Spaces:
Running
OpenMed models for Tabular data predicting survival in lung cancer patients
Hello Everyone,
First of all wonderful job with OpneMed, and good luck ahead!
I am working on a lung cancer dataset (like 400-600 rows) never been studied before (to predict survival in days), I was wondering does OpenMed have any kind of model that could help me with?. (I was going through, could not find any.)
Thank you
Hi
@Rafii
Thank you! I started this to help people who need these models and tools but never get to have them to make a difference in this world.
Can you share like a few rows to see what we are dealing with here? Like is it patient history, diagnosis, what's each row represent. Then, we'll see what we can extract from it or how it can be used.
Sure @MaziyarPanahi
, its a bit long,
So the features are demographics , clinical and Phylogenetic tree,
'age',
'sex',
'ethnicity',
'cigs_perday',
'years_smoking',
'packyears',
'smoking_status_merged',
'is.family.lung',
'ECOG_PS',
'pathologyTNM',
'pT_stage_per_patient',
'pN_stage_per_patient',
'LVI_per_patient',
'PL_per_patient',
'margin_status_per_patient',
'size_pathology_per_patient',
'Surgery_type',
'histology_lesion1',
'histology_lesion1_merged',
'lesion1_sampled',
'histology_lesion2',
'lesion2_sampled',
'histology_multi_full',
'histology_multi_full_genomically.confirmed',
'LUAD_pred_subtype'
'overall_survival' ---- Target Variable (number of days)
values are like,
0 32 68 0 8 20 35 35 0 1 0 1 2 0 1 40 2 3 1 6 2 0 0 0 1 3 0 19 12 41 2 4 0 1849 0 1849 0 1849 0 1849 6 3 1
1 107 81 1 7 44.5 49 109.03 0 0 0 3 4 0 1 75 2 4 1 6 2 7 6 8 1 3 0 19 12 41 2 0 1 1362 1 1362 1 1362 0 1362 6 3 0
2 109 60 1 7 20 38 38 2 0 0 1 2 0 1 36 2 3 1 6 2 0 0 0 0 0 4 17 11 40 0 1 1 2224 1 2224 1 1935 0 2224 5 3 0
I hope ts understandable.
thank you for responding
Thanks @Rafii for sharing these! I actually love to use those as entities, generate datasets and train NER models so it might help you where you can extract what's needed from the text via those models