Search

Context.—: Stanford Pathology began stepwise subspecialty implementation of whole slide imaging (WSI) in 2018 soon after the first US Food and Drug Administration approval. In 2020, during the COVID-19 pandemic, the Centers for Medicare & Medicaid Services waived the requirement for pathologists to perform diagnostic tests in Clinical Laboratory Improvement Amendments (CLIA)-licensed facilities. This encouraged rapid implementation of WSI across all surgical pathology subspecialties., Objective.—: To present our experience with validation and implementation of WSI at a large academic medical center encompassing a caseload of more than 50 000 cases per year., Design.—: Validation was performed independently for 3 subspecialty services with a diagnostic concordance threshold above 95%. Analysis of user experience, staffing, infrastructure, and information technology was performed after department-wide expansion., Results.—: Diagnostic concordance was achieved in 96% of neuropathology cases, 100% of gynecologic pathology cases, and 98% of immunohistochemistry cases. After full implementation, 8 high-capacity scanners were operational, with whole slide images generated on greater than 2000 slides per weekday, accounting for approximately 80% of histologic slides at Stanford Medicine. Multiple modifications in workflow and information technology were needed to improve performance. Within months of full implementation, most attending pathologists and trainees had adopted WSI for primary diagnosis., Conclusions.—: WSI across all surgical subspecialities is achievable at scale at an academic medical center; however, adoption required flexibility to adjust workflows and develop tailored solutions. WSI at scale supported the health and safety of medical staff while facilitating high-quality patient care and education during COVID-19 restrictions.

Background Radiogenomics of pediatric medulloblastoma (MB) offers an opportunity for MB risk stratification, which may aid therapeutic decision making, family counseling, and selection of patient groups suitable for targeted genetic analysis. Purpose To develop machine learning strategies that identify the four clinically significant MB molecular subgroups. Materials and Methods In this retrospective study, consecutive pediatric patients with newly diagnosed MB at MRI at 12 international pediatric sites between July 1997 and May 2020 were identified. There were 1800 features extracted from T2- and contrast-enhanced T1-weighted preoperative MRI scans. A two-stage sequential classifier was designed-one that first identifies non-wingless ( WNT ) and non-sonic hedgehog ( SHH ) MB and then differentiates therapeutically relevant WNT from SHH . Further, a classifier that distinguishes high-risk group 3 from group 4 MB was developed. An independent, binary subgroup analysis was conducted to uncover radiomics features unique to infantile versus childhood SHH subgroups. The best-performing models from six candidate classifiers were selected, and performance was measured on holdout test sets. CIs were obtained by bootstrapping the test sets for 2000 random samples. Model accuracy score was compared with the no-information rate using the Wald test. Results The study cohort comprised 263 patients (mean age ± SD at diagnosis, 87 months ± 60; 166 boys). A two-stage classifier outperformed a single-stage multiclass classifier. The combined, sequential classifier achieved a microaveraged F1 score of 88% and a binary F1 score of 95% specifically for WNT . A group 3 versus group 4 classifier achieved an area under the receiver operating characteristic curve of 98%. Of the Image Biomarker Standardization Initiative features, texture and first-order intensity features were most contributory across the molecular subgroups. Conclusion An MRI-based machine learning decision path allowed identification of the four clinically relevant molecular pediatric medulloblastoma subgroups. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Chaudhary and Bapuraj in this issue.

Background: Clinicians and machine classifiers reliably diagnose pilocytic astrocytoma (PA) on magnetic resonance imaging (MRI) but less accurately distinguish medulloblastoma (MB) from ependymoma (EP). One strategy is to first rule out the most identifiable diagnosis., Objective: To hypothesize a sequential machine-learning classifier could improve diagnostic performance by mimicking a clinician's strategy of excluding PA before distinguishing MB from EP., Methods: We extracted 1800 total Image Biomarker Standardization Initiative (IBSI)-based features from T2- and gadolinium-enhanced T1-weighted images in a multinational cohort of 274 MB, 156 PA, and 97 EP. We designed a 2-step sequential classifier - first ruling out PA, and next distinguishing MB from EP. For each step, we selected the best performing model from 6-candidate classifier using a reduced feature set, and measured performance on a holdout test set with the microaveraged F1 score., Results: Optimal diagnostic performance was achieved using 2 decision steps, each with its own distinct imaging features and classifier method. A 3-way logistic regression classifier first distinguished PA from non-PA, with T2 uniformity and T1 contrast as the most relevant IBSI features (F1 score 0.8809). A 2-way neural net classifier next distinguished MB from EP, with T2 sphericity and T1 flatness as most relevant (F1 score 0.9189). The combined, sequential classifier was with F1 score 0.9179., Conclusion: An MRI-based sequential machine-learning classifiers offer high-performance prediction of pediatric posterior fossa tumors across a large, multinational cohort. Optimization of this model with demographic, clinical, imaging, and molecular predictors could provide significant advantages for family counseling and surgical planning., (© Congress of Neurological Surgeons 2021.)

Background and Purpose: Atypical teratoid/rhabdoid tumors and medulloblastomas have similar imaging and histologic features but distinctly different outcomes. We hypothesized that they could be distinguished by MR imaging-based radiomic phenotypes., Materials and Methods: We retrospectively assembled T2-weighted and gadolinium-enhanced T1-weighted images of 48 posterior fossa atypical teratoid/rhabdoid tumors and 96 match-paired medulloblastomas from 7 institutions. Using a holdout test set, we measured the performance of 6 candidate classifier models using 6 imaging features derived by sparse regression of 900 T2WI and 900 T1WI Imaging Biomarker Standardization Initiative-based radiomics features., Results: From the originally extracted 1800 total Imaging Biomarker Standardization Initiative-based features, sparse regression consistently reduced the feature set to 1 from T1WI and 5 from T2WI. Among classifier models, logistic regression performed with the highest AUC of 0.86, with sensitivity, specificity, accuracy, and F1 scores of 0.80, 0.82, 0.81, and 0.85, respectively. The top 3 important Imaging Biomarker Standardization Initiative features, by decreasing order of relative contribution, included voxel intensity at the 90th percentile, inverse difference moment normalized, and kurtosis-all from T2WI., Conclusions: Six quantitative signatures of image intensity, texture, and morphology distinguish atypical teratoid/rhabdoid tumors from medulloblastomas with high prediction performance across different machine learning strategies. Use of this technique for preoperative diagnosis of atypical teratoid/rhabdoid tumors could significantly inform therapeutic strategies and patient care discussions., (© 2021 by American Journal of Neuroradiology.)

Background and Purpose: Fractional tumor burden better correlates with histologic tumor volume fraction in treated glioblastoma than other perfusion metrics such as relative CBV. We defined fractional tumor burden classes with low and high blood volume to distinguish tumor from treatment effect and to determine whether fractional tumor burden can inform treatment-related decision-making., Materials and Methods: Forty-seven patients with high-grade gliomas (primarily glioblastoma) with recurrent contrast-enhancing lesions on DSC-MR imaging were retrospectively evaluated after surgical sampling. Histopathologic examination defined treatment effect versus tumor. Normalized relative CBV thresholds of 1.0 and 1.75 were used to define low, intermediate, and high fractional tumor burden classes in each histopathologically defined group. Performance was assessed with an area under the receiver operating characteristic curve. Consensus agreement among physician raters reporting hypothetic changes in treatment-related decisions based on fractional tumor burden was compared with actual real-time treatment decisions., Results: Mean lower fractional tumor burden, high fractional tumor burden, and relative CBV of the contrast-enhancing volume were significantly different between treatment effect and tumor ( P = .002, P < .001, and P < .001), with tumor having significantly higher fractional tumor burden and relative CBV and lower fractional tumor burden. No significance was found with intermediate fractional tumor burden. Performance of the area under the receiver operating characteristic curve was the following: high fractional tumor burden, 0.85; low fractional tumor burden, 0.7; and relative CBV, 0.81. In comparing treatment decisions, there were disagreements in 7% of tumor and 44% of treatment effect cases; in the latter, all disagreements were in cases with scattered atypical cells., Conclusions: High fractional tumor burden and low fractional tumor burden define fractions of the contrast-enhancing lesion volume with high and low blood volume, respectively, and can differentiate treatment effect from tumor in recurrent glioblastomas. Fractional tumor burden maps can also help to inform clinical decision-making., (© 2019 by American Journal of Neuroradiology.)

We describe two cases of neonatal onset interstitial lung disease eventually diagnosed as mucopolysaccharidosis type I (MPS I). In both cases, evaluation led to lung biopsy, pathology review, and identification of glycogen deposition. Pulmonary interstitial glycogenosis (PIG) was considered as a clinical diagnosis in case one; however, further review of electron microscopy (EM) was more consistent with MPS I rather than PIG. Both cases were confirmed to have MPS I by enzyme and molecular analysis. Neonatal interstitial lung disease is an atypical presentation for MPS I which is likely under-recognized. Diagnosis through clinical guidelines and a multidisciplinary approach had a major impact on patient management. The diagnosis of MPS I prompted timely initiation of enzyme replacement therapy (ERT) and the patients ultimately underwent hematopoietic stem cell transplantation (HSCT) to improve symptomatic outcomes. In addition to treatment, immediate precautionary recommendations were made to avoid potentially catastrophic outcomes associated with cervical instability. These cases add to the clinical spectrum of MPS I in the newborn period. They further illustrate the difficulties in early recognition of the disease, and importance of a definitive diagnosis of MPS I in infants with interstitial lung disease.

Nonatherosclerotic cerebrovascular arteriopathies share epidemiologic and clinical features, but few studies directly compare histologic features of the intracranial vasculature. We studied 3 adult autopsy cases of fibromuscular dysplasia in patients who died of basilar artery aneurysm rupture, vertebral artery dissection, or Moyamoya syndrome. Fibromuscular dysplasia was only identified when multiple sections (optimally of the entire circle of Willis) were examined by microscopy. A fourth case of a massive subcutaneous scalp cirsoid aneurysm with classic "string-of-beads" gross appearance and microscopic medial hypertrophy was also compatible with fibromuscular dysplasia. Intracranial vascular changes were compared with those in 1 patient with Ehlers-Danlos type IV (vascular type) and in 4 patients with neurofibromatosis I. Distinct histologic features and distributions of cerebral vessel abnormalities were observed in all 3 disorders. Disordered collagen within the muscularis (identified using picrosirius red histochemistry) was confined to fibromuscular dysplasia; fibrocellular smooth muscle intimal proliferation within parenchymal cerebral arteries was confirmed using smooth muscle actin immunohistochemistry in the Moyamoya case; the patient with Ehlers-Danlos type IV showed aneurysm formation and eccentric intimal thickening of circle of Willis vessels but no obvious abnormalities of the muscularis; and neurofibromatosis I cases showed extensive leptomeningeal smaller-caliber arterial disease that particularly affected the spinal cord. Thus, cranial/intracranial artery involvement is not rare in these conditions but requires extensive sampling to identify the range of features.

Ventriculitis or periventriculitis as a predominant pattern of tissue involvement in cerebral toxoplasmosis was always a rare event, even at the height of the acquired immunodeficiency syndrome (AIDS) era. Ventriculitis on premortem neuroimaging or at autopsy in AIDS patients chiefly led to differential diagnoses of primary central nervous system lymphoma (PCNSL) or cytomegalovirus ventriculitis, not toxoplasmosis. Usually cerebral toxoplasmosis manifests as multifocal, necrotizing, hemorrhagic foci of cerebritis or abscesses. We report two non-AIDS patients with cerebral toxoplasmosis that presented with predominant ventriculitis/periventriculitis, with diagnosis in both cases made only at postmortem examination. A 90-year-old woman, with autoimmune hemolytic anemia and large granular lymphocytic leukemia diagnosed 2 1/2 years prior, presented with altered mental status. Neuroimaging revealed a necrotic 5.4 × 4.6 × 3.5 cm mass extending across corpus callosum and involving both periventricular frontal horn regions, diagnosed as "butterfly" glioblastoma or possible PCNSL. No consideration of infection was raised, care was withdrawn. A 44-year-old woman with systemic lupus erythematous (SLE) treated with prednisone presented with fever and generalized malaise with rapid progression to agitation and confusion. Infection was suspected, but never confirmed on extensive premortem workup. Brain autopsy in both patients revealed severe necrotizing toxoplasmosis virtually confined to periventricular regions. In the first case, necrosis extended across the corpus callosum. Large numbers of organisms were found microscopically, reflecting their immunocompromised, and untreated, status. Cerebral toxoplasmosis should be included in the differential diagnosis when encountering patients with necrotizing ventriculitis, even in the non-AIDS immunosuppressed population.