# CNS Tumor Classification on PBT Source: [CNS Tumor Classification on PBT]() Collection index: [llms.txt]() Data updated: 2026-05-26T08:02:08.669Z Evaluation of CNS tumor classification performance on the PBT dataset, reporting precision, recall, and F1 scores across specific tumor subtypes including Ependymoma, LG-CNS, HG-CNS, and NGT. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | CNS tumor classification | | Dataset (including any stated split) | PBT | | Metric | Ependymoma F1 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Ependymoma F1](): Higher is better; unit: PERCENT. - [Ependymoma Precision](): Higher is better; unit: PERCENT. - [Ependymoma Recall](): Higher is better; unit: PERCENT. - [LG-CNS Precision](): Higher is better; unit: PERCENT. - [LG-CNS Recall](): Higher is better; unit: PERCENT. - [LG-CNS F1](): Higher is better; unit: PERCENT. - [HG-CNS Precision](): Higher is better; unit: PERCENT. - [HG-CNS Recall](): Higher is better; unit: PERCENT. - [HG-CNS F1](): Higher is better; unit: PERCENT. - [NGT Precision](): Higher is better; unit: PERCENT. - [NGT Recall](): Higher is better; unit: PERCENT. - [NGT F1](): Higher is better; unit: PERCENT. - [Macro Precision](): Higher is better; unit: PERCENT. - [Macro Recall](): Higher is better; unit: PERCENT. - [Macro F1](): Higher is better; unit: PERCENT. ## Recorded results | Method | Ependymoma F1 | Result date | Report paper | | --- | --- | --- | --- | | CLAM | 60 | Not recorded | 69a66c622d603e6de3e3d206 | | EF\_W | 60 | Not recorded | 69a66c622d603e6de3e3d206 | | PM-EF\_WTG | 60 | Not recorded | 69a66c622d603e6de3e3d206 | | Early Fusion | 60 | Not recorded | 6a1528f3dbd7150506f7dd7a | | PathMoE | 60 | Not recorded | 6a1528f3dbd7150506f7dd7a | | ConceptM3oE | 60 | Not recorded | 6a1528f3dbd7150506f7dd7a | | MambaMIL | 55 | Not recorded | 69a66c622d603e6de3e3d206 | | S4MIL | 50 | Not recorded | 69a66c622d603e6de3e3d206 | | PM-EF\_WT | 50 | Not recorded | 69a66c622d603e6de3e3d206 | | PM-SG\_WTG | 50 | Not recorded | 69a66c622d603e6de3e3d206 | | TransMIL | 46.7 | Not recorded | 69a66c622d603e6de3e3d206 | | PM-SG\_WG | 40 | Not recorded | 69a66c622d603e6de3e3d206 | | MoEPP | 40 | Not recorded | 6a1528f3dbd7150506f7dd7a | | PM-EF\_WG | 36.7 | Not recorded | 69a66c622d603e6de3e3d206 | | PM-SG\_WT | 30 | Not recorded | 69a66c622d603e6de3e3d206 | | Switch Gate | 20 | Not recorded | 6a1528f3dbd7150506f7dd7a | | SG\_W | 10 | Not recorded | 69a66c622d603e6de3e3d206 | ### CLAM Ependymoma F1: 60. Result ID: 69a83f5f6dbb7075b94f02c0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Input Modality":"Image (WSI)"}. This record does not assert independent reproduction. ### EF\_W Ependymoma F1: 60. Result ID: 69a83f5f6dbb7075b94f02c4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fusion Strategy":"Early Fusion","Input Modality":"Image (WSI)"}. This record does not assert independent reproduction. ### PM-EF\_WTG Ependymoma F1: 60. Result ID: 69a83f5f6dbb7075b94f02c8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Backbone":"PathMoE","Fusion Strategy":"Early Fusion","Input Modality":"Image (WSI) + Pathology Reports (T) + Nuclei Graphs (G)"}. This record does not assert independent reproduction. ### Early Fusion Ependymoma F1: 60. Result ID: 6a155364fbc373592c0c8e87. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### PathMoE Ependymoma F1: 60. Result ID: 6a155364fbc373592c0c8e8a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ConceptM3oE Ependymoma F1: 60. Result ID: 6a155364fbc373592c0c8e8b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"variant":"flat (morph)"}. This record does not assert independent reproduction. ### MambaMIL Ependymoma F1: 55. Result ID: 69a83f5f6dbb7075b94f02c1. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Input Modality":"Image (WSI)"}. This record does not assert independent reproduction. ### S4MIL Ependymoma F1: 50. Result ID: 69a83f5f6dbb7075b94f02c2. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Input Modality":"Image (WSI)"}. This record does not assert independent reproduction. ### PM-EF\_WT Ependymoma F1: 50. Result ID: 69a83f5f6dbb7075b94f02c6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Backbone":"PathMoE","Fusion Strategy":"Early Fusion","Input Modality":"Image (WSI) + Pathology Reports (T)"}. This record does not assert independent reproduction. ### PM-SG\_WTG Ependymoma F1: 50. Result ID: 69a83f5f6dbb7075b94f02cb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Backbone":"PathMoE","Fusion Strategy":"SwitchGate","Input Modality":"Image (WSI) + Pathology Reports (T) + Nuclei Graphs (G)"}. This record does not assert independent reproduction. ### TransMIL Ependymoma F1: 46.7. Result ID: 69a83f5f6dbb7075b94f02c3. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Input Modality":"Image (WSI)"}. This record does not assert independent reproduction. ### PM-SG\_WG Ependymoma F1: 40. Result ID: 69a83f5f6dbb7075b94f02ca. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Backbone":"PathMoE","Fusion Strategy":"SwitchGate","Input Modality":"Image (WSI) + Nuclei Graphs (G)"}. This record does not assert independent reproduction. ### MoEPP Ependymoma F1: 40. Result ID: 6a155364fbc373592c0c8e88. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### PM-EF\_WG Ependymoma F1: 36.7. Result ID: 69a83f5f6dbb7075b94f02c7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Backbone":"PathMoE","Fusion Strategy":"Early Fusion","Input Modality":"Image (WSI) + Nuclei Graphs (G)"}. This record does not assert independent reproduction. ### PM-SG\_WT Ependymoma F1: 30. Result ID: 69a83f5f6dbb7075b94f02c9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Backbone":"PathMoE","Fusion Strategy":"SwitchGate","Input Modality":"Image (WSI) + Pathology Reports (T)"}. This record does not assert independent reproduction. ### Switch Gate Ependymoma F1: 20. Result ID: 6a155364fbc373592c0c8e89. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### SG\_W Ependymoma F1: 10. Result ID: 69a83f5f6dbb7075b94f02c5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fusion Strategy":"SwitchGate","Input Modality":"Image (WSI)"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.