High yield clinical applications for photon counting CT in neurovascular imaging.

Photon-counting CT (PCCT) uses a novel X-ray detection mechanism that confers many advantages over that used in traditional energy integrating CT. As PCCT becomes more available, it is important to thoroughly understand its benefits and highest yield areas for improvements in diagnosis of various diseases. Based on our early experience, we have identified several areas of neurovascular imaging in which PCCT shows promise. Here, we describe the benefits in diagnosing arterial and venous diseases in the head, neck, and spine. Specifically, we focus on applications in head and neck CT angiography (CTA), spinal CT angiography, and CT myelography for detection of CSF-venous fistulas. Each of these applications highlights the technological advantages of PCCT in neurovascular imaging. Further understanding of these applications will not only benefit institutions incorporating PCCT into their practices but will also help guide future directions for implementation of PCCT for diagnosing other pathologies in neuroimaging.

Back to the Future: Dynamic Contrast-Enhanced Photon-Counting Detector CT for the Detection of Pituitary Adenoma in Cushing Disease.

Historically, MR imaging has been unable to detect a pituitary adenoma in up to one-half of patients with Cushing disease. This issue is problematic because the standard-of-care treatment is surgical resection, and its success is correlated with finding the tumor on imaging. Photon-counting detector CT is a recent advancement that has multiple benefits over conventional energy-integrating detector CT. We present the use of dynamic contrast-enhanced imaging using photon-counting detector CT for the detection of pituitary adenomas in patients with Cushing disease.

The Number of Radiographically Positive Lymph Nodes Further Stratifies Patient Survival Among Clinical N1 Patients With Human Papillomavirus-Associated Oropharyngeal Cancer.

Purpose: Human papillomavirus-associated oropharyngeal squamous cell carcinoma (HPV[+]OPSCC) requires further study to optimize the existing clinical staging system and guide treatment selection. We hypothesize that incorporation of the number of radiographically positive lymph nodes will further stratify patients with clinical N1 (cN1) HPV(+)OPSCC. Methods and Materials: A post hoc analysis from 2 prospective clinical trials at a high-volume referral center was conducted. Patients underwent primary tumor resection and lymphadenectomy, followed by either standard-of-care radiation therapy (60 Gy in 30 fractions) with or without cisplatin (40 mg/m2 weekly) or de-escalated radiation therapy (30 Gy in 20 twice-daily fractions) with concomitant 15 mg/m2 docetaxel once weekly. Imaging studies were independently reviewed by a blinded neuroradiologist classifying radiographic extranodal extension (rENE) and the number and maximal size of involved lymph nodes. Patients without pathologic data available for assessment were excluded. Results: A total of 260 patients were included. Of these, 216 (83%) were cN1. Patients had a median of 2 radiographically positive lymph nodes (range, 0-12), and 107 (41%) had rENE. For cN1 patients, stratifying by radiographically positive lymph nodes (1-2 vs 3-4 vs >4) was predictive of progression-free survival (PFS) (P = .017), with 2-year PFS rates of 96%, 88%, and 81%, respectively. More than 2 radiographically positive lymph nodes was identified as a significant threshold for PFS (P = .0055) and overall survival (P = .029). Radiographic ENE and lymph node size were not predictive of PFS among cN1 patients. Conclusions: The number of radiographically positive lymph nodes is predictive of PFS and overall survival and could be used to meaningfully subcategorize cN1 patients with HPV(+)OPSCC. We recommend further validation of our proposal that cN1 patients with 1 to 2 radiologically positive lymph nodes be categorized as cN1a, patients with 3 to 4 radiologically positive lymph nodes categorized as cN1b, and patients with >4 radiographically positive lymph nodes categorized as cN1c.

Manifestations of radiation toxicity in the head, neck, and spine: An image-based review.

BACKGROUND AND PURPOSE: Radiation therapy is an important component of treatment in patients with malignancies of the head, neck, and spine. However, radiation to these regions has well-known potential side effects, many of which can be encountered on imaging. In this manuscript, we review selected radiographic manifestations of therapeutic radiation to the head, neck, and spine that may be encountered in the practice of radiology. METHODS: We conducted an extensive literature review of known complications of radiation therapy in the head, neck, and spine. We excluded intracranial and pulmonary radiation effects from our search. We selected complications that had salient, recognizable imaging findings. We searched our imaging database for illustrative examples of these complications. RESULTS: Based on our initial literature search and imaging database review, we selected cases of radiation-induced tumors, radiation tissue necrosis (osteoradionecrosis and soft tissue necrosis), carotid stenosis and blowout secondary to radiation, enlarging thyroglossal duct cysts, radiation myelopathy, and radiation-induced vertebral compression fractures. CONCLUSIONS: We describe the clinical and imaging features of selected sequelae of radiation therapy to the head, neck, and spine, with a focus on those with characteristic imaging findings that can be instrumental in helping to make the diagnosis. Knowledge of these entities and their imaging findings is crucial for accurate diagnosis. Not only do radiologists play a key role in early detection of these entities, but many of these entities can be misinterpreted if one is not familiar with them.

Correlation between radiographic and pathologic lymph node involvement and extranodal extension via CT and PET in HPV-associated oropharyngeal cancer.

PURPOSE/OBJECTIVES: Extranodal extension (ENE) and more than 4 pathologically involved lymph nodes (pN2) are critical prognostic factors in HPV-associated oropharyngeal cancer (HPV(+) OPSCC). We analyze a patient cohort with HPV(+) OPSCC to determine the sensitivity and specificity of CT and PET/CT in identifying involvement of more than 4 lymph nodes (rN2) compared to pN2 and radiographic ENE (rENE) compared to pathologic ENE (pENE). MATERIALS/METHODS: The dataset consisted of 261 patients enrolled in two prospective clinical trials. All imaging studies were independently reviewed by a blinded neuroradiologist, classifying the presence or absence of rENE and rN2. Secondary analyses included correlations of imaging accuracy and pathologic size of the primary tumor. RESULTS: CT demonstrated sensitivity of 59%, specificity of 92%, positive predictive value (PPV) of 53%, negative predictive value (NPV) of 94%, and accuracy of 88% for pN2. PET/CT showed similar results. Patients with up to 4 involved lymph nodes (rN0-1) had a 93-94% chance of remaining pN0-1. CT and PET/CT identified an equal number of involved nodes in 81% of patients. CT demonstrated sensitivity of 54%, specificity of 71%, PPV of 72%, NPV of 53%, and accuracy of 62% for pENE. PET/CT showed similar outcomes. Notably, when multiple radiographic criteria were used to identify rENE, PPV increased for both CT (84%) and PET/CT (79%). CONCLUSION: Patients with rN0-1 had a 93-94% chance of remaining pN0-1, suggesting rN0-1 could effectively stratify patients for clinical trials and treatment de-escalation. While CT and PET/CT were highly correlated, both showed low sensitivity for identifying pENE.

e-ASPECTS software improves interobserver agreement and accuracy of interpretation of aspects score.

INTRODUCTION: There is increased interest in the use of artificial intelligence-based (AI) software packages in the evaluation of neuroimaging studies for acute ischemic stroke. We studied whether, compared to standard image interpretation without AI, Brainomix e-ASPECTS software improved interobserver agreement and accuracy in detecting ASPECTS regions affected in anterior circulation LVO. METHODS: We included 60 consecutive patients with anterior circulation LVO who had TICI 3 revascularization within 60 minutes of their baseline CT. A total of 16 readers, including senior neuroradiologists, junior neuroradiologists and vascular neurologists participated. Readers interpreted CT scans on independent workstations and assessed final ASPECTS and evaluated whether each individual ASPECTS region was affected. Two months later, readers again evaluated the CT scans, but with assistance of e-ASPECTS software. We assessed interclass correlation coefficient for total ASPECTS and interobserver agreement with Fleiss' Kappa for each ASPECTS region with and without assistance of the e-ASPECTS. We also assessed accuracy for the readers with and without e-ASPECTS assistance. In our assessment of accuracy, ground truth was the 24 hour CT in this cohort of patients who had prompt and complete revascularization. RESULTS: Interclass correlation coefficient for total ASPECTS without e-ASPECTS assistance was 0.395, indicating fair agreement compared, to 0.574 with e-ASPECTS assistance, indicating good agreement (P < 0.01). There was significant improvement in inter-rater agreement with e-ASPECTS assistance for each individual region with the exception of M6 and caudate. The e-ASPECTS software had higher accuracy than the overall cohort of readers (with and without e-ASPECTS assistance) for every region except the caudate. CONCLUSIONS: Use of Brainomix e-ASPECTS software resulted in significant improvements in inter-rater agreement and accuracy of ASPECTS score evaluation in a large group of neuroradiologists and neurologists. e-ASPECTS software was more predictive of final infarct/ASPECTS than the overall group interpreting the CT scans with and without e-ASPECTS assistance.

Intracranial long-term complications of radiation therapy: an image-based review.

BACKGROUND AND PURPOSE: Radiation therapy is commonly utilized in the majority of solid cancers and many hematologic malignancies and other disorders. While it has an undeniably major role in improving cancer survival, radiation therapy has long been recognized to have various negative effects, ranging from mild to severe. In this manuscript, we review several intracranial manifestations of therapeutic radiation, with particular attention to those that may be encountered by radiologists. METHODS: We conducted an extensive literature review of known complications of intracranial radiation therapy. Based on this review, we selected complications that had salient, recognizable imaging findings. We searched our imaging database for illustrative examples of these complications, focusing only on patients who had a history of intracranial radiation therapy. We then selected cases that best exemplified expected imaging findings in these entities. RESULTS: Based on our initial literature search and imaging database review, we selected cases of radiation-induced meningioma, radiation-induced glioma, cavernous malformation, enlarging perivascular spaces, leukoencephalopathy, stroke-like migraine after radiation therapy, Moyamoya syndrome, radiation necrosis, radiation-induced labyrinthitis, optic neuropathy, and retinopathy. Although retinopathy is not typically apparent on imaging, it has been included given its clinical overlap with optic neuropathy. CONCLUSIONS: We describe the clinical and imaging features of selected sequelae of intracranial radiation therapy, with a focus on those most relevant to practicing radiologists. Knowledge of these complications and their imaging findings is important, because radiologists play a key role in early detection of these entities.

Risk factors for carcinoma ex pleomorphic adenoma in patients presenting with recurrence after resection of pleomorphic adenoma.

BACKGROUND: In patients presenting with recurrent pleomorphic adenoma (rPA), clinical evaluation can fail to recognize carcinoma ex PA (cxPA). We aim to identify the risk factors for cxPA. METHODS: This is a single institution retrospective case-control study from 2000 to 2015. CxPA was diagnosed based on surgical pathology. Demographics, clinical, and social histories were collected. RESULTS: A number of 13/106 (12.3%) patients were diagnosed with cxPA, of which only 4/13 (31%) had clinical features suspicious for malignancy. Compared to benign rPA, factors associated with cxPA included age >50 (odds ratio [OR] 6.67, 95% confidence interval [CI]: 1.71-25.98, P < .01), >10 pack-years of smoking history (OR 3.36, 95% CI: 1.01-11.14, P = .04), and the largest tumor being >2 cm on pathology (OR 4.42, 95% CI: 1.14-17.10, P = .03). CONCLUSIONS: In patients presenting with rPA, risk factors for malignant transformation include age >50, significant smoking history, and tumors larger than 2 cm. Clinical signs of malignancy such as rapid growth or pain are not always present.

Risk factors for subsequent recurrence after surgical treatment of recurrent pleomorphic adenoma of the parotid gland.

BACKGROUND: Recurrent pleomorphic adenoma (PA) can be a lifelong disease, and rates of subsequent recurrence are high. METHODS: Patients between 2000 and 2015 were identified. Primary outcome was subsequent recurrence after surgical salvage. RESULTS: Twenty-seven of 84 patients developed a subsequent recurrence. Risk factors for subsequent recurrence included a higher number of previous recurrences (P < .01), worse preoperative facial nerve function (P < .01), and deep parotid lesion(s) (P < .01). Interval since last surgery was protective (P < .01), specifically >10 years since last surgery (P < .01). For patients with a >10-year interval since their last surgery, the subsequent recurrence-free rate at 10 years follow-up was 80.2% vs 31.8%. CONCLUSIONS: For patients presenting with a >10-year interval since their last surgery, subsequent recurrence rates are low, which may allow for as needed surveillance recommendations. For patients presenting with recurrent PA and ≤10 years since their last surgery, a closer surveillance is warranted.

Deep neural network to locate and segment brain tumors outperformed the expert technicians who created the training data.

Purpose: Deep learning (DL) algorithms have shown promising results for brain tumor segmentation in MRI. However, validation is required prior to routine clinical use. We report the first randomized and blinded comparison of DL and trained technician segmentations. Approach: We compiled a multi-institutional database of 741 pretreatment MRI exams. Each contained a postcontrast T1-weighted exam, a T2-weighted fluid-attenuated inversion recovery exam, and at least one technician-derived tumor segmentation. The database included 729 unique patients (470 males and 259 females). Of these exams, 641 were used for training the DL system, and 100 were reserved for testing. We developed a platform to enable qualitative, blinded, controlled assessment of lesion segmentations made by technicians and the DL method. On this platform, 20 neuroradiologists performed 400 side-by-side comparisons of segmentations on 100 test cases. They scored each segmentation between 0 (poor) and 10 (perfect). Agreement between segmentations from technicians and the DL method was also evaluated quantitatively using the Dice coefficient, which produces values between 0 (no overlap) and 1 (perfect overlap). Results: The neuroradiologists gave technician and DL segmentations mean scores of 6.97 and 7.31, respectively ( p < 0.00007 ). The DL method achieved a mean Dice coefficient of 0.87 on the test cases. Conclusions: This was the first objective comparison of automated and human segmentation using a blinded controlled assessment study. Our DL system learned to outperform its "human teachers" and produced output that was better, on average, than its training data.

Recurrent pleomorphic adenoma of the parotid gland: A comparison of radiographic and pathologic tumor burden.

PURPOSE: To compare recurrent pleomorphic adenoma tumor burden as detected on magnetic resonance and computerized tomography imaging with postoperative histopathology. MATERIALS AND METHODS: 44 patients were identified at a tertiary medical center between 2000 and 2015. Patients were included if they had viewable preoperative imaging and a postoperative diagnosis of recurrent pleomorphic adenoma. Primary outcomes were differences in the number and size of lesions detected on imaging and pathology. RESULTS: The size in greatest dimension between pathology and imaging was not significant on aggregate MRI + CT (p = 0.78), MRI (p = 0.41), or CT (p = 0.69). There were more lesions found on pathology compared to both aggregate MRI + CT (p = 0.003) and CT alone (p = 0.014). The number of lesions between MRI and pathology failed to reach significance (p = 0.06). On univariate analysis, the interval between imaging and pathology (recurrent surgery) did not significantly affect the number of lesions detected (p = 0.18). On multivariable analysis, CT as the primary imaging modality and >1 recurrence was independently associated with greater inaccuracy with respect to number of lesions detected (p = 0.006; p = 0.008). CONCLUSION: The size of the largest lesion on pathology can be accurately determined with imaging. Compared to MRI, CT scans significantly underpredict the number of lesions found on pathology. MRI should be prioritized unless contraindications exist. These findings will help guide imaging choice, preoperative planning, and patient counseling.

Clinical Assessment of Metal Artifact Reduction Methods in Dual-Energy CT Examinations of Instrumented Spines.

OBJECTIVE: The purpose of this study was to evaluate the performance of three metal artifact reduction methods in dual-energy CT (DECT) examinations of instrumented spines. MATERIALS AND METHODS: Twenty patients with instrumented spines who underwent spine DECT were retrospectively identified. All scans were obtained on a dual-source 128-MDCT scanner. In addition to the original DE mixed images, DECT images were reconstructed using an iterative metal artifact reconstruction algorithm (DE iMAR), virtual monochromatic imaging (VMI) algorithm (DE Mono+), and a combination of the two algorithms DE iMAR and DE Mono+, which we refer to here as "DE iMAR Mono+." The four image series were anonymized and randomized for a reader study. Four experienced neuroradiologists rated the images in terms of artifact scores of four anatomic regions and overall image quality scores in both bone and soft-tissue display window settings. In addition, a quantitative analysis was performed to assess the performance of the three metal artifact reduction methods. RESULTS: There were statistically significant differences in the artifact scores and overall image quality scores among the four methods (both, p < 0.001). DE iMAR Mono+ showed the best artifact scores and quality scores (all, p < 0.001). The intraclass correlation coefficient for the overall image quality score was 0.779 using the bone display window and 0.892 using the soft-tissue display window (both, p < 0.001). In addition, DE iMAR Mono+ reduced the artifacts by the greatest amount in the quantitative analysis. CONCLUSION: The method that used DE iMAR Mono+ showed the best performance of spine metal artifact reduction using DECT data. These results may be specific to this CT vendor and implant type.

Structured Head and Neck CT Angiography Reporting Reduces Resident Revision Rates.

PURPOSE: This resident-driven quality improvement project was undertaken to assess the effectiveness of structured reporting to reduce revision rates for afterhours reports dictated by residents. METHODS: The first part of the study assessed baseline revision rates for head and neck CT angiography (CTA) examinations dictated by residents during afterhours call. A structured report was subsequently created based on templates on the RSNA informatics reporting website and critical findings that should be assessed for on all CTA examinations. The template was made available to residents through the speech recognition software for all head and neck CTA examinations for a duration of 2 months. Report revision rates were then compared with and without use of the structured template. RESULTS: The structured template was found to reduce revision rates by approximately 50% with 10/41 unstructured reports revised and 2/17 structured reports revised. CONCLUSIONS: We believe that structured reporting can help reduce reporting errors, particularly in term of typographical errors, train residents to evaluate complex examinations in a systematic fashion, and assist them in recalling critical findings on these examinations.

The flip-flop fungus sign: an FDG PET/CT sign of benignity.

Benign granulomatous processes such as fungal infection may mimic metastatic lung cancer on FDG PET/CT. We found that these processes often have draining lymph node(s) with equal or greater FDG activity than associated lung nodule(s), a "flip-flop" of what is commonly seen in lung cancer. The aim of this study was to examine the utility of this "flip-flop fungus" (FFF) sign for diagnosing benign pulmonary disease. FDG PET/CT scans performed between 9/09-3/13 for the indications of pulmonary nodule or mass were reviewed. Scans with at least one hilar or mediastinal FDG avid draining node were included. Patients with a history of cancer, lack of pathologic confirmation, or without at least two years of imaging follow-up were excluded. A total of 209 FDG PET/CT exams were included and reviewed in a blinded fashion. A positive FFF sign had a sensitivity of 60.0% (95% CI: 47.6-71.5%) and specificity of 84.9% (95% CI: 77.8-90.4%) (P<0.0001) for benign disease. With additional strict imaging criteria applied, the FFF sign had a specificity of 98.6% (95% CI: 94.9-99.8%) (P<0.0001) and a positive predictive value of 90.0% (95% CI: 68.3-98.5%). A positive FFF sign was predominately due to granulomatous disease (91%), mostly histoplasmosis (73%). A positive FFF sign combined with positive fungal serology (n=16) had a specificity of 100% for benign disease. The FFF sign predicts benign disease in patients with a lung nodule(s) and an FDG avid draining lymph node(s) that would otherwise be considered worrisome for cancer.