Single nucleotide polymorphism genes and mitochondrial DNA haplogroups as biomarkers for early prediction of knee osteoarthritis structural progressors: use of supervised machine learning classifiers | BMC Medicine
, 2022-09-11 19:44:48,
Participant characteristics
Comparison of the OAI cohort (Table 1) baseline characteristics between the structural progressors and no-progressors showed, for the former, a higher percentage of participants with a Kellgren-Lawrence (KL) score > 0–1, Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores and JSN, and a lower JSW. Age and BMI were also slightly higher in the structural progressor group, but although they reached statistical differences, they were not clinically significant.
For the TASOAC cohort (Table 2), a comparison between the two groups showed that the structural progressors have a higher WOMAC score and JSN, a lower JSW, and fewer men.
OAI and TASOAC cohorts have the same proportion for structural progressors (31%) and no-progressors (69%).
Comparison of the machine learning methodologies
With the OAI cohort, seven methodologies were compared using the 12 independent variables (Eq. 1). Figure 2 indicates the accuracy of the different ML methodologies in PVBSP forecasting at both the training and testing stages.
Comparison of the different machine learning methodologies in PVBSP in the OAI population. a Training (train) and b testing (test) stages accuracy for all the population. DT, decision tree; DT-SA-ELM, decision tree and self-adaptive ELM; ELM, extreme learning machine; KNN, K-nearest neighbor; OAI, Osteoarthritis Initiative; PVBSP, probability values of being structural progressors; RF, random Forest; SA-ELM,…
,
To read the original article from bmcmedicine.biomedcentral.com, Click here
