RESUMEN
Differential diagnosis of dementia remains a challenge in neurology due to symptom overlap across etiologies, yet it is crucial for formulating early, personalized management strategies. Here, we present an artificial intelligence (AI) model that harnesses a broad array of data, including demographics, individual and family medical history, medication use, neuropsychological assessments, functional evaluations and multimodal neuroimaging, to identify the etiologies contributing to dementia in individuals. The study, drawing on 51,269 participants across 9 independent, geographically diverse datasets, facilitated the identification of 10 distinct dementia etiologies. It aligns diagnoses with similar management strategies, ensuring robust predictions even with incomplete data. Our model achieved a microaveraged area under the receiver operating characteristic curve (AUROC) of 0.94 in classifying individuals with normal cognition, mild cognitive impairment and dementia. Also, the microaveraged AUROC was 0.96 in differentiating the dementia etiologies. Our model demonstrated proficiency in addressing mixed dementia cases, with a mean AUROC of 0.78 for two co-occurring pathologies. In a randomly selected subset of 100 cases, the AUROC of neurologist assessments augmented by our AI model exceeded neurologist-only evaluations by 26.25%. Furthermore, our model predictions aligned with biomarker evidence and its associations with different proteinopathies were substantiated through postmortem findings. Our framework has the potential to be integrated as a screening tool for dementia in clinical settings and drug trials. Further prospective studies are needed to confirm its ability to improve patient care.
Asunto(s)
Encéfalo , Angiopatía Amiloide Cerebral , Demencia , Encefalitis Límbica , Neoplasias Pulmonares , Carcinoma Pulmonar de Células Pequeñas , Anciano , Femenino , Humanos , Encéfalo/diagnóstico por imagen , Encéfalo/patología , Demencia/diagnóstico por imagen , Demencia/etiología , Demencia/patología , Diagnóstico Diferencial , Progresión de la Enfermedad , Resultado Fatal , Imagen por Resonancia Magnética , Encefalitis Límbica/diagnóstico por imagen , Encefalitis Límbica/etiología , Carcinoma Pulmonar de Células Pequeñas/complicaciones , Carcinoma Pulmonar de Células Pequeñas/patología , Neoplasias Pulmonares/complicaciones , Neoplasias Pulmonares/patología , Angiopatía Amiloide Cerebral/complicaciones , Angiopatía Amiloide Cerebral/patologíaRESUMEN
Objective This study aims to investigate the association between specific imaging parameters, namely, the Evans index (EI) and ventricular volume (VV), and the variation in gait speed observed in patients with idiopathic normal pressure hydrocephalus (iNPH) before and after cerebrospinal fluid (CSF) removal/lumbar drain (LD). Furthermore, it seeks to identify which imaging parameters are the most reliable predictors for significant improvements in gait speed post procedure. Methods In this retrospective analysis, the study measured the gait speed of 35 patients diagnosed with idiopathic normal pressure hydrocephalus (iNPH) before and after they underwent CSF removal. Before lumbar drain (LD), brain images were segmented to calculate the Evans index and ventricular volume. The study explored the relationship between these imaging parameters (the Evans index and ventricular volume) and the improvement in gait speed following CSF removal. Patients were divided into two categories based on the degree of improvement in gait speed, and we compared the imaging parameters between these groups. Receiver operating characteristic (ROC) curve analysis was employed to determine the optimal imaging parameter thresholds predictive of gait speed enhancement. Finally, the study assessed the predictive accuracy of these thresholds for identifying patients likely to experience improved gait speed post-LD. Results Following CSF removal/lumbar drain, the participants significantly improved in gait speed, as indicated by a paired sample t-test (p-value = 0.0017). A moderate positive correlation was observed between the imaging parameters (EI and VV) and the improvement in gait speed post-LD. Significant differences were detected between the two patient groups regarding EI, VV, and a composite score (statistical test value = 3.1, 2.8, and 2.9, respectively; p-value < 0.01). Receiver operating characteristic (ROC) curve analysis identified the optimal thresholds for the EI and VV to be 0.39 and 110.78 cm³, respectively. The classification based on these thresholds yielded significant associations between patients displaying favorable imaging parameters and those demonstrating improved gait speed post-LD, with chi-square (χ²) values of 8.5 and 7.1, respectively, and p-values < 0.01. Furthermore, these imaging parameter thresholds had a 74% accuracy rate in predicting patients who would improve post-LD. Conclusion The study demonstrates that ventricle volume and the Evans index can significantly predict gait speed improvement after lumbar drain (LD) in patients with iNPH.
RESUMEN
Differential diagnosis of dementia remains a challenge in neurology due to symptom overlap across etiologies, yet it is crucial for formulating early, personalized management strategies. Here, we present an AI model that harnesses a broad array of data, including demographics, individual and family medical history, medication use, neuropsychological assessments, functional evaluations, and multimodal neuroimaging, to identify the etiologies contributing to dementia in individuals. The study, drawing on 51,269 participants across 9 independent, geographically diverse datasets, facilitated the identification of 10 distinct dementia etiologies. It aligns diagnoses with similar management strategies, ensuring robust predictions even with incomplete data. Our model achieved a micro-averaged area under the receiver operating characteristic curve (AUROC) of 0.94 in classifying individuals with normal cognition, mild cognitive impairment and dementia. Also, the micro-averaged AUROC was 0.96 in differentiating the dementia etiologies. Our model demonstrated proficiency in addressing mixed dementia cases, with a mean AUROC of 0.78 for two co-occurring pathologies. In a randomly selected subset of 100 cases, the AUROC of neurologist assessments augmented by our AI model exceeded neurologist-only evaluations by 26.25%. Furthermore, our model predictions aligned with biomarker evidence and its associations with different proteinopathies were substantiated through postmortem findings. Our framework has the potential to be integrated as a screening tool for dementia in various clinical settings and drug trials, with promising implications for person-level management.
RESUMEN
Acute stroke imaging plays a vital and time-sensitive role in therapeutic decision-making. Current clinical workflows widely use computed tomography (CT) and magnetic resonance (MR) techniques including CT and MR perfusion to estimate the volume of ischemic penumbra at risk for infarction without acute intervention. The use of imaging techniques aimed toward evaluating the metabolic derangements underlying a developing infarct may provide additional information for differentiating the penumbra from benign oligemia and infarct core. The authors review several modalities of metabolic imaging including PET, hydrogen and oxygen spectroscopy, sodium MRI, and pH-weighted MRI.
Asunto(s)
Isquemia Encefálica , Accidente Cerebrovascular Isquémico , Accidente Cerebrovascular , Humanos , Isquemia Encefálica/patología , Oxígeno , Accidente Cerebrovascular/terapia , Imagen por Resonancia Magnética , Análisis Espectral , Tomografía de Emisión de Positrones/métodos , Infarto , Concentración de Iones de HidrógenoRESUMEN
OBJECTIVE: Radiological imaging is pivotal in diagnosing idiopathic normal pressure hydrocephalus (iNPH), given the similarity of its symptoms to other neurodegenerative diseases. We aimed to correlate the Evans index (EI), callosal angle (CA), and the volume of the lateral ventricles measured before cerebrospinal fluid removal with the resultant outcomes in gait response. METHODS: In our retrospective study, we identified 42 patients with a diagnosis of iNPH. These patients underwent gait analysis, imaging, and lumbar puncture. Radiological assessments included measurements of CA EI and lateral ventricular volume. Clinically, we assessed the following 4 gait parameters: cadence, gait speed, stride length, and timed up and go. Change in the 4 gait parameters was calculated, normalized, and compiled into a composite score, following which the group was divided into 'responders' and 'nonresponders' based on z score of 0.5. Our dependent variable was clinical improvement in gait, and our independent variables included lateral ventricular volume, EI, and CA. We performed a Wilcoxon rank-sum test to compare significant responder status using CA, EI, and lateral ventricle volume. A receiver operating characteristic analysis was employed to determine which volume measurement exhibited the strongest correlation with responder status. Determining the significant variables, a chi-square analysis was subsequently conducted.A significance threshold was set at P < 0.05. All our statistical evaluations were conducted in the Spyder environment, which is compatible with Python 3.10. RESULTS: There was a significant difference for responder status in EI and lateral ventricle volume. Evan index showing a statistic of 2.202 (P value = 0.02) and lateral ventricle volume demonstrating a statistic of 2.086 (P value = 0.03). Subsequent exploration using receiver operating characteristic analysis, with area under the curve of 0.71, identified 105.40 cm3 as the most robustly correlated volume threshold with responder status. CONCLUSIONS: The lateral ventricular volume demonstrates a stronger correlation with gait improvement compared to the CA or EI. These observations indicate that evaluating the lateral ventricle volume before lumbar puncture could serve as a predictor for gait response after lumbar puncture in individuals with normal pressure hydrocephalus.
RESUMEN
PURPOSE: Radiomics of vertebral bone structure is a promising technique for identification of osteoporosis. We aimed at assessing the accuracy of machine learning in identifying physiological changes related to subjects' sex and age through analysis of radiomics features from CT images of lumbar vertebrae, and define its generalizability across different scanners. MATERIALS AND METHODS: We annotated spherical volumes-of-interest (VOIs) in the center of the vertebral body for each lumbar vertebra in 233 subjects who had undergone lumbar CT for back pain on 3 different scanners, and we evaluated radiomics features from each VOI. Subjects with history of bone metabolism disorders, cancer, and vertebral fractures were excluded. We performed machine learning classification and regression models to identify subjects' sex and age respectively, and we computed a voting model which combined predictions. RESULTS: The model was trained on 173 subjects and tested on an internal validation dataset of 60. Radiomics was able to identify subjects' sex within single CT scanner (ROC AUC: up to 0.9714), with lower performance on the combined dataset of the 3 scanners (ROC AUC: 0.5545). Higher consistency among different scanners was found in identification of subjects' age (R2 0.568 on all scanners, MAD 7.232 years), with highest results on a single CT scanner (R2 0.667, MAD 3.296 years). CONCLUSION: Radiomics features are able to extract biometric data from lumbar trabecular bone, and determine bone modifications related to subjects' sex and age with great accuracy. However, acquisition from different CT scanners reduces the accuracy of the analysis.
Asunto(s)
Enfermedades Óseas Metabólicas , Tomografía Computarizada por Rayos X , Humanos , Niño , Tomografía Computarizada por Rayos X/métodos , Vértebras Lumbares/diagnóstico por imagen , Estudios RetrospectivosAsunto(s)
Parestesia , Adulto , Humanos , Masculino , Pierna , Parestesia/etiología , Síndrome de las Piernas Inquietas , BrazoRESUMEN
Background: Accurate and repeatable measurement of high-grade glioma (HGG) enhancing (Enh.) and T2/FLAIR hyperintensity/edema (Ed.) is required for monitoring treatment response. 3D measurements can be used to inform the modified Response Assessment in Neuro-oncology criteria. We aim to develop an HGG volumetric measurement and visualization AI algorithm that is generalizable and repeatable. Methods: A single 3D-Convoluted Neural Network, NS-HGlio, to analyze HGG on MRIs using 5-fold cross validation was developed using retrospective (557 MRIs), multicentre (38 sites) and multivendor (32 scanners) dataset divided into training (70%), validation (20%), and testing (10%). Six neuroradiologists created the ground truth (GT). Additional Internal validation (IV, three institutions) using 70 MRIs, and External validation (EV, single institution) using 40 MRIs through measuring the Dice Similarity Coefficient (DSC) of Enh., Ed. ,and Enh. + Ed. (WholeLesion/WL) tumor tissue and repeatability testing on 14 subjects from the TCIA MGH-QIN-GBM dataset using volume correlations between timepoints were performed. Results: IV Preoperative median DSC Enh. 0.89 (SD 0.11), Ed. 0.88 (0.28), WL 0.88 (0.11). EV Preoperative median DSC Enh. 0.82 (0.09), Ed. 0.83 (0.11), WL 0.86 (0.06). IV Postoperative median DSC Enh. 0.77 (SD 0.20), Ed 0.78. (SD 0.09), WL 0.78 (SD 0.11). EV Postoperative median DSC Enh. 0.75 (0.21), Ed 0.74 (0.12), WL 0.79 (0.07). Repeatability testing; Intraclass Correlation Coefficient of 0.95 Enh. and 0.92 Ed. Conclusion: NS-HGlio is accurate, repeatable, and generalizable. The output can be used for visualization, documentation, treatment response monitoring, radiation planning, intra-operative targeting, and estimation of Residual Tumor Volume among others.
Asunto(s)
Cefalea , Trastornos de la Visión , Adulto , Femenino , Cefalea/etiología , Humanos , Trastornos de la Visión/etiologíaRESUMEN
Tay-Sachs disease (TSD) is an inherited neurological disorder caused by deficiency of hexosaminidase A (HexA). Here, we describe an adeno-associated virus (AAV) gene therapy expanded-access trial in two patients with infantile TSD (IND 18225) with safety as the primary endpoint and no secondary endpoints. Patient TSD-001 was treated at 30 months with an equimolar mix of AAVrh8-HEXA and AAVrh8-HEXB administered intrathecally (i.t.), with 75% of the total dose (1 × 1014 vector genomes (vg)) in the cisterna magna and 25% at the thoracolumbar junction. Patient TSD-002 was treated at 7 months by combined bilateral thalamic (1.5 × 1012 vg per thalamus) and i.t. infusion (3.9 × 1013 vg). Both patients were immunosuppressed. Injection procedures were well tolerated, with no vector-related adverse events (AEs) to date. Cerebrospinal fluid (CSF) HexA activity increased from baseline and remained stable in both patients. TSD-002 showed disease stabilization by 3 months after injection with ongoing myelination, a temporary deviation from the natural history of infantile TSD, but disease progression was evident at 6 months after treatment. TSD-001 remains seizure-free at 5 years of age on the same anticonvulsant therapy as before therapy. TSD-002 developed anticonvulsant-responsive seizures at 2 years of age. This study provides early safety and proof-of-concept data in humans for treatment of patients with TSD by AAV gene therapy.
Asunto(s)
Enfermedad de Tay-Sachs , Anticonvulsivantes , Dependovirus/genética , Terapia Genética , Humanos , Enfermedad de Tay-Sachs/genética , Enfermedad de Tay-Sachs/terapiaRESUMEN
Immune checkpoint inhibitors are monoclonal antibodies directed against cellular pathways on T-cells to treat different types of malignancies. This new therapy can cause immune-related adverse events that can involve almost any organ system. This article will review clinical presentations, molecular mechanisms and imaging manifestations of adverse events caused by checkpoint inhibitors and also illustrate the pseudoprogression tumor response pattern.
Asunto(s)
Antineoplásicos Inmunológicos/efectos adversos , Inmunoterapia/efectos adversos , Radiólogos/educación , Anticuerpos Monoclonales , Antígeno CTLA-4 , Humanos , Neoplasias/tratamiento farmacológicoRESUMEN
Routine non-contrast material-enhanced head CT is one of the most frequently ordered studies in the emergency department. Skull base-related pathologic entities, often depicted on the first or last images of a routine head CT study, can be easily overlooked in the emergency setting if not incorporated in the interpreting radiologist's search pattern, as the findings can be incompletely imaged. Delayed diagnosis, misdiagnosis, or lack of recognition of skull base pathologic entities can negatively impact patient care. This article reviews and illustrates the essential skull base anatomy and common blind spots that are important to radiologists who interpret nonenhanced head CT images in the acute setting. The imaging characteristics of important "do not miss" lesions are emphasized and categorized by their cause and location within the skull base, and the potential differential diagnoses are discussed. An interpretation checklist to improve diagnostic accuracy is provided. ©RSNA, 2019.
Asunto(s)
Traumatismos Craneocerebrales/diagnóstico por imagen , Base del Cráneo/diagnóstico por imagen , Tomografía Computarizada por Rayos X/métodos , Lista de Verificación , Trastornos de la Conciencia/diagnóstico por imagen , Enfermedades de los Nervios Craneales/diagnóstico por imagen , Urgencias Médicas , Dolor Facial/diagnóstico por imagen , Cefalea/diagnóstico por imagen , Humanos , Hipertensión Intracraneal/diagnóstico por imagen , Órbita/diagnóstico por imagen , Convulsiones/diagnóstico por imagen , Base del Cráneo/anatomía & histología , Base del Cráneo/lesiones , Hueso Temporal/diagnóstico por imagen , Traumatismos del Sistema Nervioso/diagnóstico por imagen , Trastornos de la Visión/diagnóstico por imagenAsunto(s)
Conducta de Elección , Pautas de la Práctica en Medicina/estadística & datos numéricos , Radiografía Abdominal , Tomografía Computarizada por Rayos X , Procedimientos Innecesarios/estadística & datos numéricos , Adhesión a Directriz , Investigación sobre Servicios de Salud , Humanos , Guías de Práctica Clínica como Asunto , Sociedades Médicas , Estados UnidosRESUMEN
OBJECTIVE: The objective of this study was to evaluate the proportion of Emergency Department (ED) radiology examinations ordered or interpreted prior to a documented clinical assessment. MATERIALS AND METHODS: We collected 600 retrospective consecutive ED cases consisting equally of patients whose first ED imaging examination was computed tomography (CT), radiography (XR), or ultrasonography (US). For each patient, the following times were documented: ED arrival, ED departure, ED length of stay (LOS), imaging order entry, image availability, radiology report availability, triage note, ED provider note, and laboratory results. RESULTS: Mean age was 44.2, 66.5% female, and mean ED LOS was 326.2 min. ED LOS was longer for patients who received CT versus XR (343.9 vs. 311.3; p = 0.029). In 25.5% of XR, 10% of CT, and 8% of US cases, the imaging exam was completed before the ED provider note was started. In 20.5% of XR, 6.5% of CT, and 6% of US cases, the radiologist did not have the ED provider note available prior to completing their diagnostic interpretation. In 33.4% of all cases and 57.5% of XR cases, incomplete clinical documentation (triage note, provider note, lab results) was available during radiology report creation. CT and US exams more frequently had clinical data available prior to radiologist interpretation than XR (p < 0.0001). Radiologist turn-around-time was unaffected by clinical information availability. CONCLUSION: Eight percent of ED CT and 10% of ED US examinations were ordered and completed before documented clinical assessment. Thirty-three percent had incomplete clinical assessment performed prior to image interpretation. Further investigation is needed to determine impact on interpretation accuracy.
Asunto(s)
Diagnóstico por Imagen , Documentación , Servicio de Urgencia en Hospital , Sistemas de Entrada de Órdenes Médicas , Adulto , Femenino , Humanos , Tiempo de Internación/estadística & datos numéricos , Masculino , Estudios Retrospectivos , Factores de Tiempo , TriajeRESUMEN
Traditionally the presence of fat in closed spinal dysraphism has been referred to as spinal lipoma. Recent reports suggest that these spinal lesions are better described as spinal hamartomas due to the unencapsulated and infiltrating nature of the fat and presence of other heterotopic soft tissue. The presence of ossified bone in spinal hamartomas referred to as ossified dysraphic hamartoma, is extremely rare with only three case reports in literature, all associated with lipomyeloceles, none with lipomyelomeningoceles. We present three cases of ossified dysraphic hamartoma, two of them associated with lipomyelomeningoceles, reviewing imaging features on CT and MRI.
Asunto(s)
Huesos/patología , Hamartoma/diagnóstico , Lipoma , Meningomielocele/patología , Osificación Heterotópica , Enfermedades de la Médula Espinal/diagnóstico , Enfermedades de la Columna Vertebral/diagnóstico , Tejido Adiposo , Femenino , Hamartoma/diagnóstico por imagen , Humanos , Lactante , Lipoma/diagnóstico por imagen , Imagen por Resonancia Magnética , Meningomielocele/diagnóstico por imagen , Radiografía , Médula Espinal/patología , Enfermedades de la Médula Espinal/diagnóstico por imagen , Enfermedades de la Columna Vertebral/diagnóstico por imagen , Disrafia Espinal , Columna Vertebral/patología , Tomografía Computarizada por Rayos XRESUMEN
In this review, we discuss the clinical and radiologic findings of small bowel diverticulosis, which is infrequently encountered during practice and far less common than colonic diverticulosis. Small bowel diverticulosis can present with a range of emergent symptomatic complications including diverticulitis, perforation, or hemorrhage. Here, we focus on the clinical features, pathogenesis, radiologic findings, and treatment of small bowel diverticulitis. Although not routinely considered in the differential diagnosis of an acute abdomen, prospective radiologic diagnosis of small bowel diverticulitis is important and can lead to conservative treatment thus preventing unnecessary exploratory laparotomy.
Asunto(s)
Diagnóstico por Imagen , Diverticulitis/diagnóstico por imagen , Intestino Delgado , Diagnóstico Diferencial , Diverticulitis/epidemiología , Diverticulitis/terapia , HumanosRESUMEN
OBJECTIVE: The objective of the present study is to examine the concordance of facial fracture classifications in patients with trauma who underwent surgery and to assess the epidemiologic findings associated with facial trauma. MATERIALS AND METHODS: Patients with trauma who underwent facial CT examination and inpatient operative intervention during a 1-year period were retrospectively analyzed. Patient demographic characteristics, the mechanism of injury, the radiology report, the surgical diagnosis, and clinical indications were reviewed. Fractures were documented according to bone type and were classified into the following subtypes: LeFort 1, LeFort 2, LeFort 3, naso-orbital-ethmoidal, zygomaticomaxillary complex (ZMC), orbital, and mandibular. Concordance between the radiology and surgery reports was assessed. RESULTS: A total of 115,000 visits to the emergency department resulted in 9000 trauma activations and 3326 facial CT examinations. One hundred fifty-six patients (4.7%) underwent facial surgical intervention, and 133 cases met criteria for inclusion in the study. The mean injury severity score was 10.2 (range, 1-75). The three most frequently noted injury mechanisms were as follows: assault (77 cases [57.9%]), a traffic accident (21 cases [15.8%]), and a fall (20 cases [15%]). The three most frequently noted facial bone fractures were as follows: mandible (100 cases [75.2%]), maxilla (53 cases [39.8%]), and orbit (53 cases [39.8%]). The five descriptors most frequently found in the radiology and surgery reports were the mandibular angle (25 cases), the orbital floor (25 cases), the mandibular parasymphysis (22 cases), the mandibular body (21 cases), and ZMC fractures (19 cases). A classification was not specified in 31 of the radiologic impressions (22.5%), with 28 of 31 radiologists expecting the surgeon to read the full report. The descriptors used in the radiology and surgery reports matched in 73 cases (54.9%) and differed in 51 cases (38.3%). No classifications were used by one or both specialties in nine cases (6.8%). CONCLUSION: For 38.3% of patients needing facial surgery, descriptors used in the radiologic and surgery reports differed. Speaking a common language can potentially improve communication between the radiology and surgery services and can help expedite management of cases requiring surgery.
Asunto(s)
Traumatismos Faciales/diagnóstico por imagen , Traumatismos Faciales/cirugía , Radiólogos , Fracturas Craneales/diagnóstico por imagen , Fracturas Craneales/cirugía , Cirujanos , Tomografía Computarizada por Rayos X , Adolescente , Adulto , Anciano , Servicio de Urgencia en Hospital , Femenino , Humanos , Imagenología Tridimensional , Masculino , Persona de Mediana Edad , Estudios RetrospectivosRESUMEN
Complications related to endoscopy are commonly encountered in the emergency department (ED) due to an increased use of outpatient diagnostic and therapeutic upper gastrointestinal endoscopic procedures. A majority of these procedures are performed on an outpatient basis, and patients with post-procedural symptoms may return to the ED. Since these patients often undergo computed tomography (CT) for diagnosis of post-procedure complications, the emergency radiologist should be familiar with the spectrum of expected post-procedural findings, as well as common and rare complications. We present a pictorial review of post-endoscopy complications and review imaging protocols in different clinical scenarios.