Jefferson William Chen MD

Dr. Jefferson William Chen MD

orcid

Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
26 July 2018 in American Journal of Neuroradiology
Verified
Journal Article
Open access
PDF file
PCP.D. Chang
P.D. Chang
EKE. Kuoy
E. Kuoy
JGjack grinband
jack grinband
Brent WeinbergBrent Weinberg
Brent Weinberg
MTM. Thompson
M. Thompson
DCDaniel Chow
Daniel Chow
BACKGROUND AND PURPOSE: Convolutional neural networks are a powerful technology for image recognition. This study evaluates a convolutional neural network optimized for the detection and quantification of intraparenchymal, epidural/subdural, and subarachnoid hemorrhages on noncontrast CT. MATERIALS AND METHODS: This study was performed in 2 phases. First, a training cohort of all NCCTs acquired at a single institution between January 1, 2017, and July 31, 2017, was used to develop and cross-validate a custom hybrid 3D/2D mask ROI-based convolutional neural network architecture for hemorrhage evaluation. Second, the trained network was applied prospectively to all NCCTs ordered from the emergency department between February 1, 2018, and February 28, 2018, in an automated inference pipeline. Hemorrhage-detection accuracy, area under the curve, sensitivity, specificity, positive predictive value, and negative predictive value were assessed for full and balanced datasets and were further stratified by hemorrhage type and size. Quantification was assessed by the Dice score coefficient and the Pearson correlation. RESULTS: A 10,159-examination training cohort (512,598 images; 901/8.1% hemorrhages) and an 862-examination test cohort (23,668 images; 82/12% hemorrhages) were used in this study. Accuracy, area under the curve, sensitivity, specificity, positive predictive value, and negative-predictive value for hemorrhage detection were 0.975, 0.983, 0.971, 0.975, 0.793, and 0.997 on training cohort cross-validation and 0.970, 0.981, 0.951, 0.973, 0.829, and 0.993 for the prospective test set. Dice scores for intraparenchymal hemorrhage, epidural/subdural hemorrhage, and SAH were 0.931, 0.863, and 0.772, respectively. CONCLUSIONS: A customized deep learning tool is accurate in the detection and quantification of hemorrhage on NCCT. Demonstrated high performance on prospective NCCTs ordered from the emergency department suggests the clinical viability of the proposed deep learning tool.

6 May 2021 in Frontiers in Neurology
Verified
Journal Article
Open access
PDF file
JMJefferson William Chen MD
Jefferson William Chen MD
JXJordan C. Xu
Jordan C. Xu
DMDennis Malkasian
Dennis Malkasian
MPMari A. Perez-Rosendahl
Mari A. Perez-Rosendahl
DTDiem Kieu Tran
Diem Kieu Tran
Background: Chronic subdural hematomas (cSDH) are increasingly prevalent worldwide with the increased aging population and anticoagulant use. Different surgical, medical, and endovascular treatments have had varying success rates. Primary neurosurgical interventions include burr hole drainage of the cSDH and mini-craniotomies/craniotomies with or without fenestration of the inner membrane. A key assessment of the success or failure of cSDH treatments has been symptomatic recurrence rates which have historically ranged from 5 to 30%. Pre-operative prediction of the inner subdural membrane by CT scan was used to guide our decision to perform mini-craniotomies. Release of the inner membrane facilitates the expansion of the brain and likely improves glymphatic flow.Methods: Consecutive mini-craniotomies (N = 34) for cSDH evacuation performed by a single neurosurgeon at a quaternary academic medical center/Level I trauma center from July 2018-September 2020 were retrospectively reviewed. Patient characteristics [age, gender, presenting GCS, GOS, initial CTs noting the inner subdural membrane, midline shift (MLS), cSDH width, inner membrane fenestration, cSDH recurrence, post-operative seizures, infections, length of stay] were extracted from the EMR.Results: Twenty nine patients had mini-craniotomies as primary treatment of the cSDH. Mean age = 68.9 ± 19.7 years (range 22–102), mean pre-operative GCS = 14.5 ± 1.1, mean MLS = 6.75 ± 4.2 mm, and mean maximum thickness of cSDH = 17.7 ± 6.0 mm. Twenty four were unilateral, five bilateral, 34 total craniotomies were performed. Thirty three had inner membrane signs on pre-operative head CTs and an inner subdural membrane was fenestrated in all cases except for the one craniotomy that didn't show these characteristic CT findings. Mean operating time = 79.5 ± 26.0 min. Radiographic and clinical improvement occurred in all patients. Mean improvement in MLS = 3.85 ± 2.69. There were no symptomatic recurrences, re-operations, surgical site infections, or deaths during the 6 months of follow-up. One patient was treated for post-operative seizures with AEDs for 6 months.Conclusion: Pre-operative CT scans demonstrating inner subdural membranes may guide one to target the treatment to allow release of this tension band. Mini-craniotomy with careful fenestration of the inner membrane is very effective for this. Brain re-expansion and re-establishment of normal brain interstitial flow may be important in long term outcomes with cSDH and may be related to the recent interests in brain glymphatics and dural lymphatics.

10 January 2022 in BME Frontiers
Verified
Journal Article
Open access
PDF file
AZAngela Zhang
Angela Zhang
AKAmil Khan
Amil Khan
SMSaisidharth Majeti
Saisidharth Majeti
JPJudy Pham
Judy Pham
CNChristopher Nguyen
Christopher Nguyen
BMB. S. Manjunath
B. S. Manjunath
Objective and Impact Statement . We propose an automated method of predicting Normal Pressure Hydrocephalus (NPH) from CT scans. A deep convolutional network segments regions of interest from the scans. These regions are then combined with MRI information to predict NPH. To our knowledge, this is the first method which automatically predicts NPH from CT scans and incorporates diffusion tractography information for prediction. Introduction . Due to their low cost and high versatility, CT scans are often used in NPH diagnosis. No well-defined and effective protocol currently exists for analysis of CT scans for NPH. Evans’ index, an approximation of the ventricle to brain volume using one 2D image slice, has been proposed but is not robust. The proposed approach is an effective way to quantify regions of interest and offers a computational method for predicting NPH. Methods . We propose a novel method to predict NPH by combining regions of interest segmented from CT scans with connectome data to compute features which capture the impact of enlarged ventricles by excluding fiber tracts passing through these regions. The segmentation and network features are used to train a model for NPH prediction. Results . Our method outperforms the current state-of-the-art by 9 precision points and 29 recall points. Our segmentation model outperforms the current state-of-the-art in segmenting the ventricle, gray-white matter, and subarachnoid space in CT scans. Conclusion . Our experimental results demonstrate that fast and accurate volumetric segmentation of CT brain scans can help improve the NPH diagnosis process, and network properties can increase NPH prediction accuracy.

9 May 2022 in World Neurosurgery
Verified
Journal Article
Closed access
JPJudy Pham
Judy Pham
ATAntalique Tran
Antalique Tran
KOKevin S. O'LEary
Kevin S. O'LEary
JYJulie Youm
Julie Youm
DTDiem Kieu Tran
Diem Kieu Tran
JMJefferson William Chen MD
Jefferson William Chen MD

1 February 2013 in The Journal of Emergency Medicine
Verified
Journal Article
Closed access
JHJonathan G. Hill
Jonathan G. Hill
SHSavanna J. Hardekopf
Savanna J. Hardekopf
JMJefferson William Chen MD
Jefferson William Chen MD
JKJames C. Krieg
James C. Krieg
RBRaymond B. Bracis
Raymond B. Bracis
RERichard F. Edlich
Richard F. Edlich

27 December 2019 in Surgical Neurology International
Verified
Journal Article
Open access
JXJordan Xu
Jordan Xu
CPCassie Poole
Cassie Poole
RSRonald Sahyouni
Ronald Sahyouni
JMJefferson William Chen MD
Jefferson William Chen MD
Background: Ventriculoperitoneal shunts (VPSs) have been the mainstay of treating hydrocephalus since the 1950s. However, shunts have a reported complication rate reaching nearly 50%. Devices have been developed that utilize noninvasive thermal transcutaneous diffusion technology. These shunt evaluation devices measure temperature gradients to detect shunt cerebrospinal fluid flow. We assessed the utility using a thermal diffusion technique to work up shunt failure in the emergency room (ER).Methods: This was a retrospective case series at a single medical center ER. We included consecutive patients with possible VPS malfunction who were evaluated with a thermal sensor. The time period of data collection included September 2015–April 2018.Results: Eight patients were reviewed, including four males and four females. The average age of reviewed patients was 35.1 (+/− ; 16.5). In our case series, three patients had adequate shunt flow as assessed by the shunt evaluation device, and the decision was made to discharge the patient and follow-up in clinic. In two patients, the flow was diminished, but due to other reassuring signs, the patients were still discharged with follow-up. Two patients were noted to have poor flow and were admitted for shunt revision.Conclusion: It is important to determine whether a malfunction is present and whether an intervention is necessary for patients who present to the emergency department with possible symptoms of shunt failure. A thermal sensor is a fast and noninvasive tool in the evaluation of shunt flow and helps determine whether it is safe to send a patient home or intervene appropriately.