Mixed-dictionary models and variational inference in task fMRI for shorter scans and better image quality.

The propensity of task-based functional magnetic resonance imaging (T-fMRI) to large physiological fluctuations, measurement noise, and imaging artifacts entail longer scans and higher temporal resolution (trading off spatial resolution) to alleviate the effects of degradation. This paper focuses on methods towards reducing scan times and enabling higher spatial resolution in T-fMRI. We propose a novel mixed-dictionary model combining (i) the task-based design matrix, (ii) a learned dictionary from resting-state fMRI, and (iii) an analytically-defined wavelet frame. For model fitting, we propose a novel adaptation of the inference framework relying on variational Bayesian expectation maximization with nested minorization. We leverage the mixed-dictionary model coupled with variational inference to enable 2×shorter scan times in T-fMRI, improving activation-map estimates towards the same quality as those resulting from longer scans. We also propose a scheme with potential to increase spatial resolution through temporally undersampled acquisition. Results on motor-task fMRI and gambling-task fMRI show that our framework leads to improved activation-map estimates over the state of the art.

R-fMRI reconstruction from k-t undersampled data using a subject-invariant dictionary model and VB-EM with nested minorization.

Higher spatial resolution in resting-state functional magnetic resonance imaging (R-fMRI) can give reliable information about the functional networks in the cerebral cortex. Typical methods can achieve higher spatial or temporal resolution by speeding up scans using either (i) complex pulse-sequence designs or (ii) k-space undersampling coupled with priors on the signal. We propose to undersample the R-fMRI acquisition in k-space and time to speedup scans in order to improve spatial resolution. We propose a novel model-based R-fMRI reconstruction framework using a robust, subject-invariant, spatially regularized dictionary prior on the signal. Furthermore, we propose a novel inference framework based on variational Bayesian expectation maximization with nested minorization (VB-EM-NM). Our inference framework allows us to provide an estimate of uncertainty of the reconstruction, unlike typical reconstruction methods. Empirical evaluation of (i) simulated R-fMRI reconstruction and (ii) functional-network estimates from brain R-fMRI reconstructions demonstrate that our framework improves over the state of the art, and, additionally, enables significantly higher spatial resolution.

Evaluation of weather-based rice yield models in India.

The objective of this study was to compare two different rice simulation models--standalone (Decision Support System for Agrotechnology Transfer [DSSAT]) and web based (SImulation Model for RIce-Weather relations [SIMRIW])--with agrometeorological data and agronomic parameters for estimation of rice crop production in southern semi-arid tropics of India. Studies were carried out on the BPT5204 rice variety to evaluate two crop simulation models. Long-term experiments were conducted in a research farm of Acharya N G Ranga Agricultural University (ANGRAU), Hyderabad, India. Initially, the results were obtained using 4 years (1994-1997) of data with weather parameters from a local weather station to evaluate DSSAT simulated results with observed values. Linear regression models used for the purpose showed a close relationship between DSSAT and observed yield. Subsequently, yield comparisons were also carried out with SIMRIW and DSSAT, and validated with actual observed values. Realizing the correlation coefficient values of SIMRIW simulation values in acceptable limits, further rice experiments in monsoon (Kharif) and post-monsoon (Rabi) agricultural seasons (2009, 2010 and 2011) were carried out with a location-specific distributed sensor network system. These proximal systems help to simulate dry weight, leaf area index and potential yield by the Java based SIMRIW on a daily/weekly/monthly/seasonal basis. These dynamic parameters are useful to the farming community for necessary decision making in a ubiquitous manner. However, SIMRIW requires fine tuning for better results/decision making.

Segmentation of left ventricle in short-axis echocardiographic sequences by weighted radial edge filtering and adaptive recovery of dropout regions.

In this paper, we present a weighted radial edge filtering algorithm with adaptive recovery of dropout regions for the semi-automatic delineation of endocardial contours in short-axis echocardiographic image sequences. The proposed algorithm requires minimal user intervention at the end diastolic frame of the image sequence for specifying the candidate points of the contour. The region of interest is identified by fitting an ellipse in the region defined by the specified points. Subsequently, the ellipse centre is used for originating the radial lines for filtering. A weighted radial edge filter is employed for the detection of edge points. The outliers are corrected by global as well as local statistics. Dropout regions are recovered by incorporating the important temporal information from the previous frame by means of recursive least squares adaptive filter. This ensures fairly accurate segmentation of the cardiac structures for further determination of the functional cardiac parameters. The proposed algorithm was applied to 10 data-sets over a full cardiac cycle and the results were validated by comparing computer-generated boundaries to those manually outlined by two experts using Hausdorff distance (HD) measure, radial mean square error (rmse) and contour similarity index. The rmse was 1.83 mm with a HD of 5.12 ± 1.21 mm. We have also compared our results with two existing approaches, level set and optical flow. The results indicate an improvement when compared with ground truth due to incorporation of temporal clues. The weighted radial edge filtering algorithm in conjunction with adaptive dropout recovery offers semi-automatic segmentation of heart chambers in 2D echocardiography sequences for accurate assessment of global left ventricular function to guide therapy and staging of the cardiovascular diseases.

Application of crisp and fuzzy clustering algorithms for identification of hidden patterns from plethysmographic observations on the radial pulse.

Radial Pulse forms the most basic and essential physical sign in clinical medicine. The paper proposes the application of crisp and fuzzy clustering algorithms under supervised and unsupervised learning scenarios for identifying non-trivial regularities and relationships of the radial pulse patterns obtained by using the Impedance Plethysmographic technique. The objective of our paper is to unearth the hidden patterns to capture the physiological variabilities from the arterial pulse for clinical analysis, thus providing a very useful tool for disease characterization. A variety of fuzzy algorithms including Gustafson-Kessel (GK) and Gath-Geva (GG)have been intensively tested over a diverse group of subjects and over 4855 data sets. Exhaustive testing over the data set show that about 80 % of the patterns are successfully classified thus providing promising results. A Rank Index of 0.7739 is obtained under supervised learning, which provides an excellent conformity of our process with the results of plethysmographic experts. A correlation of the patterns with the diseases of heart, liver and lungs is judiciously performed.

Head rotation evoked tinnitus due to superior semicircular canal dehiscence.

INTRODUCTION: Superior semicircular canal dehiscence affects the auditory and vestibular systems due to a partial defect in the canal's bony wall. In most cases, sound- and pressure-induced vertigo are present, and are sometimes accompanied by pulse-synchronous tinnitus. CASE PRESENTATION: We describe a 50-year-old man with superior semicircular canal dehiscence whose only complaints were head rotation induced tinnitus and autophony. Head rotation in the plane of the right semicircular canal with an angular velocity exceeding 600 degrees/second repeatedly induced a 'cricket' sound in the patient's right ear. High resolution temporal bone computed tomography changes, and an elevated umbo velocity, supported the diagnosis of superior semicircular canal dehiscence. CONCLUSION: In addition to pulse-synchronous or continuous tinnitus, head rotation induced tinnitus can be the only presenting symptom of superior semicircular canal dehiscence without vestibular complaints. We suggest that, in our patient, the bony defect of the superior semicircular canal ('third window') might have enhanced the flow of inner ear fluid, possibly producing tinnitus.

Isolated fracture of the manubrium of the malleus.

OBJECTIVE: To describe a series of five patients with isolated fracture of the manubrium of the malleus. DESIGN: Retrospective case series. SUBJECTS: Five patients aged 44-64 years with isolated fracture of the manubrium who presented to our institution over a five-year period (2000-2005). RESULTS: All patients presented with a history of digitally manipulating the external auditory canal, leading to the manubrial fracture, which we presume was due to a suction-type mechanism. Otomicroscopy often revealed a break in the smooth contour of the manubrium. All patients had air-bone gaps on audiometry, especially at higher frequencies. Tympanometry showed hypermobility of the tympanic membrane in four patients who were tested. Laser-Doppler vibrometry revealed increased umbo velocity in four out of five patients. Four patients were treated conservatively. One patient underwent exploratory tympanotomy with successful ossiculoplasty. CONCLUSIONS: Isolated fracture of the manubrium is a rare condition which may present as sudden-onset hearing loss after digital manipulation of the external auditory canal. The diagnosis can be made on the basis of otomicroscopy, audiometry, tympanometry and laser-Doppler vibrometry. Conservative treatment is often successful.

Testing a method for quantifying the output of implantable middle ear hearing devices.

This report describes tests of a standard practice for quantifying the performance of implantable middle ear hearing devices (also known as implantable hearing aids). The standard and these tests were initiated by the Food and Drug Administration of the United States Government. The tests involved measurements on two hearing devices, one commercially available and the other home built, that were implanted into ears removed from human cadavers. The tests were conducted to investigate the utility of the practice and its outcome measures: the equivalent ear canal sound pressure transfer function that relates electrically driven middle ear velocities to the equivalent sound pressure needed to produce those velocities, and the maximum effective ear canal sound pressure. The practice calls for measurements in cadaveric ears in order to account for the varied anatomy and function of different human middle ears.

A decrease in effective diameter of rat mesenteric venules due to leukocyte margination after a bolus injection of pentoxifylline--digital image analysis of an intravital microscopic observation.

The ability of leukocytes to adhere to endothelial cells (EC) and then to migrate out of the blood stream into tissues enable them to perform their surveillance functions. Adhesion of leukocytes to EC is, however, only possible if the cells have marginated as a result of rheological interaction with other blood cells in flow. Using Pentoxifylline (PTX), a rheologically active drug, to manipulate this interaction, we have imaged and quantified this margination phenomenon in vivo. A system has been developing to perform this imaging via an intravital microscope connected to an image processing system. Albino rats were anesthetized and cannulated for intravenous bolus injection (0.5 ml) of PTX (1.25 mg/ml) through the femoral vein. A longitudinal incision exposed the mesentery, part of which was observed under microscope to visualize microcirculation. The image of interest was then stored on computer hard drive. Individual leukocyte velocities were determined before and after PTX infusion. The leukocytes, marginating and sticking after PTX infusion either remained attached, constituting the peripheral marginating leukocyte pool in the postcapillary venules, or detached with different step velocities. The reduction in effective venular diameters as a result of leukocyte margination was estimated to be 32-44%. These results demonstrate the biological importance of hemodynamic displacement leading to docking, adhesion, rolling and migration processes of leukocytes in blood.

Adult-onset xeroderma pigmentosum neurological disease--observations in an autopsy case.

Xeroderma pigmentosum (XP) is an inherited disease with defective DNA repair. Patients develop skin cancer because of unrepaired DNA damage produced by the ultraviolet radiation (UV) in sunlight. Many XP children also develop XP neurological disease (ND), consisting of sensorineural hearing loss (SNHL) and a primary neuronal degeneration of the central and peripheral nervous systems. Since the harmful UV in sunlight cannot reach the nervous system, the cause of the death of XP neurons has been hypothesized to result from the inability to repair their DNA that has been damaged by endogenous metabolites. Progressive XP ND originating in an adult has been identified in only a single case. Although clinically asymptomatic at the age of 47 years, the patient had audiometric evidence of a developing mild SNHL together with elicited signs and electrophysiologic evidence of a peripheral neuropathy. She died of metastatic endocervical adenocarcinoma at 49 years of age. We describe here the neuropathological findings in this patient, including examination of the inner ear. Despite clinical evidence of SNHL, there were no anatomic abnormalities of the inner ear. However, the dorsal root ganglia (DRG) showed ongoing neuronal loss. Our findings indicate that XP ND originating in this adult is, like XP ND in children, a primary neuronal degeneration that manifests first in the peripheral nervous system.

Decreasing hair cell counts in aging humans.

Deterioration of balance with advancing age is a well-known fact of life. Some investigators have reported a 50% prevalence of dizziness in the elderly. Clinically, progressive dysequilibrium of aging presents as gradually worsening balance due to age-related decline in function of the peripheral vestibular system, central nervous system, vision, and musculoskeletal system. Vestibular function testing has shown clear evidence of age-related changes in peripheral and central sites. Histopathologic changes in the vestibular sensory organs include progressive hair cell degeneration, otoconial degeneration in the otolith organs, and decreasing number of Scarpa's ganglion neurons. Recently, a new quantitative method of assessing vestibular otopathology has been described, utilizing Nomarski differential interference contrast microscopy. This technique has been applied to 67 human temporal bones of individuals from birth to age 100 to create a normative database of total, type I, and type II hair cell counts as a function of age. Results show a highly significant continuous decrease in all counts from birth to age 100, best fit by a linear regression model. Type I hair cell counts in all three semicircular canal cristae decrease at a similar rate, significantly faster than the degeneration observed in type I hair cells of the maculae. Type II hair cell counts decline at the same rate for all 5 sensory epithelia. These normative data provide the basis for comparisons to hair cell counts made in temporal bones from subjects with known vestibular disorders. They also provide a basis for drawing correlations between vestibular function testing and vestibular otopathology.

Middle-ear function with tympanic-membrane perforations. I. Measurements and mechanisms.

Sound transmission through ears with tympanic-membrane (TM) perforations is not well understood. Here, measurements on human-cadaver ears are reported that describe sound transmission through the middle ear with experimentally produced perforations, which range from 0.5 to 5.0 mm in diameter. Three response variables were measured with acoustic stimulation at the TM: stapes velocity, middle-ear cavity sound pressure, and acoustic impedance at the TM. The stapes-velocity measurements show that perforations cause frequency-dependent losses; at low frequencies losses are largest and increase as perforation size increases. Measurements of middle-ear cavity pressure coupled with the stapes-velocity measurements indicate that the dominant mechanism for loss with TM perforations is reduction in pressure difference across the TM; changes in TM-to-ossicular coupling generally contribute less than 5 dB to the loss. Measurements of middle-ear input impedance indicate that for low frequencies, the input impedance with a perforation approximates the impedance of the middle-ear cavity; as the perforation size increases, the similarity to the cavity's impedance extends to higher frequencies. The collection of results suggests that the effects of perforations can be represented by the path for air-volume flow from the ear canal to the middle-ear cavity. The quantitative description of perforation-induced losses may help clinicians determine, in an ear with a perforation, whether poor hearing results only from the perforation or whether other pathology should be expected.

Middle-ear function with tympanic-membrane perforations. II. A simple model.

A quantitative model of the human middle ear with a tympanic-membrane (TM) perforation is developed. The model is constrained by several types of acoustic measurements made on human cadaver ears, which indicate that perforation-induced changes in transmission result primarily from changes in driving pressure across the TM and that perforation-induced change in the structure of the TM and its coupling to the ossicles contributes a substantially smaller component. The model represents the effect of a perforation on the pressure difference across the TM by inclusion of a path for sound coupling through the perforation from the ear canal to the middle-ear cavity. The model implies that hearing loss with perforations depends primarily on three quantities: the perforation diameter, sound frequency, and the volume of air in the middle-ear cavity. For the conditions that produce the largest hearing loss (low frequency and large perforation), the model yields a simple dependence of loss on frequency, perforation diameter, and middle-ear cavity volume. Predictions from this model may be useful to clinicians in determining whether, in particular cases, hearing losses are explainable by the observed perforations or if additional pathology must be involved.

Histopathology and molecular genetics of hearing loss in the human.

Hearing loss is among the most common disabilities of man. It has been estimated that over 70 million individuals in the world are hearing impaired with pure tone averages greater than 55 dB. A genetic etiology is thought to be responsible for over half of early onset hearing loss and at least one third of late onset hearing loss. In this review, examples of the histopathology of the inner ear in known genetic syndromes in the human will be presented in order to provide a structural basis for understanding molecular mechanisms of development and maintenance in the inner ear, and to serve the essential function of validating the applicability of animal genetic models of hearing loss to the human condition.

Otopathology in a case of type I Waardenburg's syndrome.

We report a case of type I Waardenburg's syndrome that provides insight into the etiopathogenesis of sensorineural hearing loss (SNHL) in this syndrome. The subject, a 76-year-old woman with type I Waardenburg's syndrome (dystopia canthorum, heterochromia irides, and white hair), had congenital low-frequency SNHL in her right ear only, which had remained relatively stable throughout her life. Blood leukocyte DNA studies revealed a PAX-3 mutation with a 1 base pair C-to-A substitution in exon 5 at base 602. Light microscopic studies of the right cochlea showed intact neurosensory structures in only the lower basal turn, with the remainder of the cochlea showing absence of melanocytes, absence of stria vascularis, missing hair cells, dysmorphogenesis of the tectorial membrane, and lack of peripheral processes of the spiral ganglion cells. There was pathological alteration of the vestibular dark cells with marked reduction of melanocytes associated with these dark cells. The left inner ear was normal, with a full complement of neurosensory structures, including melanocytes. Because the PAX-3 gene is involved in neural crest development and melanocytes migrate from the neural crest to the ear, the findings in this case are consistent with the hypothesis that defective melanocyte migration or defective melanocyte function results in defective development of the stria vascularis (and perhaps other structures of the ear), leading to SNHL.

Temporal bone histopathologic and genetic studies in Mohr-Tranebjaerg syndrome (DFN-1).

OBJECTIVE: To describe the temporal bone histopathologic and genetic abnormalities in a case of Mohr-Tranebjaerg syndrome. BACKGROUND: Mohr-Tranebjaezrg syndrome (DFN-1) is an X-linked, recessive, syndromic hearing loss, characterized by postlingual sensorineural hearing loss with onset in childhood, followed in adult life by progressive dystonia, spasticity, dysphagia, and optic atrophy. The syndrome is caused by mutations in the DDP (deafness/dystonia peptide) gene, which are thought to result in mitochondrial dysfunction with subsequent neurodegeneration. The temporal bone pathologic changes in this syndrome have not been reported. METHODS: Hearing loss developed in the patient at age 4, blindness at age 48, and dystonia at age 57. Genetic studies on peripheral blood showed a l51delT mutation in his DDP gene. He died at age 66. The right temporal bone was subjected to light microscopy and polymerase chain reaction-based analysis of the DDP gene sequence. RESULTS: There was near complete loss of spiral ganglion cells with loss of nearly all peripheral and central processes. Only 1,765 spiral ganglion cells remained (8.5% of mean normal for age). The organ of Corti (including hair cells), stria vascularis, and spiral ligament were preserved. There was also a severe loss of Scarpa's ganglion cells with preservation of vestibular hair cells. The population of geniculate and trigeminal ganglion cells appeared normal. Sequence analysis from temporal bone DNA showed the 15ldelT DDP gene mutation. CONCLUSION: Sensorineural hearing loss in Mohr-Tranebjaerg syndrome is the result of a postnatal, progressive, severe auditory neuropathy.

Histologic studies of the posterior stapediovestibular joint in otosclerosis.

OBJECTIVE: To determine the prevalence of ankylosis or otosclerosis at the posterior stapediovestibular joint (SVJ) in temporal bones with otosclerosis, with special reference to stapes surgery. BACKGROUND: Long-term success of the laser stapedotomy minus prosthesis (STAMP) procedure, anterior crurotomy, and similar partial stapedectomy procedures depends on lack of ankylosis and lack of otosclerosis involving the posterior SVJ. Previous work has shown that the air-bone gap in otosclerosis correlates with narrowing and loss of the SVJ space. However, the prevalence and histologic features of otosclerotic involvement of the posterior SVJ space have not been well characterized. METHODS: Histologic assessment of serial sections through the oval window niche in 140 temporal bones with otosclerosis that had been sectioned in the axial plane (age range 20-95 years, mean 68). Bones with stapes mobilization or stapedectomy were excluded. RESULTS AND CONCLUSIONS: Two of 140 bones had otosclerosis exclusively at the posterior SVJ. Of the remaining 138 bones, all of which had otosclerosis at the anterior SVJ, 82 bones also had otosclerosis at the posterior joint. Of the 56 bones without otosclerosis of the posterior joint, there was bony ankylosis of the posterior joint in 3 bones. Thus, 53 bones (38%) had neither ankylosis nor otosclerosis involving the posterior joint, and they would be potentially suitable for a laser STAMP or a similar procedure. There was no correlation between otosclerosis at the posterior SVJ and age, sex, or duration of conductive hearing loss. Otosclerosis at the posterior joint in one ear was significantly associated with its presence at the posterior joint in the opposite ear (p = 0.01). The audiogram could not be used to reliably predict otosclerotic involvement of the posterior SVJ or the degree of footplate pathologic changes, such as ankylosis.

How do tympanic-membrane perforations affect human middle-ear sound transmission?

Although tympanic-membrane (TM) perforations are common sequelae of middle-ear disease, the hearing losses they cause have not been accurately determined, largely because additional pathological conditions occur in these ears. Our measurements of acoustic transmission before and after making controlled perforations in cadaver ears show that perforations cause frequency-dependent loss that: (1) is largest at low frequencies; (2) increases as perforation size increases; and (3) does not depend on perforation location. The dominant loss mechanism is the reduction in sound-pressure difference across the TM. Measurements of middle-ear air-space sound pressures show that transmission via direct acoustic stimulation of the oval and round windows is generally negligible. A quantitative model predicts the influence of middle-ear air-space volume on loss; with larger volumes, loss is smaller.

Acoustic responses of the human middle ear.

Measurements on human cadaver ears are reported that describe sound transmission through the middle ear. Four response variables were measured with acoustic stimulation at the tympanic membrane: stapes velocity, middle-ear cavity sound pressure, acoustic impedance at the tympanic membrane and acoustic impedance of the middle-ear cavity. Measurements of stapes velocity at different locations on the stapes suggest that stapes motion is predominantly 'piston-like', for frequencies up to at least 2000 Hz. The measurements are generally consistent with constraints of existing models. The measurements are used (1) to show how the cavity pressure and the impedance at the tympanic membrane are related, (2) to develop a measurement-based middle-ear cavity model, which shows that the middle-ear cavity has only small effects on the motion of the tympanic membrane and stapes in the normal ear, although it may play a more prominent role in pathological ears, and (3) to show that inter-ear variations in the impedance at the tympanic membrane and the stapes velocity are not well correlated.