Diagnostic testing is experiencing a significant transformation, particularly in the semiconductor and medical sectors. A new research initiative from the Texas McCombs School of Business proposes a testing methodology that could redefine efficiency in both fields. With the global market for semiconductor testing projected to reach $39 billion by 2025 and the medical laboratory testing market estimated at $125 billion, the stakes are high.
Dr. Rohan Ghuge, an assistant professor of decision science at Texas McCombs, emphasizes commonalities between testing chips and medical diagnostics. Both processes involve intricate systems with numerous components. Current testing protocols often rely on sequential approaches, which can be time-consuming and costly. Ghuge explains that this can lead to delays in critical medical decision-making, stating, “First, you might check the vital signs. Then, you come back the next day and do an ECG, then we do blood work, step by step. That’s going to take a lot of time, which we don’t really want to waste for a patient.”
Ghuge’s research, titled “Nonadaptive Stochastic Score Classification and Explainable Half-Space Evaluation,” published in the journal Operations Research, introduces an innovative solution. The study proposes a method that could enable clinicians and engineers to gather essential information in a single round of testing, potentially revolutionizing how both medical and semiconductor systems are evaluated.
Innovative Testing Methodology
The core of Ghuge’s theory lies in selecting a limited number of tests designed to quickly classify a system’s risk as low, medium, or high. Collaborating with Anupam Gupta from New York University and Viswanath Nagarajan from the University of Michigan, Ghuge developed a protocol that combines two distinct test sets: one aimed at diagnosing operational functionality and the other at identifying failures. This novel approach allows for a comprehensive risk assessment in a streamlined manner.
He describes the process: “You create two lists, say, a success list and a failure list. You combine a fraction of the first list and a fraction of the second list. You want to come up with a single batch of tests that tell you at the same time whether the system is working or failing.”
An existing medical application of this concept is the HEART Score, which evaluates five factors to quickly assess the risk of major cardiac events in patients presenting with chest pain. In simulations, Ghuge’s algorithm demonstrated remarkable efficiency, achieving results over 100 times faster than traditional sequential testing methods, albeit at a cost averaging 22% higher.
While the tests may incur higher costs, Ghuge notes that they could ultimately reduce overall expenses by minimizing the logistical challenges associated with setting up multiple tests sequentially.
Future Applications and Real-World Testing
Looking ahead, Ghuge aims to apply this algorithm to real-world testing scenarios. He envisions its use in environments such as broadband internet networks, including major providers like Google Fiber or Spectrum, facilitating rapid diagnostics of system functionality.
“I come from a more theoretical background that focuses on the right model,” Ghuge explains. “There’s a gap between that and applying it in practice. I’m excited to speak with people, to talk to practitioners and see if these can be applied.”
As the demand for efficient diagnostic testing continues to grow, Ghuge’s research could represent a pivotal step forward, offering a scalable and uniform solution that meets the needs of both the semiconductor and healthcare industries.
