Myeloid Malignancy Predictor

Advanced LightGBM-based diagnostic tool

Differentiate between leukemoid reaction and myeloid malignancy in patients with leukocytosis

Machine learning model based on 13 clinical and laboratory parameters. Sensitivity: 100.0%, Specificity: 96.3% (validation set).

Demographics & Biochemistry

?
Patient's age in years
years
?
Lactate dehydrogenase - elevated in tissue damage, cellular turnover, and various malignancies
(94-250)
IU/L

Complete Blood Count

?
Initial white blood cell count on hospital admission
(4-10)
k/μL
?
White blood cell count must be greater than 50 for accurate results
k/μL
?
Hemoglobin - oxygen-carrying protein in red blood cells
(11-16)
g/dL
?
Mean corpuscular volume - average size of red blood cells
(82-98)
fL
?
Cell fragments involved in blood clotting
(150-400)
k/μL

Differential Count

?
Neutrophils - white blood cells that fight bacterial infections
(34-71%)
%
0
100
?
Band neutrophils - immature neutrophils, increase in infections and inflammation
(0-5%)
%
0
30
?
Lymphocytes - white blood cells involved in immune response
(19-53%)
%
0
100
?
Monocytes - white blood cells that fight infections and remove dead cells
(5-13%)
%
0
30
?
Eosinophils - white blood cells that fight parasites and are involved in allergic reactions
(1-7%)
%
0
20
?
Basophils - least common white blood cells, involved in inflammatory reactions
(0-1%)
%
0
5
Prediction Score
-
Diagnosis
-
Model Confidence
-
No prediction generated yet. Submit data in the Patient Data tab first.