of patients with olfactory loss were included. Dataset 1 (N = 66) and 3 (N = 181) were collected at the Smell and Taste Center in Ede (NL) on a 3 T scanner; dataset 2 (N = 42) was collected at the Smell and Taste Clinic in Dresden (DE) on a 1.5 T scanner. To define the reference standard, manual annotation of the OBs was performed in Dataset 1 and 2. OBs were segmented with a method that employs two consecutive convolutional neural networks (CNNs) that the first localize the OBs in an MRI scan and subsequently segment them.In Dataset 1 and 2, the method accurately segmented the OBs, resulting in a Dice coefficient above 0.7 and average symmetrical surface distance below 0.3 mm. Volumes determined from manual and automatic segmentations showed a strong correlation (Dataset 1: r = 0.79, p < 0.001; Dataset 2: r = 0.72, p = 0.004). In addition, the method was able to recognize the absence of an OB. In Dataset 3, OB volumes computed from automatic segmentations obtained with our method were related to clinical outcome measures, i.e. duration and etiology of olfactory loss, and olfactory ability. We found that OB volume was significantly related to age of the patient, duration and etiology of olfactory loss, and olfactory ability (F(5, 172) = 11.348, p < 0.001, R2 = 0.248).In conclusion, the results demonstrate that automatic segmentation of the OBs and subsequent computation of their volumes in MRI scans can be performed accurately and can be applied in clinical and research population studies. Automatic evaluation may lead to more insight in the role of OB volume in diagnosis, prognosis and treatment of olfactory loss.http://www.sciencedirect.com/science/article/pii/S2213158223001006Olfactory lossDeep learningConvolutional neural networksSegmentationOlfactory bulb volume |
| spellingShingle | Elbrich M. Postma Julia M.H. Noothout Wilbert M. Boek Akshita Joshi Theresa Herrmann Thomas Hummel Paul A.M. Smeets Ivana Išgum Sanne Boesveldt The potential for clinical application of automatic quantification of olfactory bulb volume in MRI scans using convolutional neural networks NeuroImage: Clinical Olfactory loss Deep learning Convolutional neural networks Segmentation Olfactory bulb volume |
| title | The potential for clinical application of automatic quantification of olfactory bulb volume in MRI scans using convolutional neural networks |
| title_full | The potential for clinical application of automatic quantification of olfactory bulb volume in MRI scans using convolutional neural networks |
| title_fullStr | The potential for clinical application of automatic quantification of olfactory bulb volume in MRI scans using convolutional neural networks |
| title_full_unstemmed | The potential for clinical application of automatic quantification of olfactory bulb volume in MRI scans using convolutional neural networks |
| title_short | The potential for clinical application of automatic quantification of olfactory bulb volume in MRI scans using convolutional neural networks |
| title_sort | potential for clinical application of automatic quantification of olfactory bulb volume in mri scans using convolutional neural networks |
| topic | Olfactory loss Deep learning Convolutional neural networks Segmentation Olfactory bulb volume |
| url | http://www.sciencedirect.com/science/article/pii/S2213158223001006 |
| work_keys_str_mv | AT elbrichmpostma thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT juliamhnoothout thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT wilbertmboek thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT akshitajoshi thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT theresaherrmann thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT thomashummel thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT paulamsmeets thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT ivanaisgum thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT sanneboesveldt thepotentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT elbrichmpostma potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT juliamhnoothout potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT wilbertmboek potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT akshitajoshi potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT theresaherrmann potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT thomashummel potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT paulamsmeets potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT ivanaisgum potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks AT sanneboesveldt potentialforclinicalapplicationofautomaticquantificationofolfactorybulbvolumeinmriscansusingconvolutionalneuralnetworks |