Michael Buchfelder
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Michael Buchfelder has a total of 337 co-authors. Here are the top co-authors by number of shared publications:
- Andreas Stadlbauer
Karl Landsteiner University of Health Sciences, 3500 Krems, Austria
10 shared publications - Thomas Mehari Kinfe
Department of Neurosurgery, Mannheim Center for Neuromodulation and Neuroprosthetics (MCNN), Medical Faculty Mannheim, Ruprecht-Karls-University Heidelberg, Mannheim, Germany.
9 shared publications - Andrea Kleindienst
Department of Neurosurgery, Friedrich-Alexander University, 91054 Erlangen, Germany
7 shared publications - Stefan Rampp
Department of Neuroradiology Friedrich Alexander University Erlangen‐Nürnberg, partner of European Reference Network EpiCARE Erlangen Germany
5 shared publications - Arnd Doerfler
Institute of Neuroradiology, University Hospital Erlangen, Friedrich-Alexander-University Erlangen-Nürnberg, 91054 Erlangen, Germany
4 shared publications - Franz Marhold
Karl Landsteiner University of Health Sciences, 3500 Krems, Austria
4 shared publications
Latest Publications
- Top Cited PapersOpen Access
- 29 April 2026
- Published by Springer Nature in Pituitary
- Vol. 29 (3) , 76
Consensus on acromegaly complications: an update
The 16th Acromegaly Consensus Conference in September 2024 updated recommendations on diagnosis and treatment of acromegaly comorbidities. Since the 2020 acromegaly comorbidity management guideline was published, new evidence has emerged on novel and known comorbidities and new treatment approaches. Forty-three experts in the management of acromegaly reviewed the current literature and assessed changes in clinical practice standards and management. Current outcome goals were considered and updated, with a focus on the impact of current and emerging treatments of these comorbidities. Participants assessed factors that determine pharmacological choices, as well as use of specific agents in the management of the most relevant acromegaly comorbidities. We present consensus recommendations highlighting optimization of evidence-based acromegaly comorbidities management.
- 29 April 2026
- Published by Springer Nature in Pituitary
- Vol. 29 (3) , 76
Top Cited PapersOpen AccessConsensus on acromegaly complications: an update
The 16th Acromegaly Consensus Conference in September 2024 updated recommendations on diagnosis and treatment of acromegaly comorbidities. Since the 2020 acromegaly comorbidity management guideline was published, new evidence has emerged on novel and known comorbidities and new treatment approaches. Forty-three experts in the management of acromegaly reviewed the current literature and assessed changes in clinical practice standards and management. Current outcome goals were considered and updated, with a focus on the impact of current and emerging treatments of these comorbidities. Participants assessed factors that determine pharmacological choices, as well as use of specific agents in the management of the most relevant acromegaly comorbidities. We present consensus recommendations highlighting optimization of evidence-based acromegaly comorbidities management.
- Open Access
- 3 June 2024
- Published by The Endocrine Society in The Journal of Clinical Endocrinology & Metabolism
- Vol. 110 (1) , 228-237
Acromegaly Disease Control Maintained After Switching From Injected Somatostatin Receptor Ligands to Oral Paltusotine
Context Paltusotine is a nonpeptide selective somatostatin receptor 2 agonist in development as once-daily oral treatment for acromegaly. Objective To evaluate the efficacy and safety of paltusotine in the treatment of patients with acromegaly previously controlled with injected somatostatin receptor ligands (SRLs). Methods This phase 3, randomized, double-blind, placebo-controlled trial enrolled adults with acromegaly who had insulin-like growth factor I (IGF-I) ≤1.0 times the upper limit of normal (×ULN) while receiving a stable dose of depot octreotide or lanreotide. Patients were switched from injected SRLs and randomized to receive paltusotine or placebo orally for 36 weeks. The primary endpoint was proportion of patients maintaining IGF-I ≤1.0×ULN. Secondary endpoints were change in IGF-I level, change in Acromegaly Symptom Diary (ASD) score, and maintenance of mean 5-sample growth hormone (GH) 999.99; P<.0001). Paltusotine was also superior to placebo for all secondary endpoints: mean (±SE) change in IGF-I of 0.04±0.09×ULN versus 0.83±0.1×ULN (P<.0001); mean (±SE) change in ASD score of –0.6±1.5 versus 4.6±1.6 (P=.02); mean GH maintained at P=.0003). The most common adverse events were acromegaly symptoms and gastrointestinal effects characteristic of SRLs. Conclusion Replacement of injected SRLs by once-daily oral paltusotine was effective in maintaining both biochemical and symptom control in patients with acromegaly and was well tolerated.
- 3 June 2024
- Published by The Endocrine Society in The Journal of Clinical Endocrinology & Metabolism
- Vol. 110 (1) , 228-237
Open AccessAcromegaly Disease Control Maintained After Switching From Injected Somatostatin Receptor Ligands to Oral Paltusotine
Context Paltusotine is a nonpeptide selective somatostatin receptor 2 agonist in development as once-daily oral treatment for acromegaly. Objective To evaluate the efficacy and safety of paltusotine in the treatment of patients with acromegaly previously controlled with injected somatostatin receptor ligands (SRLs). Methods This phase 3, randomized, double-blind, placebo-controlled trial enrolled adults with acromegaly who had insulin-like growth factor I (IGF-I) ≤1.0 times the upper limit of normal (×ULN) while receiving a stable dose of depot octreotide or lanreotide. Patients were switched from injected SRLs and randomized to receive paltusotine or placebo orally for 36 weeks. The primary endpoint was proportion of patients maintaining IGF-I ≤1.0×ULN. Secondary endpoints were change in IGF-I level, change in Acromegaly Symptom Diary (ASD) score, and maintenance of mean 5-sample growth hormone (GH) 999.99; P<.0001). Paltusotine was also superior to placebo for all secondary endpoints: mean (±SE) change in IGF-I of 0.04±0.09×ULN versus 0.83±0.1×ULN (P<.0001); mean (±SE) change in ASD score of –0.6±1.5 versus 4.6±1.6 (P=.02); mean GH maintained at P=.0003). The most common adverse events were acromegaly symptoms and gastrointestinal effects characteristic of SRLs. Conclusion Replacement of injected SRLs by once-daily oral paltusotine was effective in maintaining both biochemical and symptom control in patients with acromegaly and was well tolerated.
- Open Access
- 30 October 2023
- Published by MDPI AG in Diagnostics
- Vol. 13 (21) , 3345
Immunometabolic Profiling of Chronic Subdural Hematoma through Untargeted Mass Spectrometry Analysis: Preliminary Findings of a Novel Approach
Objective: Metabolomics has growing importance in the research of inflammatory processes. Chronic subdural hematoma (cSDH) is considered to be, at least in part, of inflammatory nature, but no metabolic analyses yet exist. Therefore, a mass spectrometry untargeted metabolic analysis was performed on hematoma samples from patients with cSDH. Methods: A prospective analytical cross-sectional study on the efficacy of subperiosteal drains in cSDH was performed. Newly diagnosed patients had the option of granting permission for the collection of a hematoma sample upon its removal. The samples were analyzed using liquid chromatography–mass spectrometry to obtain different types of metabolites from diverse biochemical classes. The statistical analysis included data cleaning, imputation, and log transformation, followed by PCA, PLS-DA, HCA, and ANOVA. The postoperative course of the disease was followed for 3 months. The metabolite concentrations in the hematoma fluid were compared based on whether a recurrence of the disease was recorded within this time frame. Results: Fifty-nine samples from patients who were operated on because of a cSDH were gathered. Among those, 8 samples were eliminated because of missing metabolites, and only 51 samples were analyzed further. Additionally, 39 samples were from patients who showed no recurrence over the course of a 3-month follow-up, and 12 samples were from a group with later recurrence. We recorded a noticeable drop (35%) in the concentration of acylcarnitines in the ”recurrence group“, where 10 of the 22 tested metabolites showed a significant reduction (p < 0.05). Furthermore, a noticeable reduction in different Acyl-CoA-dehydrogenases was detected (VLCAD-deficiency p < 0.05, MCAD-deficiency p = 0.07). No further changes were detected between both populations. Conclusions: The current study presents a new approach to the research of cSDH. The measurements presented us with new data, which, to date, are without any reference values. Therefore, it is difficult to interpret the information, and our conclusions should be considered to be only speculative. The results do, however, point in the direction of impaired fatty acid oxidation for cases with later recurrence. As fatty acid oxidation plays an important role in inflammatory energy metabolism, the results suggest that inflammatory processes could be aggravated in cases with recurrence. Because our findings are neither proven through further analyses nor offer an obvious therapy option, their implications would not change everyday practice in the management of cSDH. They do, however, present a further possibility of research that might, in the future, be relevant to the therapy.
- 30 October 2023
- Published by MDPI AG in Diagnostics
- Vol. 13 (21) , 3345
Open AccessImmunometabolic Profiling of Chronic Subdural Hematoma through Untargeted Mass Spectrometry Analysis: Preliminary Findings of a Novel Approach
Objective: Metabolomics has growing importance in the research of inflammatory processes. Chronic subdural hematoma (cSDH) is considered to be, at least in part, of inflammatory nature, but no metabolic analyses yet exist. Therefore, a mass spectrometry untargeted metabolic analysis was performed on hematoma samples from patients with cSDH. Methods: A prospective analytical cross-sectional study on the efficacy of subperiosteal drains in cSDH was performed. Newly diagnosed patients had the option of granting permission for the collection of a hematoma sample upon its removal. The samples were analyzed using liquid chromatography–mass spectrometry to obtain different types of metabolites from diverse biochemical classes. The statistical analysis included data cleaning, imputation, and log transformation, followed by PCA, PLS-DA, HCA, and ANOVA. The postoperative course of the disease was followed for 3 months. The metabolite concentrations in the hematoma fluid were compared based on whether a recurrence of the disease was recorded within this time frame. Results: Fifty-nine samples from patients who were operated on because of a cSDH were gathered. Among those, 8 samples were eliminated because of missing metabolites, and only 51 samples were analyzed further. Additionally, 39 samples were from patients who showed no recurrence over the course of a 3-month follow-up, and 12 samples were from a group with later recurrence. We recorded a noticeable drop (35%) in the concentration of acylcarnitines in the ”recurrence group“, where 10 of the 22 tested metabolites showed a significant reduction (p < 0.05). Furthermore, a noticeable reduction in different Acyl-CoA-dehydrogenases was detected (VLCAD-deficiency p < 0.05, MCAD-deficiency p = 0.07). No further changes were detected between both populations. Conclusions: The current study presents a new approach to the research of cSDH. The measurements presented us with new data, which, to date, are without any reference values. Therefore, it is difficult to interpret the information, and our conclusions should be considered to be only speculative. The results do, however, point in the direction of impaired fatty acid oxidation for cases with later recurrence. As fatty acid oxidation plays an important role in inflammatory energy metabolism, the results suggest that inflammatory processes could be aggravated in cases with recurrence. Because our findings are neither proven through further analyses nor offer an obvious therapy option, their implications would not change everyday practice in the management of cSDH. They do, however, present a further possibility of research that might, in the future, be relevant to the therapy.
- Open Access
Cortical Activity Associated With Acute Development of Phantom Auditory Percepts via Unilateral Deprivation
Subjective tinnitus describes the experience of hearing phantom sounds (e.g., tones, buzzing, noise). While the majority of those who report having experienced phantom sounds claim that these percepts have not lasted but are transient, some experience this chronically, with others even describing their tinnitus as severe enough to negatively impact their well-being and daily lives. Currently, no permanent solution has been discovered for preventing or curing tinnitus. This is due to several factors, including an insufficient understanding of the mechanisms at play that give rise to such auditory sensations, as well as a lack of research investigating the corresponding changes in neural activity associated with the onset and development of tinnitus. Taking advantage of the high spatial and temporal resolution of magnetoencephalography (MEG), we measured cortical activity associated with the development of acute tinnitus-like percepts induced via unilateral auditory deprivation. Over the course of four days, participants continuously wore a silicone earplug in one ear, which led to the experience of phantom sounds in 15 of 16 participants. Frequency analysis of source-localized continuous MEG data revealed a significant increase of gamma power in primary auditory cortices (A1) during the tinnitus condition (p=0.02), which most likely reflects the neuronal processing correlated with tinnitus perception.
Open AccessCortical Activity Associated With Acute Development of Phantom Auditory Percepts via Unilateral Deprivation
Subjective tinnitus describes the experience of hearing phantom sounds (e.g., tones, buzzing, noise). While the majority of those who report having experienced phantom sounds claim that these percepts have not lasted but are transient, some experience this chronically, with others even describing their tinnitus as severe enough to negatively impact their well-being and daily lives. Currently, no permanent solution has been discovered for preventing or curing tinnitus. This is due to several factors, including an insufficient understanding of the mechanisms at play that give rise to such auditory sensations, as well as a lack of research investigating the corresponding changes in neural activity associated with the onset and development of tinnitus. Taking advantage of the high spatial and temporal resolution of magnetoencephalography (MEG), we measured cortical activity associated with the development of acute tinnitus-like percepts induced via unilateral auditory deprivation. Over the course of four days, participants continuously wore a silicone earplug in one ear, which led to the experience of phantom sounds in 15 of 16 participants. Frequency analysis of source-localized continuous MEG data revealed a significant increase of gamma power in primary auditory cortices (A1) during the tinnitus condition (p=0.02), which most likely reflects the neuronal processing correlated with tinnitus perception.
- Open Access
- 14 March 2023
- Published by MDPI AG in Brain Sciences
- Vol. 13 (3) , 492
Deep Brain Stimulation for Chronic Facial Pain: An Individual Participant Data (IPD) Meta-Analysis
Despite available, advanced pharmacological and behavioral therapies, refractory chronic facial pain of different origins still poses a therapeutic challenge. In circumstances where there is insufficient responsiveness to pharmacological/behavioral therapies, deep brain stimulation should be considered as a potential effective treatment option. We performed an individual participant data (IPD) meta-analysis including searches on PubMed, Embase, and the Cochrane Library (2000–2022). The primary endpoint was the change in pain intensity (visual analogue scale; VAS) at a defined time-point of ≤3 months post-DBS. In addition, correlation and regression analyses were performed to identify predictive markers (age, duration of pain, frequency, amplitude, intensity, contact configuration, and the DBS target). A total of seven trials consisting of 54 screened patients met the inclusion criteria. DBS significantly reduced the pain levels after 3 months without being related to a specific DBS target, age, contact configuration, stimulation intensity, frequency, amplitude, or chronic pain duration. Adverse events were an infection or lead fracture (19%), stimulation-induced side effects (7%), and three deaths (unrelated to DBS—from cancer progression or a second stroke). Although comparable long-term data are lacking, the current published data indicate that DBS (thalamic and PVG/PAG) effectively suppresses facial pain in the short-term. However, the low-quality evidence, reporting bias, and placebo effects must be considered in future randomized-controlled DBS trials for facial pain.
- 14 March 2023
- Published by MDPI AG in Brain Sciences
- Vol. 13 (3) , 492
Open AccessDeep Brain Stimulation for Chronic Facial Pain: An Individual Participant Data (IPD) Meta-Analysis
Despite available, advanced pharmacological and behavioral therapies, refractory chronic facial pain of different origins still poses a therapeutic challenge. In circumstances where there is insufficient responsiveness to pharmacological/behavioral therapies, deep brain stimulation should be considered as a potential effective treatment option. We performed an individual participant data (IPD) meta-analysis including searches on PubMed, Embase, and the Cochrane Library (2000–2022). The primary endpoint was the change in pain intensity (visual analogue scale; VAS) at a defined time-point of ≤3 months post-DBS. In addition, correlation and regression analyses were performed to identify predictive markers (age, duration of pain, frequency, amplitude, intensity, contact configuration, and the DBS target). A total of seven trials consisting of 54 screened patients met the inclusion criteria. DBS significantly reduced the pain levels after 3 months without being related to a specific DBS target, age, contact configuration, stimulation intensity, frequency, amplitude, or chronic pain duration. Adverse events were an infection or lead fracture (19%), stimulation-induced side effects (7%), and three deaths (unrelated to DBS—from cancer progression or a second stroke). Although comparable long-term data are lacking, the current published data indicate that DBS (thalamic and PVG/PAG) effectively suppresses facial pain in the short-term. However, the low-quality evidence, reporting bias, and placebo effects must be considered in future randomized-controlled DBS trials for facial pain.
- Open Access
- 28 February 2023
- Published by MDPI AG in Diagnostics
- Vol. 13 (5) , 911
Prediction of the Topography of the Corticospinal Tract on T1-Weighted MR Images Using Deep-Learning-Based Segmentation
Introduction: Tractography is an invaluable tool in the planning of tumor surgery in the vicinity of functionally eloquent areas of the brain as well as in the research of normal development or of various diseases. The aim of our study was to compare the performance of a deep-learning-based image segmentation for the prediction of the topography of white matter tracts on T1-weighted MR images to the performance of a manual segmentation. Methods: T1-weighted MR images of 190 healthy subjects from 6 different datasets were utilized in this study. Using deterministic diffusion tensor imaging, we first reconstructed the corticospinal tract on both sides. After training a segmentation model on 90 subjects of the PIOP2 dataset using the nnU-Net in a cloud-based environment with graphical processing unit (Google Colab), we evaluated its performance using 100 subjects from 6 different datasets. Results: Our algorithm created a segmentation model that predicted the topography of the corticospinal pathway on T1-weighted images in healthy subjects. The average dice score was 0.5479 (0.3513–0.7184) on the validation dataset. Conclusions: Deep-learning-based segmentation could be applicable in the future to predict the location of white matter pathways in T1-weighted scans.
- 28 February 2023
- Published by MDPI AG in Diagnostics
- Vol. 13 (5) , 911
Open AccessPrediction of the Topography of the Corticospinal Tract on T1-Weighted MR Images Using Deep-Learning-Based Segmentation
Introduction: Tractography is an invaluable tool in the planning of tumor surgery in the vicinity of functionally eloquent areas of the brain as well as in the research of normal development or of various diseases. The aim of our study was to compare the performance of a deep-learning-based image segmentation for the prediction of the topography of white matter tracts on T1-weighted MR images to the performance of a manual segmentation. Methods: T1-weighted MR images of 190 healthy subjects from 6 different datasets were utilized in this study. Using deterministic diffusion tensor imaging, we first reconstructed the corticospinal tract on both sides. After training a segmentation model on 90 subjects of the PIOP2 dataset using the nnU-Net in a cloud-based environment with graphical processing unit (Google Colab), we evaluated its performance using 100 subjects from 6 different datasets. Results: Our algorithm created a segmentation model that predicted the topography of the corticospinal pathway on T1-weighted images in healthy subjects. The average dice score was 0.5479 (0.3513–0.7184) on the validation dataset. Conclusions: Deep-learning-based segmentation could be applicable in the future to predict the location of white matter pathways in T1-weighted scans.
- Open Access
- 2 February 2023
- Published by MDPI AG in Biomedicines
- Vol. 11 (2) , 438
MEG Node Degree for Focus Localization: Comparison with Invasive EEG
Epilepsy surgery is a viable therapy option for patients with pharmacoresistant focal epilepsies. A prerequisite for postoperative seizure freedom is the localization of the epileptogenic zone, e.g., using electro- and magnetoencephalography (EEG/MEG). Evidence shows that resting state MEG contains subtle alterations, which may add information to the workup of epilepsy surgery. Here, we investigate node degree (ND), a graph-theoretical parameter of functional connectivity, in relation to the seizure onset zone (SOZ) determined by invasive EEG (iEEG) in a consecutive series of 50 adult patients. Resting state data were subjected to whole brain, all-to-all connectivity analysis using the imaginary part of coherence. Graphs were described using parcellated ND. SOZ localization was investigated on a lobar and sublobar level. On a lobar level, all frequency bands except alpha showed significantly higher maximal ND (mND) values inside the SOZ compared to outside (ratios 1.11–1.20, alpha 1.02). Area-under-the-curve (AUC) was 0.67–0.78 for all expected alpha (0.44, ns). On a sublobar level, mND inside the SOZ was higher for all frequency bands (1.13–1.38, AUC 0.58–0.78) except gamma (1.02). MEG ND is significantly related to SOZ in delta, theta and beta bands. ND may provide new localization tools for presurgical evaluation of epilepsy surgery.
- 2 February 2023
- Published by MDPI AG in Biomedicines
- Vol. 11 (2) , 438
Open AccessMEG Node Degree for Focus Localization: Comparison with Invasive EEG
Epilepsy surgery is a viable therapy option for patients with pharmacoresistant focal epilepsies. A prerequisite for postoperative seizure freedom is the localization of the epileptogenic zone, e.g., using electro- and magnetoencephalography (EEG/MEG). Evidence shows that resting state MEG contains subtle alterations, which may add information to the workup of epilepsy surgery. Here, we investigate node degree (ND), a graph-theoretical parameter of functional connectivity, in relation to the seizure onset zone (SOZ) determined by invasive EEG (iEEG) in a consecutive series of 50 adult patients. Resting state data were subjected to whole brain, all-to-all connectivity analysis using the imaginary part of coherence. Graphs were described using parcellated ND. SOZ localization was investigated on a lobar and sublobar level. On a lobar level, all frequency bands except alpha showed significantly higher maximal ND (mND) values inside the SOZ compared to outside (ratios 1.11–1.20, alpha 1.02). Area-under-the-curve (AUC) was 0.67–0.78 for all expected alpha (0.44, ns). On a sublobar level, mND inside the SOZ was higher for all frequency bands (1.13–1.38, AUC 0.58–0.78) except gamma (1.02). MEG ND is significantly related to SOZ in delta, theta and beta bands. ND may provide new localization tools for presurgical evaluation of epilepsy surgery.
- Open Access
Improved Detection of Cavernous Sinus Invasion of Pituitary Macroadenomas with Ultra-High-Field 7 T MRI
To compare 7 T magnetic resonance imaging (MRI) of pituitary macroadenomas (PMA) with standard MRI and intraoperative findings regarding tumor detection, localization, size, and extension. Patients with suspected pituitary adenoma underwent pre-operative 1.5 T or 3 T and 7 T MRI; 14 patients with a PMA were included. A qualitative (lesion detection, location, cavernous sinus infiltration) and quantitative (lesion size, depth of cavernous sinus infiltration) analysis of 1.5 T, 3 T and 7 T MRI was performed and compared with intraoperative findings. Both 1.5/3 T and 7 T MRI enabled the detection of all PMAs; lesion size determination was equal. 7 T MRI enables more precise assessments of cavernous sinus infiltration of PMA (ncorrect 7T = 78.6%, ncorrect 1.5/3T = 64.3%). Ultra-high-field MRI is a reliable imaging modality for evaluation of PMAs providing exact information on lesion location and size. 7 T MRI yielded more accurate information on cavernous sinus infiltration with better agreement with intraoperative findings than standard MRI.
Open AccessImproved Detection of Cavernous Sinus Invasion of Pituitary Macroadenomas with Ultra-High-Field 7 T MRI
To compare 7 T magnetic resonance imaging (MRI) of pituitary macroadenomas (PMA) with standard MRI and intraoperative findings regarding tumor detection, localization, size, and extension. Patients with suspected pituitary adenoma underwent pre-operative 1.5 T or 3 T and 7 T MRI; 14 patients with a PMA were included. A qualitative (lesion detection, location, cavernous sinus infiltration) and quantitative (lesion size, depth of cavernous sinus infiltration) analysis of 1.5 T, 3 T and 7 T MRI was performed and compared with intraoperative findings. Both 1.5/3 T and 7 T MRI enabled the detection of all PMAs; lesion size determination was equal. 7 T MRI enables more precise assessments of cavernous sinus infiltration of PMA (ncorrect 7T = 78.6%, ncorrect 1.5/3T = 64.3%). Ultra-high-field MRI is a reliable imaging modality for evaluation of PMAs providing exact information on lesion location and size. 7 T MRI yielded more accurate information on cavernous sinus infiltration with better agreement with intraoperative findings than standard MRI.
- Open Access
- 14 December 2022
- Published by MDPI AG in Metabolites
- Vol. 12 (12) , 1264
Differentiation of Glioblastoma and Brain Metastases by MRI-Based Oxygen Metabolomic Radiomics and Deep Learning
Glioblastoma (GB) and brain metastasis (BM) are the most frequent types of brain tumors in adults. Their therapeutic management is quite different and a quick and reliable initial characterization has a significant impact on clinical outcomes. However, the differentiation of GB and BM remains a major challenge in today’s clinical neurooncology due to their very similar appearance in conventional magnetic resonance imaging (MRI). Novel metabolic neuroimaging has proven useful for improving diagnostic performance but requires artificial intelligence for implementation in clinical routines. Here; we investigated whether the combination of radiomic features from MR-based oxygen metabolism (“oxygen metabolic radiomics”) and deep convolutional neural networks (CNNs) can support reliably pre-therapeutic differentiation of GB and BM in a clinical setting. A self-developed one-dimensional CNN combined with radiomic features from the cerebral metabolic rate of oxygen (CMRO2) was clearly superior to human reading in all parameters for classification performance. The radiomic features for tissue oxygen saturation (mitoPO2; i.e., tissue hypoxia) also showed better diagnostic performance compared to the radiologists. Interestingly, both the mean and median values for quantitative CMRO2 and mitoPO2 values did not differ significantly between GB and BM. This demonstrates that the combination of radiomic features and DL algorithms is more efficient for class differentiation than the comparison of mean or median values. Oxygen metabolic radiomics and deep neural networks provide insights into brain tumor phenotype that may have important diagnostic implications and helpful in clinical routine diagnosis.
- 14 December 2022
- Published by MDPI AG in Metabolites
- Vol. 12 (12) , 1264
Open AccessDifferentiation of Glioblastoma and Brain Metastases by MRI-Based Oxygen Metabolomic Radiomics and Deep Learning
Glioblastoma (GB) and brain metastasis (BM) are the most frequent types of brain tumors in adults. Their therapeutic management is quite different and a quick and reliable initial characterization has a significant impact on clinical outcomes. However, the differentiation of GB and BM remains a major challenge in today’s clinical neurooncology due to their very similar appearance in conventional magnetic resonance imaging (MRI). Novel metabolic neuroimaging has proven useful for improving diagnostic performance but requires artificial intelligence for implementation in clinical routines. Here; we investigated whether the combination of radiomic features from MR-based oxygen metabolism (“oxygen metabolic radiomics”) and deep convolutional neural networks (CNNs) can support reliably pre-therapeutic differentiation of GB and BM in a clinical setting. A self-developed one-dimensional CNN combined with radiomic features from the cerebral metabolic rate of oxygen (CMRO2) was clearly superior to human reading in all parameters for classification performance. The radiomic features for tissue oxygen saturation (mitoPO2; i.e., tissue hypoxia) also showed better diagnostic performance compared to the radiologists. Interestingly, both the mean and median values for quantitative CMRO2 and mitoPO2 values did not differ significantly between GB and BM. This demonstrates that the combination of radiomic features and DL algorithms is more efficient for class differentiation than the comparison of mean or median values. Oxygen metabolic radiomics and deep neural networks provide insights into brain tumor phenotype that may have important diagnostic implications and helpful in clinical routine diagnosis.
- 4 October 2022
- Published by Elsevier in Neuromodulation: Technology at the Neural Interface
- Vol. 25 (7) , S110
O082 / #527 MOLECULAR INFLAMMATORY AND SENSORY PHENOTYPING IN CRPS PATIENTS TREATED WITH UNILATERAL L4-DORSAL ROOT GANGLION STIMULATION
- 4 October 2022
- Published by Elsevier in Neuromodulation: Technology at the Neural Interface
- Vol. 25 (7) , S110
O082 / #527 MOLECULAR INFLAMMATORY AND SENSORY PHENOTYPING IN CRPS PATIENTS TREATED WITH UNILATERAL L4-DORSAL ROOT GANGLION STIMULATION