Can physiology zap therapeutic sweet spots in hypertension?

M. Joyner

DOI: https://doi.org/10.1161/HYPERTENSIONAHA.112.202754

IF: 9.8968

2012-12-01

Hypertension

Abstract:See related article, pp 1485–1490 Resistant hypertension is a vexing medical problem. It is defined as poorly controlled blood pressure in spite of treatment with optimal doses of ≥3 different classes of drugs.1 In this context, the archetypical patient with resistant hypertension is an obese person in late middle age who might or might not have coexisting diseases, such as diabetes mellitus, peripheral vascular disease, and mildly impaired ventricular function or diastolic dysfunction. A blood pressure value of 150 to 160/95 mm Hg or greater would not be uncommon, and such a value puts these individuals at markedly increased risk for heart attack and stroke. So, the question is what to do about these difficult-to-treat patients? Although many of these patients might respond to truly intensive multimodal lifestyle modification, this has been difficult to implement in a widespread way in real populations.2 Our obesogenic modern world is also a blood pressure raising one. In this context, device therapy, most notably catheter-based renal denervation (also called renal nerve ablation), has moved to fill this therapeutic breach.3 There are 2 basic theories about how renal denervation …

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the **Balanced Number Partitioning Problem**. Specifically, given a set of numbers, the goal is to partition these numbers into two subsets such that the sum of the elements in each subset is as close as possible, and the difference in the number of elements between the two subsets does not exceed 1.

### Problem Background and Importance

The number partitioning problem is a classic problem in theoretical computer science and is one of the six basic NP - complete problems proposed by Garey and Johnson. It not only has important theoretical significance but also has a wide range of applications in practical applications, such as multiprocessor scheduling, VLSI circuit design, etc.

### Main Contributions of the Paper

The main contribution of the paper is to propose a brand - new algorithm, which combines the **Balanced Largest Difference Method (BLDM)** and **Korf's Complete Karmarkar - Karp Algorithm (CKK)**, thus obtaining a new **complete anytime algorithm** that can effectively solve the balanced number partitioning problem.

### Algorithm Characteristics

1. **Initial Solution**: This algorithm first returns the BLDM solution, which is an approximate solution.
2. **Optimization Process**: As the running time increases, the algorithm will continue to look for better solutions until it finally finds and verifies the optimal solution.
3. **Scope of Application**: For numbers with 12 or fewer significant digits, this algorithm can solve balanced partitioning problems of any size in practice. For numbers with higher precision, the algorithm first returns the BLDM solution and then continues to optimize.

### Key Formulas

- Partition Difference:
\[
\Delta(A)=\left|\sum_{i \in A} x_{i}-\sum_{i \notin A} x_{i}\right|
\]
- Card - Degree Difference:
\[
m = |A|-(n - |A|)=2|A|-n
\]

### Experimental Results

The experimental results show that for randomly generated 25 - bit integers, when the number of elements \(n\geq30\), the algorithm can quickly find the perfect partition. For numbers with higher precision (such as 150 - bit), even if the perfect partition does not exist, the algorithm can significantly improve the quality of the initial solution within a few minutes.

### Summary

This paper successfully adapts Korf's CKK algorithm to the constrained number partitioning problem by introducing an effective card - degree tracking mechanism and extending the BLDM heuristic method. This improvement enables the algorithm to efficiently solve the balanced number partitioning problem in practice and find the optimal solution under certain conditions.