Try a new search

Format these results:

Searched for:



Total Results:


Artificial Intelligence Applied to Osteoporosis: A Performance Comparison of Machine Learning Algorithms in Predicting Fragility Fractures From MRI Data

Ferizi, Uran; Besser, Harrison; Hysi, Pirro; Jacobs, Joseph; Rajapakse, Chamith S; Chen, Cheng; Saha, Punam K; Honig, Stephen; Chang, Gregory
BACKGROUND:A current challenge in osteoporosis is identifying patients at risk of bone fracture. PURPOSE/OBJECTIVE:To identify the machine learning classifiers that predict best osteoporotic bone fractures and, from the data, to highlight the imaging features and the anatomical regions that contribute most to prediction performance. STUDY TYPE/METHODS:Prospective (cross-sectional) case-control study. POPULATION/METHODS:. Field Strength/ Sequence: 3D FLASH at 3T. ASSESSMENT/RESULTS:Quantitative MRI outcomes by software algorithms. Mechanical and topological microstructural parameters of the trabecular bone were calculated for five femoral regions, and added to the vector of features together with bone mineral density measurement, fracture risk assessment tool (FRAX) score, and personal characteristics such as age, weight, and height. We fitted 15 classifiers using 200 randomized cross-validation datasets. Statistical Tests: Data: Kolmogorov-Smirnov test for normality. Model Performance: sensitivity, specificity, precision, accuracy, F1-test, receiver operating characteristic curve (ROC). Two-sided t-test, with P < 0.05 for statistical significance. RESULTS:The top three performing classifiers are RUS-boosted trees (in particular, performing best with head data, F1 = 0.64 ± 0.03), the logistic regression and the linear discriminant (both best with trochanteric datasets, F1 = 0.65 ± 0.03 and F1 = 0.67 ± 0.03, respectively). A permutation of these classifiers comprised the best three performers for four out of five anatomical datasets. After averaging across all the anatomical datasets, the score for the best performer, the boosted trees, was F1 = 0.63 ± 0.03 for All-features dataset, F1 = 0.52 ± 0.05 for the no-MRI dataset, and F1 = 0.48 ± 0.06 for the no-FRAX dataset. Data Conclusion: Of many classifiers, the RUS-boosted trees, the logistic regression, and the linear discriminant are best for predicting osteoporotic fracture. Both MRI and FRAX independently add value in identifying osteoporotic fractures. The femoral head, greater trochanter, and inter-trochanter anatomical regions within the proximal femur yielded better F1-scores for the best three classifiers. LEVEL OF EVIDENCE/METHODS:2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018.
PMID: 30252971
ISSN: 1522-2586
CID: 3316002

Response to: The "RACE" national database for recurrent acute rhinosinusitis may need a relook [Letter]

Jacobs, Joseph; Bleier, Benjamin S; Hopkins, Claire; Hwang, Peter; Poetker, David; Schlosser, Rodney; Stewart, Michael; Varshney, Rickul
PMID: 27383827
ISSN: 2042-6984
CID: 2278802

Frontal sinus stenting

Chapter by: Wei, CC; Kanowitz, SJ; Lebowitz, RA; Jacobs, JB
in: The Frontal Sinus by
pp. 393-402
ISBN: 9783662485231
CID: 2567222

Criterion-based (proficiency) training to improve surgical performance

Fried, Marvin P; Kaye, Rachel J; Gibber, Marc J; Jackman, Alexis H; Paskhover, Boris P; Sadoughi, Babak; Schiff, Bradley; Fraioli, Rebecca E; Jacobs, Joseph B
OBJECTIVE To investigate whether training otorhinolaryngology residents to criterion performance levels (proficiency) on the Endoscopic Sinus Surgery Simulator produces individuals whose performance in the operating room is at least equal to those who are trained by performing a fixed number of surgical procedures. DESIGN Prospective cohort. SETTING Two academic medical centers in New York City. PARTICIPANTS Otorhinolaryngology junior residents composed of 8 experimental subjects and 6 control subjects and 6 attending surgeons. INTERVENTION Experimental subjects achieved benchmark proficiency criteria on the Endoscopic Sinus Surgery Simulator; control subjects repeated the surgical procedure twice. MAIN OUTCOME MEASURES Residents completed validated objective tests to assess baseline abilities. All subjects were videotaped performing an initial standardized surgical procedure. Residents were videotaped performing a final surgery. Videotapes were assessed for metrics by an expert panel. RESULTS Attendings outperformed the residents in most parameters on the initial procedure. Experimental and attending groups outperformed controls in some parameters on the final procedure. There was no difference between resident groups in initial performance, but the experimental subjects outperformed the control subjects in navigation in the final procedure. Most important, there was no difference in final performance between subgroups of the experimental group on the basis of the number of trials needed to attain proficiency. CONCLUSIONS Simulator training can improve resident technical skills so that each individual attains a proficiency level, despite the existence of an intrinsic range of abilities. This proficiency level translates to at least equal, if not superior, operative performance compared with that of current conventional training with finite repetition of live surgical procedures.
PMID: 23069788
ISSN: 1538-361x
CID: 2207512

Coronary stent thrombosis in patients undergoing multidigit replantation

Jacobs J; Shah A; Zinn A; Levine J
OBJECTIVE:: The development of drug-eluting stents has decreased the rate of in-stent restenosis. However, there have been reports of late stent thrombosis in patients with drug-eluting stents, especially when dual antiplatelet therapy is interrupted. The high mortality rate associated with cardiac stent thrombosis has led to recent recommendations regarding duration of antiplatelet therapy as well as timing of elective surgery in patients with both drug-eluting stents and bare metal stents. However, in patients requiring emergency operations, delaying surgery is not an option. DATA SOURCES:: In a retrospective review of 65 patients undergoing replantation from 2005 to 2010, only two patients with coronary stents presented, both with drug-eluting stents. Both of these patients developed acute in-stent thrombosis postoperatively on days 5 and 2 despite continuing dual antiplatelet therapy while undergoing multidigit replantation. CONCLUSIONS:: Several factors including large transfusion requirements and the complex pharmacogenetics of clopidogrel may have played a role. These cases bring to light the increasing number of patients with indwelling drug-eluting stents in whom the need for massive surgical or trauma type management will become more frequent
PMID: 21926607
ISSN: 1530-0293
CID: 139042

Measurement of Mycotoxins in Patients with Chronic Rhinosinusitis

Lieberman SM; Jacobs JB; Lebowitz RA; Fitzgerald MB; Crawford J; Feigenbaum BA
Mycotoxins are toxic secondary metabolites produced by a variety of fungi including Aspergillus, Alternaria, and Penicillium species. The presence of mycotoxins in sinonasal tissue and secretions and any possible link to chronic rhinosinusitis (CRS) or other diseases of the head and neck have not been reported. The authors performed an exploratory study to determine the presence and levels of mycotoxins in the sinonasal tissue and secretions of 18 subjects undergoing endoscopic sinus surgery for CRS. Using commercial enzyme-linked immunosorbent assay kits, samples were analyzed for the following mycotoxins: aflatoxin, deoxynivalenol, zearalenone, ochratoxin, and fumonisin. All specimens were negative for aflatoxin, deoxynivalenol, zearalenone, and fumonisin. Four (22%) of 18 specimens were positive for ochratoxin. The clinical significance of this finding remains to be determined
PMID: 21493263
ISSN: 1097-6817
CID: 137919

Measurement of mycotoxins in patients with chronic rhinosinusitis [Meeting Abstract]

Lieberman S.M.; Jacobs J.B.; Lebowitz R.A.; Feigenbaum B.A.
RATIONALE: Rhinosinusitis is one of the most common chronic conditions in the US. The etiology of chronic rhinosinusitis (CRS) remains unknown and controversial. Mycotoxins are toxic secondary metabolites produced by fungi including aspergillus, alternaria, and penicillium species. The presence of mycotoxins in sinonasal tissue and secretions and any possible link to CRS has not been reported. METHODS: Sinonasal tissue and mucus specimens, predominantly from the ethmoid sinuses, were collected from 18 subjects undergoing endoscopic sinus surgery for CRS. The specimens were pulverized and centrifuged, then the resultant supernatant fraction was collected. The following mycotoxins were analyzed using commercial ELISA test kits: aflatoxin, deoxynivalenol, zearalenone, ochratoxin, and fumonisin. Mycotoxin concentrations were quantified from a standard curve. All standards and samples were analyzed in duplicate. We considered a sample positive when the mean value of the sample was two standard deviations above the limit of detection for the test kit. RESULTS: Four (22%) of 18 specimens were positive for ochratoxin. All specimens were negative for aflatoxin, deoxynivalenol, zearalenone, and fumonisin. CONCLUSIONS: Ochratoxin was identified in the sinonasal tissue and/or mucus of some subjects with CRS. The clinical significance of this is not known
ISSN: 0091-6749
CID: 127252

Three-dimensional reconstruction based on computed tomography images of the frontal sinus drainage pathway

Reitzen, S D; Wang, E Y; Butros, S R; Babb, J; Lebowitz, R A; Jacobs, J B
Objective:This study aimed to investigate the utility of three-dimensional reconstructions of paranasal sinus computed tomography data in depicting the anatomy of the frontal sinus drainage pathway.Methods:Twenty-nine patients underwent imaging of the sinuses for various clinical indications. Variations in frontal sinus recess anatomy were determined from 0.75-mm thick coronal, axial and sagittal computed tomography images. Three-dimensional, reformatted images were generated from manually segmented volumes of interest. Observations were made on the variation and usefulness of these reconstructions.Results:Three-dimensional, reformatted images of segmented volumes aided delineation of the spatial relationships of the frontal sinus, frontal sinus drainage pathway, infundibular and meatal direction of drainage, agger nasi cells, ethmoid bulla cells, supraorbital cells, and suprabullar cells.Conclusion:Three-dimensional, reformatted images of frontonasal anatomy enable improved understanding of the frontal sinus drainage pathway anatomy and of the spatial relationships between ethmoid air cells in this region. Such images may provide a useful adjunct to surgical planning and education
PMID: 19922707
ISSN: 0022-2151
CID: 107265

From virtual reality to the operating room: the endoscopic sinus surgery simulator experiment

Fried, Marvin P; Sadoughi, Babak; Gibber, Marc J; Jacobs, Joseph B; Lebowitz, Richard A; Ross, Douglas A; Bent, John P 3rd; Parikh, Sanjay R; Sasaki, Clarence T; Schaefer, Steven D
OBJECTIVE: Establish the feasibility of a predictive validity study in sinus surgery simulation training and demonstrate the effectiveness of the Endoscopic Sinus Surgery Simulator (ES3) as a training device. STUDY DESIGN: Prospective, multi-institutional controlled trial. SETTING: Four tertiary academic centers with accredited otolaryngology-head and neck surgery residency programs. SUBJECTS: Twelve ES3-trained novice residents were compared with 13 control novice residents. METHODS: Subjects were assessed on the performance of basic sinus surgery tasks. Their first in vivo procedure was video recorded and submitted to a blinded panel of independent experts after the panel established a minimum inter-rater reliability of 80 percent. The recordings were reviewed by using a standardized computer-assisted method and customized metrics. Results were analyzed with the Mann-Whitney U test. Internal rater consistency was verified with Pearson moment correlation. RESULTS: Completion time was significantly shorter in the experimental group (injection P = 0.003, dissection P < 0.001), which, according to the rater panel, also demonstrated higher confidence (P = 0.009), demonstrated skill during instrument manipulation (P = 0.011), and made fewer technical mistakes during the injection task (P = 0.048) compared with the control group. The raters' post hoc internal consistency was deemed adequate (r > 0.5 between serial measurements). CONCLUSION: The validity of the ES3 as an effective surgical trainer was verified in multiple instances, including those not depending on subjective rater evaluations. The ES3 is one of the few virtual reality simulators with a comprehensive validation record. Advanced simulation technologies need more rapid implementation in otolaryngology training, as they present noteworthy potential for high-quality surgical education while meeting the necessity of patient safety
PMID: 20115975
ISSN: 1097-6817
CID: 126560

Allergic fungal sinusitis with extensive bone erosion of the clivus presenting with diplopia [Case Report]

Reitzen, S D; Lebowitz, R A; Jacobs, J B
OBJECTIVE: We report a case of allergic fungal sinusitis causing bone erosion and diplopia. CASE REPORT: A 43-year-old man presented with a four-month history of increased nasal congestion and progressive diplopia. Clinical examination revealed bilateral nasal polyposis and a right lateral gaze deficit, consistent with a VIth cranial nerve palsy. Computed tomography of the paranasal sinuses demonstrated a large sellar mass with extensive bony erosion and both supra- and infra-sellar extension. An endoscopic approach to the sphenoid sinus, clivus and posterior cranial fossa with image guidance was performed, enabling surgical treatment involving nasal polypectomy, wide marsupialisation of the sphenoid sinus and removal of the extensive allergic fungal mucin. The patient awoke from anaesthesia with complete resolution of his diplopia. CONCLUSION: Otolaryngologists should be aware that approximately 20 per cent of patients with allergic fungal sinusitis demonstrate paranasal sinus expansion and bone erosion involving surrounding anatomical structures. Such patients may have clinical findings involving the orbit and cranial vault
PMID: 18761768
ISSN: 0022-2151
CID: 100473