REL
2025-02-21
2025-11-25
2025-11-25
Source data for AI-directed voxel extraction and volume EM identify intrusions as sites of mitochondrial contact
0000-0002-5710-6100
Benjamin
Scott
Padman
The Kids Institute & University of Western Australia, Nedlands, Western Australia, Australia.
Stirling Hwy
Perth
Western Australia
Australia
6009
0000-0003-2150-5545
Michael
Lazarou
Walter and Eliza Hall Institute
Elizabeth Street
Parkville
Victoria
Australia
3000
Padman BS
Lindblom R
Lazarou M
Source data and segmentation generated for "AI-directed voxel extraction and volume EM identify intrusions as sites of mitochondrial contact"
Test ROI from Dataset 1 used for AIVE benchmarking.
/data/DATASET1_TEST_ROI
micrographs - multiframe
TIFF
TIFF
1
121
UNSIGNED BYTE
740
3.2552
370
3.2552
Test dataset (cropped from dataset 1) for AIVE benchmarking. Voxel values are inverted from original BSE signal. Spatial units are in nm. 10nm per slice.
Raw model predictions & AIVE processed data for comparisons between 6 trained models
/data/DATASET1_TEST_ROI/Model_comparisons_-_Membrane_Segmentation
reconstructed volumes
TIFF
TIFF
12
121
UNSIGNED BYTE
740
3.2552
370
3.2552
All raw membrane predictions and AIVE processed data from Figures 1 & 2 of manuscript, showing direct comparison of results generated via different models. The trained models are RF (Random Forest), J48 (the java compatible extension of Ross Quinlan’s C4.5 classifier), MLP (Multi-Layer Perceptron), DT (Decision Table), JRip (java compatible propositional rule-based RIPPER), and PART (Projective Adaptive Resonance Theory neural network). Slice thickness is 10nm.
Raw model predictions & human annotations for comparisons after AIVE processing
/data/DATASET1_TEST_ROI/Human_Vs_Unet_-_Mito_Classification
reconstructed volumes
TIFF
TIFF
9
121
UNSIGNED BYTE
740
3.2552
370
3.2552
Raw annotations and AIVE processed data for mitochondria from Figure 3, using classifications generated by one human on consecutive days, or two U-Nets with 3D anisotropic architecture (different random seeds).
Membrane predictions by a random forest model, which were used to conduct various forms of AIVE, are also provided.
Slice thickness is 10nm.
Overview Dataset 1
/data/Set1_-_AIVE_SOURCE_DATA
micrographs - multiframe
TIFF
TIFF
1
724
UNSIGNED BYTE
2960
6.5104
720
6.5104
Tiff stack for half-scale overview FIB-SEM dataset 1. Slice thickness is 10nm.
data/Set1 - AIVE SOURCE DATA/Dataset1_overview - ORGANELLE CLASS LABELS.tif
Organelle classes for Overview Dataset1. Numerical value of each class are identified as follows:
1=MITOS
2=LDS
3=EARLYENDO
4=GOLGI
5=LATEENDO
6=NUC
7=CYTOSKEL
8=ER
9=PLASMEM
10=VESC
TIFF
data/Set1 - AIVE SOURCE DATA/Dataset1_overview_MembranePredictions.tif
Raw membrane predictions for Overview Dataset1, as generated by a Random Forest Model. Values are scalars between 0 and 1, indicating the probability of membrane being present at that location.
TIFF
Overview Dataset 2
/data/Set2_-_AIVE_SOURCE_DATA
micrographs - multiframe
TIFF
TIFF
1
858
UNSIGNED BYTE
2930
6.5104
806
6.5104
Tiff stack for half-scale overview FIB-SEM dataset 2. Slice thickness is 10nm.
data/Set2 - AIVE SOURCE DATA/Dataset2_overview - ORGANELLE CLASS LABELS.tif
Organelle classes for Overview Dataset2. Numerical value of each class are identified as follows:
1=NUC
2=MITOS
3=LDS
4=PLASMEM
5=CHROMATIN
6=ER
7=CYTOSKEL
8=LATEENDO
9=EARLYENDO
10=VESC
TIFF
data/Set2 - AIVE SOURCE DATA/Dataset2_overview_MembranePredictions.tif
Raw membrane predictions for Overview Dataset2, as generated by a Random Forest Model. Values are scalars between 0 and 1, indicating the probability of membrane being present at that location.
TIFF
Overview Dataset 3
/data/Set3_-_AIVE_SOURCE_DATA
micrographs - multiframe
TIFF
TIFF
1
618
UNSIGNED BYTE
2948
6.5104
830
6.5104
Tiff stack for half-scale overview FIB-SEM dataset 3. Slice thickness is 10nm.
data/Set3 - AIVE SOURCE DATA/Dataset3_overview - ORGANELLE CLASS LABELS.tif
Organelle classes for Overview Dataset3. Numerical value of each class are identified as follows:
1=MITOS
2=LDS
3=LATEENDOS
4=EARLYENDOS
5=NUC
6=CYTOSKEL
7=GOLGI
8=ER
9=PLASMEM
10=VESC
TIFF
data/Set3 - AIVE SOURCE DATA/Dataset3_overview_MembranePredictions.tif
Raw membrane predictions for Overview Dataset3, as generated by a Random Forest Model. Values are scalars between 0 and 1, indicating the probability of membrane being present at that location.
TIFF
Overview Dataset 4 - Muscle
/data/Set4_-_AIVE_OUTPUTS
micrographs - multiframe
TIFF
TIFF
1
311
UNSIGNED BYTE
500
16.0
500
16.0
Tiff stack for muscle tissue. Slice thickness is 20nm.
data/Set4 - AIVE SOURCE DATA/Dataset4_MitoClasss.tif
Mitochondrial class for muscle tissue overview.
TIFF
data/Dataset4_overview_MembranePredictions.tif
Raw membrane predictions for Overview Dataset3, as generated by a Random Forest Model. Values are scalars between 0 and 1, indicating the probability of membrane being present at that location.
TIFF