Differentiating kidney stones from phleboliths in unenhanced low-dose computed tomography using radiomics and machine learning.

OBJECTIVES: Distinguishing between kidney stones and phleboliths can constitute a diagnostic challenge in patients undergoing unenhanced low-dose CT (LDCT) for acute flank pain. We sought to investigate the accuracy of radiomics and a machine-learning classifier in differentiating between kidney stones and phleboliths on LDCT. METHODS: Radiomics features were extracted following a semi-automatic segmentation of kidney stones and phleboliths for two independent consecutive cohorts of patients undergoing LDCT for acute flank pain. Radiomics features from the first cohort of patients (n = 369) were ultimately used to train a machine-learning model designed to distinguish kidney stones (n = 211) from phleboliths (n = 201). Classification performance was assessed on the second independent cohort (i.e., testing set) (kidney stones n = 24; phleboliths n = 23) using positive and negative predictive values (PPV and NPV), area under the receiver operating curves (AUC), and permutation testing. RESULTS: Our machine-learning classification model trained on radiomics features achieved an overall accuracy of 85.1% on the independent testing set, with an AUC of 0.902, PPV of 81.5%, and NPV of 90.0%. Classification accuracy was significantly better than chance on permutation testing (p < 0.05, permutation p value). CONCLUSION: Radiomics and machine learning enable accurate differentiation between kidney stones and phleboliths on LDCT in patients presenting with acute flank pain. KEY POINTS: • Combining a machine-learning algorithm with radiomics features extracted for abdominopelvic calcification on LDCT offers a highly accurate method for discriminating phleboliths from kidney stones. • Our radiomics and machine-learning model proved robust for CT acquisition and reconstruction protocol when tested in comparison with an external independent cohort of patients with acute flank pain. • The high performance of the radiomics-based automatic classification model in differentiating phleboliths from kidney stones indicates its potential as a future diagnostic tool for equivocal abdominopelvic calcifications in the setting of suspected renal colic.

Impact of iterative reconstructions on objective and subjective emphysema assessment with computed tomography: a prospective study.

OBJECTIVES: To prospectively evaluate the impact of iterative reconstruction (IR) algorithms on pulmonary emphysema assessment as compared to filtered back projection (FBP). METHODS: One hundred ten unenhanced chest CT examinations were obtained on two different scanners. Image reconstructions from a single acquisition were done with different levels of IR and compared with FBP on the basis of the emphysema index (EI), lung volume and voxel densities. Objective emphysema assessment was performed with 3D software provided by each manufacturer. Subjective assessment of emphysema was performed as a blinded evaluation. Quantitative and subjective values were compared using repeated ANOVA analysis, Bland-Altman analysis and Kendall's coefficient of concordance (W). RESULTS: Lung volumes are stable on both units, throughout all IR levels (P ≥ 0.057). EI significantly decreases on both units with the use of any level of IR (P < 0.001). The highest levels of IR are responsible for a decrease of 33-36 % of EI. Significant differences in minimal lung density are found between the different algorithms (P < 0.003). Intra- and inter-reader concordance for emphysema characterisation is generally good (W ≥ 0.77 and W ≥ 0.86, respectively). CONCLUSIONS: Both commercially available IR algorithms used in this study significantly changed EI but did not alter visual assessment compared to standard FBP reconstruction at identical radiation exposure. KEY POINTS: • Objective quantification of pulmonary emphysema is sensitive to iterative reconstructions • Subjective evaluation of pulmonary emphysema is not influenced by iterative reconstructions • Consistency in reconstruction algorithms is of paramount importance for pulmonary emphysema monitoring.

Generative Adversarial Networks Improve the Reproducibility and Discriminative Power of Radiomic Features.

PURPOSE: To assess the contribution of a generative adversarial network (GAN) to improve intermanufacturer reproducibility of radiomic features (RFs). MATERIALS AND METHODS: The authors retrospectively developed a cycle-GAN to translate texture information from chest radiographs acquired using one manufacturer (Siemens) to chest radiographs acquired using another (Philips), producing fake chest radiographs with different textures. The authors prospectively evaluated the ability of this texture-translation cycle-GAN to reduce the intermanufacturer variability of RFs extracted from the lung parenchyma. This study assessed the cycle-GAN's ability to fool several machine learning (ML) classifiers tasked with recognizing the manufacturer on the basis of chest radiography inputs. The authors also evaluated the cycle-GAN's ability to mislead radiologists who were asked to perform the same recognition task. Finally, the authors tested whether the cycle-GAN had an impact on radiomic diagnostic accuracy for chest radiography in patients with congestive heart failure (CHF). RESULTS: RFs, extracted from chest radiographs after the cycle-GAN's texture translation (fake chest radiographs), showed decreased intermanufacturer RF variability. Using cycle-GAN-generated chest radiographs as inputs, ML classifiers categorized the fake chest radiographs as belonging to the target manufacturer rather than to a native one. Moreover, cycle-GAN fooled two experienced radiologists who identified fake chest radiographs as belonging to a target manufacturer class. Finally, reducing intermanufacturer RF variability with cycle-GAN improved the discriminative power of RFs for patients without CHF versus patients with CHF (from 55% to 73.5%, P < .001). CONCLUSION: Both ML classifiers and radiologists had difficulty recognizing the chest radiographs' manufacturer. The cycle-GAN improved RF intermanufacturer reproducibility and discriminative power for identifying patients with CHF. This deep learning approach may help counteract the sensitivity of RFs to differences in acquisition.Supplemental material is available for this article.© RSNA, 2020See also the commentary by Alderson in this issue.

Emphysema quantification using hybrid versus model-based generations of iterative reconstruction: SAFIRE versus ADMIRE.

To compare 2 incompatible generations of iterative reconstructions from the same raw dataset based on automatic emphysema quantification and noise reduction: a hybrid algorithm called sinogram affirmed iterative reconstruction (SAFIRE) versus a model-based algorithm called advanced modeled iterative reconstruction (ADMIRE).Raw datasets of 40 non-contrast thoracic computed tomography scanners obtained from a single acquisition on a SOMATOM Definition Flash unit (Siemens Healthcare, Forchheim) were reconstructed with 3 levels of SAFIRE and ADMIRE algorithms resulting in a total of 240 datasets. Emphysema index (EI) and image noise were compared using repeated analysis of variance (ANOVA) analysis with a P value <.05 considered statistically significant.EI and image noise were stable between both generations of IR when reconstructed with the same level (P ≥0.31 and P ≥0.06, respectively).SAFIRE and ADMIRE perform equally in terms of emphysema quantification and noise reduction.

Evaluating the effect of increased pitch, iterative reconstruction and dual source CT on dose reduction and image quality.

OBJECTIVE: To compare radiation dose and image quality of thoracoabdominal scans obtained with a high-pitch protocol (pitch 3.2) and iterative reconstruction (Sinogram Affirmed Iterative Reconstruction) in comparison to standard pitch reconstructed with filtered back projection (FBP) using dual source CT. METHODS: 114 CT scans (Somatom Definition Flash, Siemens Healthineers, Erlangen, Germany), 39 thoracic scans, 54 thoracoabdominal scans and 21 abdominal scans were performed. Analysis of three protocols was undertaken; pitch of 1 reconstructed with FBP, pitch of 3.2 reconstructed with SAFIRE, pitch of 3.2 with stellar detectors reconstructed with SAFIRE. Objective and subjective image analysis were performed. Dose differences of the protocols used were compared. RESULTS: Dose was reduced when comparing scans with a pitch of 1 reconstructed with FBP to high-pitch scans with a pitch of 3.2 reconstructed with SAFIRE with a reduction of volume CT dose index of 75% for thoracic scans, 64% for thoracoabdominal scans and 67% for abdominal scans. There was a further reduction after the implementation of stellar detectors reflected in a reduction of 36% of the dose-length product for thoracic scans. This was not at the detriment of image quality, contrast-to-noise ratio, signal-to-noise ratio and the qualitative image analysis revealed a superior image quality in the high-pitch protocols. CONCLUSION: The combination of a high pitch protocol with iterative reconstruction allows significant dose reduction in routine chest and abdominal scans whilst maintaining or improving diagnostic image quality, with a further reduction in thoracic scans with stellar detectors. Advances in knowledge: High pitch imaging with iterative reconstruction is a tool that can be used to reduce dose without sacrificing image quality.

Radiological findings of complications after lung transplantation.

Complications following lung transplantation may impede allograft function and threaten patient survival. The five main complications after lung transplantation are primary graft dysfunction, post-surgical complications, alloimmune responses, infections, and malignancy. Primary graft dysfunction, a transient ischemic/reperfusion injury, appears as a pulmonary edema in almost every patient during the first three days post-surgery. Post-surgical dysfunction could be depicted on computed tomography (CT), such as bronchial anastomosis dehiscence, bronchial stenosis and bronchomalacia, pulmonary artery stenosis, and size mismatch. Alloimmune responses represent acute rejection or chronic lung allograft dysfunction (CLAD). CLAD has three different forms (bronchiolitis obliterans syndrome, restrictive allograft syndrome, acute fibrinoid organizing pneumonia) that could be differentiated on CT. Infections are different depending on their time of occurrence. The first post-operative month is mostly associated with bacterial and fungal pathogens. From the second to sixth months, viral pneumonias and fungal and parasitic opportunistic infections are more frequent. Different patterns according to the type of infection exist on CT. Malignancy should be depicted and corresponded principally to post-transplantation lymphoproliferative disease (PTLD). In this review, we describe specific CT signs of these five main lung transplantation complications and their time of occurrence to improve diagnosis, follow-up, medical management, and to correlate these findings with pathology results. KEY POINTS: • The five main complications are primary graft dysfunction, surgical, alloimmune, infectious, and malignancy complications. • CT identifies anomalies in the setting of unspecific symptoms of lung transplantation complications. • Knowledge of the specific CT signs can allow a prompt diagnosis. • CT signs maximize the yield of bronchoscopy, transbronchial biopsy, and bronchoalveolar lavage. • Radiopathological correlation helps to understand CT signs after lung transplantation complications.

Noninvasive pulmonary nodule characterization using transcutaneous bioconductance: Preliminary results of an observational study.

We sought to assess the use of an electro pulmonary nodule (EPN) scanner (FreshMedx, Salt Lake City, UT) in the noninvasive characterization of pulmonary nodules using transcutaneous bioconductance.Monocentric prospective study including patients with a pulmonary nodule identified on a chest computed tomography scan. Study protocol approved by the institutional review board and written consent was obtained for every patient. 32 patients (12 females and 20 males), average age 65 years, and average lesion size 33.1 mm (range: 9-123 mm). Data collection by a trained physician, 62 skin surface measurements on the chest, arms, and hands bilaterally. Results were anonymized and mailed to a central data center for analysis and compared to histopathology.Pathology results obtained by percutaneous biopsy (n = 14), surgical biopsy (n = 1), or surgical resection (n = 17) showed 29 malignant lesions (adenocarcinoma n = 21, squamous cell carcinoma n = 5, typical carcinoid n = 1, metastasis n = 2), and 3 benign lesions (necrotic granuloma n = 1, no malignant cells on biopsy n = 2). EPN scanner results had a specificity of 66.67% (95% confidence interval [CI] 0.09-0.99), sensitivity 72.41% (95% CI 0.53-0.87), positive predictive value 95.45% (95% CI 0.81-0.99), and a negative predictive value 20.00% (95% CI 0.08-0.40).This pilot study showed a high positive predictive value of the EPN scanner, allowing aggressive management of lung nodules characterized as malignant. The low negative predictive value warrants further investigation of nodules that are characterized as benign.

Prostatic Utricle Cyst as the Most Likely Cause in a Case of Recurrent Episodes of Hematospermia.

Hematospermia is a clinical symptom that raises anxiety in patients and has various causes, benign and malignant. We report a case of hematospermia for which appropriate multidisciplinary expertise favored a conservative management of a benign prostatic cyst, namely, a prostatic utricle cyst. A cystic lesion found by transrectal ultrasound in the context of hematospermia related to masturbation in a young virgin male patient was investigated with a high-field magnetic resonance imaging (MRI) and an endorectal coil. The association of high-field MRI and endorectal coil leads to high quality images.

Unusual Presentation of Elastofibroma Dorsi on 18F-FDG-PET/CT: A Case Report.

A 70-year-old male patient underwent an Fluorodeoxyglucose-positron emission tomography-computed tomography for staging of a left parahilar lung neoplasm found during work-up for fatigue and asthenia. The scan demonstrated a hypermetabolic lung tumor, a hypermetabolic pleural effusion and 4 hypermetabolic bilateral soft tissue lesions of the chest wall corresponding to 4 elastofibroma dorsi. Initially, the oncologic disease was classified as stage IV because of the hypermetabolic pleural effusion. A transbronchial biopsy showed squamous cell carcinoma and the cytology of the pleural effusion revealed no malignant cells. As the other 4 hypermetabolic thoracic wall lesions were correctly diagnosed as benign despite their unusual presentation, the patient underwent surgery by left pneumonectomy and mediastinal lymphadenectomy. The lymph node involvement required adjuvant chemotherapy. Diagnostic confidence of the benignity of the hypermetabolic chest wall lesions allowed a more aggressive treatment with a better outcome after a malignant pleural effusion was excluded.

Total pancreatectomy and autologous islet cell transplantation as a means to treat severe chronic pancreatitis.

Autologous islet cell transplantation after near-total or total pancreatic resection can alleviate pain in patients with severe chronic pancreatitis and preserve endocrine function. From February 2000 to February 2003, a total of 22 patients, whose median age was 38 years, underwent pancreatectomy and autologous islet cell transplantation. Postoperative complications, metabolic studies, insulin usage, pain scores, and quality of life were recorded for all of these patients. The average number of islet cells harvested was 245,457 (range 20,850 to 607,466). Operative data revealed a mean estimated blood loss of 635 ml, an average operative time of 9 hours, and a mean length of hospital stay of 15 days. Sixty-eight percent of the patients had either a minor or major complication. Major complications included acute respiratory distress syndrome (n=2), intra-abdominal abscess (n=1), and pulmonary embolism (n=1). There were no deaths in our series. All patients demonstrated C-peptide and insulin production indicating graft function. Forty-one percent are insulin independent, and 27% required minimal amount of insulin or a sliding scale. All patients had preoperative pain and had been taking opioid analgesics; 82% no longer required analgesics postoperatively. Pancreatectomy with autologous islet cell transplantation can alleviate pain for patients with chronic pancreatitis and preserve endocrine function.