Evaluating ChatGPT-4V in chest CT diagnostics: a critical image interpretation assessment.

PURPOSE: To assess the diagnostic accuracy of ChatGPT-4V in interpreting a set of four chest CT slices for each case of COVID-19, non-small cell lung cancer (NSCLC), and control cases, thereby evaluating its potential as an AI tool in radiological diagnostics. MATERIALS AND METHODS: In this retrospective study, 60 CT scans from The Cancer Imaging Archive, covering COVID-19, NSCLC, and control cases were analyzed using ChatGPT-4V. A radiologist selected four CT slices from each scan for evaluation. ChatGPT-4V's interpretations were compared against the gold standard diagnoses and assessed by two radiologists. Statistical analyses focused on accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), along with an examination of the impact of pathology location and lobe involvement. RESULTS: ChatGPT-4V showed an overall diagnostic accuracy of 56.76%. For NSCLC, sensitivity was 27.27% and specificity was 60.47%. In COVID-19 detection, sensitivity was 13.64% and specificity of 64.29%. For control cases, the sensitivity was 31.82%, with a specificity of 95.24%. The highest sensitivity (83.33%) was observed in cases involving all lung lobes. The chi-squared statistical analysis indicated significant differences in Sensitivity across categories and in relation to the location and lobar involvement of pathologies. CONCLUSION: ChatGPT-4V demonstrated variable diagnostic performance in chest CT interpretation, with notable proficiency in specific scenarios. This underscores the challenges of cross-modal AI models like ChatGPT-4V in radiology, pointing toward significant areas for improvement to ensure dependability. The study emphasizes the importance of enhancing these models for broader, more reliable medical use.

Improved detection of small pulmonary embolism on unenhanced computed tomography using an artificial intelligence-based algorithm - a single centre retrospective study.

To preliminarily verify the feasibility of a deep-learning (DL) artificial intelligence (AI) model to localize pulmonary embolism (PE) on unenhanced chest-CT by comparison with pulmonary artery (PA) CT angiography (CTA). In a monocentric study, we retrospectively reviewed 99 oncological patients (median age in years: 64 (range: 28-92 years); percentage of female: 39.4%) who received unenhanced and contrast-enhanced chest CT examinations in one session between January 2020 and October 2022 and who were diagnosed incidentally with PE. Findings in the unenhanced images were correlated with the contrast-enhanced images, which were considered the gold standard for central, segmental and subsegmental PE. The new algorithm was trained and tested based on the 99 unenhanced chest-CT image data sets. Based on them, candidate boxes, which were output by the model, were post-processed by evaluating whether the predicted box intersects with the patient's lung segmentation at any position. The AI-based algorithm proved to have an overall sensitivity of 54.5% for central, of 81.9% for segmental and 80.0% for subsegmental PE if taking n = 20 candidate boxes into account. Depending on the localization of the pulmonary embolism, the detection rate for only one box was: 18.1% central, 34.7% segmental and 0.0% subsegmental. The median volume of the clots differed significantly between the three subgroups and was 846.5 mm3 (IQR:591.1-964.8) in central, 201.3 mm3 (IQR:98.3-390.9) in segmental and 110.6 mm3 (IQR:94.3-128.0) in subsegmental PA (p < 0.05). The new algorithm proved to have high sensitivity in detecting PE in particular in segmental/subsegmental localization and may guide to decide whether a second contrast enhanced CT is necessary.

Enhancing Cone-Beam CT Image Quality in TIPSS Procedures Using AI Denoising.

Purpose: This study evaluates a deep learning-based denoising algorithm to improve the trade-off between radiation dose, image noise, and motion artifacts in TIPSS procedures, aiming for shorter acquisition times and reduced radiation with maintained diagnostic quality. Methods: In this retrospective study, TIPSS patients were divided based on CBCT acquisition times of 6 s and 3 s. Traditional weighted filtered back projection (Original) and an AI denoising algorithm (AID) were used for image reconstructions. Objective assessments of image quality included contrast, noise levels, and contrast-to-noise ratios (CNRs) through place-consistent region-of-interest (ROI) measurements across various critical areas pertinent to the TIPSS procedure. Subjective assessments were conducted by two blinded radiologists who evaluated the overall image quality, sharpness, contrast, and motion artifacts for each dataset combination. Statistical significance was determined using a mixed-effects model (p ≤ 0.05). Results: From an initial cohort of 60 TIPSS patients, 44 were selected and paired. The mean dose-area product (DAP) for the 6 s acquisitions was 5138.50 ± 1325.57 µGy·m2, significantly higher than the 2514.06 ± 691.59 µGym2 obtained for the 3 s series. CNR was highest in the 6 s-AID series (p < 0.05). Both denoised and original series showed consistent contrast for 6 s and 3 s acquisitions, with no significant noise differences between the 6 s Original and 3 s AID images (p > 0.9). Subjective assessments indicated superior quality in 6 s-AID images, with no significant overall quality difference between the 6 s-Original and 3 s-AID series (p > 0.9). Conclusions: The AI denoising algorithm enhances CBCT image quality in TIPSS procedures, allowing for shorter scans that reduce radiation exposure and minimize motion artifacts.