Quantum Machine Learning Achieves 91.90% Accuracy in Lung Cancer Prediction With Limited Data

Quantum Computing Tackles Healthcare's Data Scarcity Challenge

Researchers at Florida Atlantic University have demonstrated that quantum machine learning can predict lung cancer with notable accuracy even when training data is severely limited. Led by Arslan Munir, professor in FAU's Department of Electrical Engineering and Computer Science, the team achieved 91.90% accuracy on quantum hardware using D-Wave's annealing quantum computers—a finding that addresses one of healthcare AI's most persistent obstacles.1

1

Source: News-Medical

The study, published in Discover Quantum Science, establishes a common framework for evaluating quantum computing approaches under identical data preparation and testing conditions. This standardized methodology enables meaningful comparisons across different quantum machine learning techniques when datasets are deliberately constrained.2

2

Experimental Design With Constrained Training Data

The research team started with a publicly available dataset containing 309 records that included risk-factor data such as smoking history, coughing, wheezing, fatigue, and chest pain. To balance cancer and non-cancer categories, they applied statistical techniques to generate synthetic examples, expanding the experimental dataset to 540 samples. An optimization-based feature-selection method then narrowed the input variables to 10 key indicators.1

1

Across five evaluation rounds, models were trained on just 108 samples and tested on the remaining 432—a deliberately limited data scenario that mirrors real-world constraints in healthcare settings. The team evaluated three classification methods: quantum support vector machines, quantum k-nearest neighbors, and quantum boosting.2

2

Comparing Gate-Based and Annealing Approaches

The study examined two major quantum computing paradigms: gate-based computing and quantum annealing. Implementations were evaluated across multiple software frameworks including Qiskit, Cirq, PennyLane, OpenJij, and D-Wave, enabling comprehensive comparisons of computational approaches and platform performance.1

1

D-Wave's annealing quantum computers, deployed within a quantum support vector machine framework, delivered the highest accuracy: 94.31% in simulation and 91.90% on actual quantum hardware. Testing on physical quantum hardware proved critical, allowing researchers to assess performance beyond simulated environments and examine real-world viability.2

2

Source: Newswise

Industry Perspective on Quantum-Powered AI in Healthcare

Catherine Potts, manager of technical account management at D-Wave Quantum Inc., noted that "FAU's research highlights how D-Wave's annealing quantum computers can be used to advance AI and machine learning. By applying our technology to a challenging healthcare use case, the researchers demonstrate its potential to extract valuable insights in data-constrained environments, opening new possibilities for quantum-powered AI across healthcare and other data-intensive industries."1

1

Related Stories

Quantum Computing Shows Promise for Kidney Disease Detection Despite Current Limitations

Quantum computers could boost AI by processing large datasets with minimal memory requirements

AI Detects Lung Cancer Precursors Years Before Symptoms Appear, Study Finds

AI Tool Achieves 91% Accuracy Predicting Lung Cancer Immunotherapy Outcomes Across Global Study

Beyond Accuracy: Framework and Methodology Contributions

The study's significance extends beyond its accuracy metrics. By applying a unified evaluation framework, researchers examined how classification method selection, variable choices, and computing platforms affected performance in data-constrained healthcare applications. These findings establish a foundation for continued investigation into quantum machine learning for structured healthcare data.2

2

Stella Batalama, dean of FAU's College of Engineering and Computer Science, emphasized that "this study reflects Florida Atlantic's efforts to connect quantum computing research with meaningful application questions. Careful comparisons and testing on quantum hardware help build the evidence needed to assess where these technologies may be useful and what challenges remain."1

1

Limitations and Future Research Directions

The researchers acknowledge their findings are preliminary. The dataset contains symptoms and risk factors but lacks medical images, pathology findings, laboratory results, or cancer-stage information. The reported accuracy reflects the study's augmented dataset and specific evaluation procedure rather than clinical diagnostic performance.2

2

Future work will examine larger datasets from multiple institutions and explore combining symptom information with imaging or genomic data. The team also plans to investigate model interpretability methods to make predictions more transparent for clinical workflows. "These results provide a starting point for further evaluation," said Arslan Munir. "The next step is to determine how well these approaches generalize to independent clinical datasets and whether they can offer useful support within carefully validated healthcare workflows."1

1