Fabien de Oliveira

Background: Foot deformities, particularly hallux valgus, significantly impact patients’ quality of life. Conventional radiographs are essential for their assessment, but manual measurements are time-consuming and variable. This study assessed the reliability of a deep learning-based solution (Milvue, France) that automates podiatry angle measurements from radiographs compared to manual measurements made by radiologists. Methods: A retrospective, non-interventional study at Perpignan Hospital analyzed the weight-bearing foot radiographs of 105 adult patients (August 2017–August 2022). The deep learning (DL) model’s measurements were compared to those of two radiologists for various angles (M1-P1, M1-M2, M1-M5, and P1-P2 for Djian–Annonier, calcaneal slope, first metatarsal slope, and Meary–Tomeno angles). Statistical analyses evaluated DL performance and inter-observer variability. Results: Of the 105 patients included (29 men and 76 women; mean age 55), the DL solution showed excellent consistency with manual measurements, except for the P1-P2 angle. The mean absolute error (MAE) for the frontal view was lowest for M1-M2 (0.96°) and highest for P1-P2 (3.16°). Intraclass correlation coefficients (ICCs) indicated excellent agreement for M1-P1, M1-M2, and M1-M5. For the lateral view, the MAE was 0.92° for calcaneal pitch and 2.83° for Meary–Tomeno, with ICCs ≥ 0.93. For hallux valgus detection, accuracy was 94%, sensitivity was 91.1%, and specificity was 97.2%. Manual measurements averaged 203 s per patient, while DL processing was nearly instantaneous. Conclusions: The DL solution reliably automates foot alignment assessments, significantly reducing time without compromising accuracy. It may improve clinical efficiency and consistency in podiatric evaluations.

Images from 64 patients undergoing an enhanced abdominal-pelvis scan at portal phase in dual-energy CT mode for the diagnosis of colitis or bowel obstruction were retrospectively analyzed. Acquisitions were performed on a third-generation dual-source CT (DSCT) 100/Sn150 kVp. Mixed images were generated, as well as virtual monoenergetic images (VMIs) at 40/50/60/70 keV. Objective image quality was assessed on VMIs and mixed images by measuring contrast, noise and contrast-to-noise ratio (CNR). Noise, smoothing and overall image quality were subjectively analyzed by two radiologists using Likert scales. For both patient groups, the noise decreased significantly according to the energy level from 40 to 60 keV by −47.2 ± 24.0% for bowel obstruction and −50.4 ± 18.2% for colitis. It was similar between 60 and 70 keV (p = 0.475 and 0.059, respectively). Noise values were significantly higher in VMIs than in mixed images, except for 70 keV (p = 0.53 and 0.071, respectively). Similar results were observed for contrast values, with a decrease between 40 and 70 keV of −56.3 ± 7.9% for bowel obstruction −56.2 ± 10.9% for colitis. The maximum CNR value was found at 60 keV compared to other energy levels and mixed images, but there was no significant difference with the other energy levels apart from 70 keV (−9.7 ± 9.8%) for bowel obstruction and 40 keV (−6.6 ± 8.2%) and 70 keV (−5.8 ± 9.2%) for colitis. The VMIs at 60 keV presented higher scores for all criteria for bowel obstruction and colitis, with no significant difference in smoothing score compared to mixed images (p = 0.119 and p = 0.888, respectively).

The study’s aim was to assess the impact of a deep learning image reconstruction algorithm (Precise Image; DLR) on image quality and liver metastasis conspicuity compared with an iterative reconstruction algorithm (IR). This retrospective study included all consecutive patients with at least one liver metastasis having been diagnosed between December 2021 and February 2022. Images were reconstructed using level 4 of the IR algorithm (i4) and the Standard/Smooth/Smoother levels of the DLR algorithm. Mean attenuation and standard deviation were measured by placing the ROIs in the fat, muscle, healthy liver, and liver tumor. Two radiologists assessed the image noise and image smoothing, overall image quality, and lesion conspicuity using Likert scales. The study included 30 patients (mean age 70.4 ± 9.8 years, 17 men). The mean CTDI_vol was 6.3 ± 2.1 mGy, and the mean dose-length product 314.7 ± 105.7 mGy.cm. Compared with i4, the HU values were similar in the DLR algorithm at all levels for all tissues studied. For each tissue, the image noise significantly decreased with DLR compared with i4 (p < 0.01) and significantly decreased from Standard to Smooth (−26 ± 10%; p < 0.01) and from Smooth to Smoother (−37 ± 8%; p < 0.01). The subjective image assessment confirmed that the image noise significantly decreased between i4 and DLR (p < 0.01) and from the Standard to Smoother levels (p < 0.01), but the opposite occurred for the image smoothing. The highest scores for overall image quality and conspicuity were found for the Smooth and Smoother levels.