Artificial Intelligence for Medical Image Analysis: A Guide for Authors and Reviewers.

OBJECTIVE: The purpose of this article is to highlight best practices for writing and reviewing articles on artificial intelligence for medical image analysis. CONCLUSION: Artificial intelligence is in the early phases of application to medical imaging, and patient safety demands a commitment to sound methods and avoidance of rhetorical and overly optimistic claims. Adherence to best practices should elevate the quality of articles submitted to and published by clinical journals.

The Value of Needle-Guidance Technology in Ultrasound-Guided Percutaneous Procedures Performed by Radiology Residents: A Comparison of Freehand, In-Plane, Fixed-Angle, and Electromagnetic Needle Tracking Techniques.

OBJECTIVES: Radiology residents typically learn ultrasound-guided procedures by performing supervised procedures on patients who may experience longer procedure times and higher complication rates. The purpose of this study was to determine if existing technologies, such as in-plane, fixed-angle guidance (IPFA) and electromagnetic needle tracking (ENT), can improve resident procedure time and accuracy. METHODS: Radiology residents (18 total) were randomized to 1 of 3 ultrasound-guidance technique groups-freehand, IPFA, or ENT-and instructed to place a needle into 4 liver lesions in a humanoid phantom, each increasing in difficulty. For each lesion, residents were timed from skin puncture to needle placement, and the number of times the needle was pulled back and redirected (pullbacks) was recorded. Primary outcomes were total time and total number of pullbacks for all 4 lesions. Secondary outcomes were individual time and number of pullbacks for each lesion. RESULTS: Compared to the freehand group, the IPFA and ENT groups demonstrated lower procedural time and number of pullbacks both in total and for each individual lesion. Differences in total time and total number of pullbacks were significant (P < .001), as were differences for lesion 3 (P = .002-.02) and lesion 4 (P < .001). Differences for lesions 1 and 2 were not statistically significant. CONCLUSIONS: Radiology resident procedure time and procedure accuracy (as judged by number of pullbacks) are significantly improved by the use IPFA and ENT guidance technologies.

Detection of Traumatic Pediatric Elbow Joint Effusion Using a Deep Convolutional Neural Network.

OBJECTIVE: The purpose of this study is to determine whether a deep convolutional neural network (DCNN) trained on a dataset of limited size can accurately diagnose traumatic pediatric elbow effusion on lateral radiographs. MATERIALS AND METHODS: A total of 901 lateral elbow radiographs from 882 pediatric patients who presented to the emergency department with upper extremity trauma were divided into a training set (657 images), a validation set (115 images), and an independent test set (129 images). The training set was used to train DCNNs of varying depth, architecture, and parameter initialization, some trained from randomly initialized parameter weights and others trained using parameter weights derived from pretraining on an ImageNet dataset. Hyperparameters were optimized using the validation set, and the DCNN with the highest ROC AUC on the validation set was selected for further performance testing on the test set. RESULTS: The final trained DCNN model had an ROC AUC of 0.985 (95% CI, 0.966-1.000) on the validation set and 0.943 (95% CI, 0.884-1.000) on the test set. On the test set, sensitivity was 0.909 (95% CI, 0.788-1.000), specificity was 0.906 (95% CI, 0.844-0.958), and accuracy was 0.907 (95% CI, 0.843-0.951). CONCLUSION: Accurate diagnosis of traumatic pediatric elbow joint effusion can be achieved using a DCNN.

Structure of RCC1 chromatin factor bound to the nucleosome core particle.

The small GTPase Ran enzyme regulates critical eukaryotic cellular functions including nuclear transport and mitosis through the creation of a RanGTP gradient around the chromosomes. This concentration gradient is created by the chromatin-bound RCC1 (regulator of chromosome condensation) protein, which recruits Ran to nucleosomes and activates Ran's nucleotide exchange activity. Although RCC1 has been shown to bind directly with the nucleosome, the molecular details of this interaction were not known. Here we determine the crystal structure of a complex of Drosophila RCC1 and the nucleosome core particle at 2.9 Å resolution, providing an atomic view of how a chromatin protein interacts with the histone and DNA components of the nucleosome. Our structure also suggests that the Widom 601 DNA positioning sequence present in the nucleosomes forms a 145-base-pair nucleosome core particle, not the expected canonical 147-base-pair particle.

Salvage Therapies After 18F-Fluciclovine Detected Prostate Cancer Recurrences.

BACKGROUND: F-Fluciclovine is the most recent prostate cancer (PCa)-directed PET radiotracer approved by the US Food and Drug Administration for detection of recurrent PCa. We report the treatments and outcomes of patients at our institution with PCa recurrences detected on F-fluciclovine PET/CT. METHODS: We identified men with recurrent PCa detected on F-fluciclovine PET/CT performed between 2017 and 2018 who were previously treated definitively and analyzed their patterns of care and cancer-specific outcomes. RESULTS: We identified 28 men with recurrent PCa detected on F-fluciclovine PET/CT. Twenty-three were initially treated with surgery and 13 also received postoperative radiation therapy (RT). Five patients were initially treated with definitive radiation. After surgery, the median time to F-fluciclovine PET/CT was 67 months (median prostate-specific antigen [PSA] of 1.63 ng/mL). After RT, the median time to F-fluciclovine PET/CT was 95 months with median PSA of 13.31 ng/mL. Six men recurred locally, 9 recurred in the pelvic nodes, 9 had distant nodal recurrences, and 4 had osseous metastases. Of the patients initially treated with surgery, 4 received salvage radiation and 3 received androgen deprivation therapy (ADT). Of the patients initially treated with surgery and postoperative RT, 3 received salvage pelvic nodal dissection, 4 received salvage radiation, and 2 received ADT. Of the patients initially treated with radiation, 4 received salvage ADT. All had PSA decline after salvage therapy. CONCLUSIONS: F-fluciclovine PET/CT can localize PCa recurrences, and subsequent salvage therapies appear effective with decreasing PSA. Longer follow-up will reveal if these diagnostic tests and subsequent therapies will improve PCa survival.

18F-Fluciclovine PET/CT Detection of Recurrent Prostate Carcinoma in Patients With Serum PSA ≤ 1 ng/mL After Definitive Primary Treatment.

PURPOSE: The aims of this study were to report on our initial experience using F-fluciclovine PET/CT to detect recurrent prostate carcinoma in patients with low serum prostate-specific antigen (PSA) after definitive treatment of primary disease and to conduct a preliminary investigation for factors associated with positive scan findings. PATIENTS AND METHODS: In this retrospective study, F-fluciclovine PET/CT scans from 28 men with suspected recurrence of prostate carcinoma and PSA values of 1 ng/mL or less were examined to identify the site(s) of disease recurrence. Differences in detection rate for Gleason scores of 7 and greater than 7, T2 and T3 disease, negative and positive surgical margins, and negative and positive seminal vesicle invasion were compared using the Fisher exact test. Mean PSA and mean PSA doubling time of patients with positive scans and negative scans were compared using the independent 2-group t test. RESULTS: At least one site of disease recurrence was identified in 13 (46.4%) of 28 patients. Disease detection rate was significantly higher in patients with history of Gleason score greater than 7 (Fisher exact test, P = 0.004). Mean PSA and PSA doubling time were not significantly different between patients with positive and negative F-fluciclovine PET/CT scans (P = 0.29 and 0.70, respectively). CONCLUSIONS: Detection of recurrent prostate cancer using F-fluciclovine PET/CT is possible in patients with low but rising PSA levels of 1 ng/mL or less. In such patients, local and nodal recurrences are more common than distant metastasis, and Gleason score greater than 7 is associated with positive scan results.

Automated Reporting of DXA Studies Using a Custom-Built Computer Program.

Dual-energy x-ray absorptiometry (DXA) scans are a critical population health tool and relatively simple to interpret but can be time consuming to report, often requiring manual transfer of bone mineral density and associated statistics into commercially available dictation systems. We describe here a custom-built computer program for automated reporting of DXA scans using Pydicom, an open-source package built in the Python computer language, and regular expressions to mine DICOM tags for patient information and bone mineral density statistics. This program, easy to emulate by any novice computer programmer, has doubled our efficiency at reporting DXA scans and has eliminated dictation errors.

RCC1 uses a conformationally diverse loop region to interact with the nucleosome: a model for the RCC1-nucleosome complex.

The binding of RCC1 (regulator of chromosome condensation 1) to chromatin is critical for cellular processes such as mitosis, nucleocytoplasmic transport, and nuclear envelope formation because RCC1 recruits the small GTPase Ran (Ras-related nuclear protein) to chromatin and sets up a Ran-GTP gradient around the chromosomes. However, the molecular mechanism by which RCC1 binds to nucleosomes, the repeating unit of chromatin, is not known. We have used biochemical approaches to test structural models for how the RCC1 beta-propeller protein could bind to the nucleosome. In contrast to the prevailing model, RCC1 does not appear to use the beta-propeller face opposite to its Ran-binding face to interact with nucleosomes. Instead, we find that RCC1 uses a conformationally flexible loop region we have termed the switchback loop in addition to its N-terminal tail to bind to the nucleosome. The juxtaposition of the RCC1 switchback loop to its Ran binding surface suggests a novel mechanism for how nucleosome-bound RCC1 recruits Ran to chromatin. Furthermore, this model accounts for previously unexplained observations for how Ran can interact with the nucleosome both dependent and independent of RCC1 and how binding of the nucleosome can enhance RCC1's Ran nucleotide exchange activity.