I have a C# function that converts a DataTable into a list of ExpandoObjects, where each object represents a row with properties dynamically assigned from the column names and values

public static IEnumerable<ExpandoObject> ToExpandoObject(this DataTable dataTable)
{
    List<ExpandoObject> dynamicDt = [];
    var splitColumnNames = dataTable.Columns.Cast<DataColumn>().Select(c => c.ColumnName.Split("_")[0]).ToArray();
    foreach (DataRow row in dataTable.Rows)
    {
        dynamic dyn = new ExpandoObject();
        dynamicDt.Add(dyn);
        var columnIndex = 0;
        foreach (DataColumn column in dataTable.Columns)
        {
            var dic = (IDictionary<string, object>)dyn;
            dic[column.ColumnName] = ConvertColumnValue(row[column], splitColumnNames[columnIndex]);
            columnIndex++;
        }
    }

    return dynamicDt;
}

I want to improve the performance of this function by utilizing parallel processing How can I modify this function to run efficiently in parallel