Identification of Pulmonary Hypertension Animal Models Using a New Evolutionary Machine Learning Framework Based on...

Area of Science:

• Biomedical research
• Veterinary medicine
• Computational biology

Background:

• Pulmonary Hypertension (PH) affects 1% of the global population, necessitating effective diagnostic tools.
• Animal models are crucial for understanding PH pathophysiology.
• Accurate identification of PH mouse models is vital for research progression.

Purpose of the Study:

• To develop and validate a Kernel Extreme Learning Machine (KELM) model for predicting pulmonary hypertension (PH) in mouse models.
• To identify key blood indicators for PH detection in mice.
• To enhance the accuracy and efficiency of PH mouse model identification.

Main Methods:

• Utilized an improved Whale Optimization Algorithm (WOA) to optimize a Kernel Extreme Learning Machine (KELM) model.
• Employed a feature selection method to identify critical blood indicators for PH prediction.
• Selected indicators included Haemoglobin (HGB), Hematocrit (HCT), Mean Platelet Volume (MPV), Platelet Distribution Width (PDW), and Platelet-Large Cell Ratio (P-LCR).

Main Results:

• The proposed feature selection method identified essential blood indicators for PH mouse model identification.
• The KELM model achieved 100.0% accuracy in classifying PH mouse models.
• The model demonstrated 100.0% specificity, indicating high reliability in identifying PH models.

Conclusions:

• The developed WOA-KELM model accurately predicts pulmonary hypertension in mouse models.
• Key blood indicators like HGB, HCT, MPV, PDW, and P-LCR are significant for PH identification.
• This computational approach shows significant potential for evaluating and identifying mouse PH models in research settings.