Human-Level Differentiation of Medulloblastoma from Pilocytic Astrocytoma : A Real-World Multicenter Pilot Study
Medulloblastoma and pilocytic astrocytoma are the two most common pediatric brain tumors with overlapping imaging features. In this proof-of-concept study, we investigated using a deep learning classifier trained on a multicenter data set to differentiate these tumor types. We developed a patch-based 3D-DenseNet classifier, utilizing automated tumor segmentation. Given the heterogeneity of imaging data (and available sequences), we used all individually available preoperative imaging sequences to make the model robust to varying input. We compared the classifier to diagnostic assessments by five readers with varying experience in pediatric brain tumors. Overall, we included 195 preoperative MRIs from children with medulloblastoma (n = 69) or pilocytic astrocytoma (n = 126) across six university hospitals. In the 64-patient test set, the DenseNet classifier achieved a high AUC of 0.986, correctly predicting 62/64 (97%) diagnoses. It misclassified one case of each tumor type. Human reader accuracy ranged from 100% (expert neuroradiologist) to 80% (resident). The classifier performed significantly better than relatively inexperienced readers (p < 0.05) and was on par with pediatric neuro-oncology experts. Our proof-of-concept study demonstrates a deep learning model based on automated tumor segmentation that can reliably preoperatively differentiate between medulloblastoma and pilocytic astrocytoma, even in heterogeneous data.
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2024 |
|---|---|
| Erschienen: |
2024 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:16 |
|---|---|
| Enthalten in: |
Cancers - 16(2024), 8 vom: 11. Apr. |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Wiestler, Benedikt [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
Artificial intelligence |
|---|
| Anmerkungen: |
Date Revised 29.04.2024 published: Electronic Citation Status PubMed-not-MEDLINE |
|---|
| doi: |
10.3390/cancers16081474 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM371614007 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM371614007 | ||
| 003 | DE-627 | ||
| 005 | 20240429232534.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 240427s2024 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.3390/cancers16081474 |2 doi | |
| 028 | 5 | 2 | |a pubmed24n1392.xml |
| 035 | |a (DE-627)NLM371614007 | ||
| 035 | |a (NLM)38672556 | ||
| 035 | |a (PII)1474 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Wiestler, Benedikt |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Human-Level Differentiation of Medulloblastoma from Pilocytic Astrocytoma |b A Real-World Multicenter Pilot Study |
| 264 | 1 | |c 2024 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Revised 29.04.2024 | ||
| 500 | |a published: Electronic | ||
| 500 | |a Citation Status PubMed-not-MEDLINE | ||
| 520 | |a Medulloblastoma and pilocytic astrocytoma are the two most common pediatric brain tumors with overlapping imaging features. In this proof-of-concept study, we investigated using a deep learning classifier trained on a multicenter data set to differentiate these tumor types. We developed a patch-based 3D-DenseNet classifier, utilizing automated tumor segmentation. Given the heterogeneity of imaging data (and available sequences), we used all individually available preoperative imaging sequences to make the model robust to varying input. We compared the classifier to diagnostic assessments by five readers with varying experience in pediatric brain tumors. Overall, we included 195 preoperative MRIs from children with medulloblastoma (n = 69) or pilocytic astrocytoma (n = 126) across six university hospitals. In the 64-patient test set, the DenseNet classifier achieved a high AUC of 0.986, correctly predicting 62/64 (97%) diagnoses. It misclassified one case of each tumor type. Human reader accuracy ranged from 100% (expert neuroradiologist) to 80% (resident). The classifier performed significantly better than relatively inexperienced readers (p < 0.05) and was on par with pediatric neuro-oncology experts. Our proof-of-concept study demonstrates a deep learning model based on automated tumor segmentation that can reliably preoperatively differentiate between medulloblastoma and pilocytic astrocytoma, even in heterogeneous data | ||
| 650 | 4 | |a Journal Article | |
| 650 | 4 | |a MRI | |
| 650 | 4 | |a artificial intelligence | |
| 650 | 4 | |a brain | |
| 650 | 4 | |a deep learning | |
| 650 | 4 | |a pediatric brain tumor | |
| 700 | 1 | |a Bison, Brigitte |e verfasserin |4 aut | |
| 700 | 1 | |a Behrens, Lars |e verfasserin |4 aut | |
| 700 | 1 | |a Tüchert, Stefanie |e verfasserin |4 aut | |
| 700 | 1 | |a Metz, Marie |e verfasserin |4 aut | |
| 700 | 1 | |a Griessmair, Michael |e verfasserin |4 aut | |
| 700 | 1 | |a Jakob, Marcus |e verfasserin |4 aut | |
| 700 | 1 | |a Schlegel, Paul-Gerhardt |e verfasserin |4 aut | |
| 700 | 1 | |a Binder, Vera |e verfasserin |4 aut | |
| 700 | 1 | |a von Luettichau, Irene |e verfasserin |4 aut | |
| 700 | 1 | |a Metzler, Markus |e verfasserin |4 aut | |
| 700 | 1 | |a Johann, Pascal |e verfasserin |4 aut | |
| 700 | 1 | |a Hau, Peter |e verfasserin |4 aut | |
| 700 | 1 | |a Frühwald, Michael |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Cancers |d 2009 |g 16(2024), 8 vom: 11. Apr. |w (DE-627)NLM198667213 |x 2072-6694 |7 nnns |
| 773 | 1 | 8 | |g volume:16 |g year:2024 |g number:8 |g day:11 |g month:04 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.3390/cancers16081474 |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 16 |j 2024 |e 8 |b 11 |c 04 | ||