Development and Validation of a Low-Dose CT Protocol for Chest Imaging in a Clinical Radiology Facility

Thesis Abstract

The escalating demand for diagnostic accuracy in chest imaging has heightened concerns about ionizing radiation exposure in radiology departments, where repeated computed tomography (CT) scans pose potential long-term risks to patients. This study addresses the imperative to minimize radiation dose without compromising diagnostic quality by developing and validating a low-dose CT protocol tailored for chest imaging within a clinical radiology facility. The aims are to (1) formulate a low-dose CT protocol guided by dose-optimization principles, (2) evaluate image quality and diagnostic performance relative to standard-dose acquisitions, and (3) establish robust validation criteria for clinical implementation. Specific objectives include optimizing tube current modulation, tube voltage, and iterative reconstruction settings; quantifying radiation dose reductions using CTDIvol and Dose-Length Product (DLP) metrics; assessing objective image quality through signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR); and determining diagnostic concordance with conventional imaging for common chest pathologies, including pulmonary nodules, interstitial changes, and thoracic vascular structures. A contingent theoretical framework integrates the ALARA (as low as reasonably achievable) principle with the Technology Acceptance Model (TAM) to gauge clinical adoption dynamics. The study adopts a prospective, multi-phase design conducted in a tertiary hospital radiology department over 18 months. The population comprises adult patients undergoing clinically indicated chest CT for various indications. A sample size of 250 participants is calculated to detect a non-inferiority margin of 5% for diagnostic accuracy between low-dose and standard-dose protocols with 80% power and a 5% significance level. A stratified random sampling approach ensures representation across body mass index (BMI) categories and scanning indications. Data collection instruments include a standardized chest CT protocol record, a calibrated dosimetry log, and imaging assessment sheets completed by two independent radiologists blinded to acquisition protocol. Objective image quality metrics—SNR and CNR—are computed from predefined thoracic regions of interest, while qualitative diagnostic confidence is rated on a 5-point Likert scale. The instruments' validity and reliability are established through pilot testing and inter-rater reliability analysis (Cohen’s kappa). Data analysis employs descriptive statistics to summarize dose metrics and image quality parameters. Inferential analyses include paired t-tests or Wilcoxon signed-rank tests for continuous outcomes (SNR, CNR, and radiologist confidence scores), and McNemar tests for diagnostic decision concordance. Regression modeling (multivariate linear and logistic) investigates the relationship between BMI, scanner settings, and diagnostic quality outcomes, adjusting for potential confounders. Non-inferiority analyses compare low-dose and standard-dose protocols on key diagnostic endpoints, with equivalence testing applied to image quality indices. A subset analysis examines radiation-dose reduction achieved by adaptive prospective dose modulation across BMI strata. Ethical considerations adhere to the Declaration of Helsinki, with informed consent obtained and institutional review board approval secured. Data integrity is maintained through double data entry and secure storage. Expected findings anticipate substantial dose reduction (approximately 40–60% decrease in CTDIvol and DLP) with minimal impact on objective image quality (SNR/CNR within 10% of standard-dose values) and non-inferior diagnostic performance for most chest conditions. It is anticipated that nodular detection and characterization will remain clinically acceptable across BMI categories when iterative reconstruction algorithms (e.g., model-based or deep learning–assisted reconstructions) are optimized. The study contributes to knowledge by providing a validated, implementable low-dose chest CT protocol with explicit parameter settings and decision rules for clinical radiology practice, clarifying how dose reductions influence diagnostic confidence and patient safety. The findings are expected to inform policy on radiation stewardship, inform radiographer training, and guide scanner manufacturers in refining dose-modulation strategies. The main conclusion anticipates that a carefully optimized low-dose CT protocol can achieve substantial radiation-sparing benefits without compromising diagnostic integrity in chest imaging, thus supporting broader adoption in clinical workflows. Recommendations include standardization of the protocol across similar facilities, routine quality assurance cycles with periodic revalidation, continued exploration of advanced reconstruction techniques to further enhance image quality, and ongoing assessment of cost-benefit implications from a patient-centered radiation-ethics perspective.