Neural models for auditory localization based on spectral cues.

In this paper we analyze several auditory localization neural models that are based on head related transfer functions (HRTFs). HRTFs represent the combined directional-spectral response of the pinnae head and torso. The role of HRTFs in such modeling has hitherto been underestimated despite substantial experimental evidence to its relevance in spatial hearing, especially in determining direction of high-frequency sound sources. In the first section we suggest a neural model that links the physiology of binaural processing to a neural network that extracts spectral ratios. These ratios correspond to HRTFs ratios and can provide auditory directional cues. Next, we compare several methods of matching HRTFs ratios using discriminative matching measure (DMM). We consider several solutions to the matching problem from a neural signal processing viewpoint. We compare correlation based approaches with DMM optimization approach and with a non-linear approach based on neural back-propagation algorithm. All three models can be implemented by neural networks. Finally, we include experimental results of simulations that are conducted using these methods. Experiments show that the back-propagation based neural network yields the best results in terms of DMM both for narrow-band and broad band excitation. The back-propagation neural network is also superior in matching noisy HRTF ratio vectors.

Shape from recognition: a novel approach for 3-D face shape recovery.

In this paper, we develop a novel framework for robust recovery of three-dimensional (3-D) surfaces of faces from single images. The underlying principle is shape from recognition, i.e., the idea that pre-recognizing face parts can constrain the space of possible solutions to the image irradiance equation, thus allowing robust recovery of the 3-D structure of a specific part. Parts of faces like nose, lips and eyes are recognized and localized using robust expansion matching filter templates under varying pose and illumination. Specialized backpropagation based neural networks are then employed to recover the 3-D shape of particular face parts. Representation using principal components allows to efficiently encode classes of objects such as nose, lips, etc. The specialized networks are designed and trained to map the principal component coefficients of the part images to another set of principal component coefficients that represent the corresponding 3-D surface shapes. To achieve robustness to viewing conditions, the network is trained with a wide range of illumination and viewing directions. A method for merging recovered 3-D surface regions by minimizing the sum squared error in overlapping areas is also derived. Quantitative analysis of the reconstruction of the surface parts in varying illumination and pose show relatively small errors, indicating that the method is robust and accurate. Several examples showing recovery of the complete face also illustrate the efficacy of the approach.

Detection and segmentation of generic shapes based on affine modeling of energy in eigenspace.

This paper presents a novel approach for detection and segmentation of man made generic shapes in cluttered images. The set of shapes to be detected are members of affine transformed versions of basic geometric shapes such as rectangles, circles etc. The shape set is represented by its vectorial edge map transformed over a wide range of affine parameters. We use vectorial boundary instead of regular boundary to improve the robustness to noise, background clutter and partial occlusion. Our approach consists of a detection stage and a verification stage. In the detection stage, we first derive the energy from the principal eigenvectors of the set. Next, an a posteriori probability map of energy distribution is computed from the projection of the edge map representation in a vectorial eigen-space. Local peaks of the posterior probability map are located and indicate candidate detections. We use energy/probability based detection since we find that the underlying distribution is not Gaussian and resembles a hypertoroid. In the verification stage, each candidate is verified using a fast search algorithm based on a novel representation in angle space and the corresponding pose information of the detected shape is obtained. The angular representation used in the verification stage yields better results than a Euclidean distance representation. Experiments are performed in various interfering distortions, and robust detection and segmentation are achieved.

Estimation of 3-D motion using eigen-normalization and expansion matching.

This correspondence describes a novel approach to three-dimensional (3-D) motion estimation of planar objects based on eigen-normalization, expansion matching (EXM), and a scaled orthographic projection model. Our approach leads to a comprehensive temporal description of all the degrees of freedom in 3-D (three rotations and three translations). Experiments with video streams show robust estimation of the real 3-D rotations and translations of the objects in motion.

Generalized feature extraction using expansion matching.

A novel generalized feature extraction method based on the expansion matching (EXM) method and on the Karhunen-Loeve transform (KLT) is presented. The method provides an efficient way to locate complex features of interest like corners and junctions with reduced number of filtering operations. The EXM method is used to design optimal detectors for a set of model elementary features. The KL representation of these model EXM detectors is used to filter the image and detect candidate interest points from the energy peaks of the eigen coefficients. The KL coefficients at these candidate points are then used to efficiently reconstruct the response and differentiate real junctions and corners from arbitrary features in the image. The method is robust to additive noise and is able to successfully extract, classify, and find the myriad compositions of corner and junction features formed by combinations of two or more edges or lines. This method differs from previous works in several aspects. First, it treats the features not as distinct entities, but as combinations of elementary features. Second, it employs an optimal set of elementary feature detectors based on the EM approach. Third, the method incorporates a significant reduction in computational complexity by representing a large set of EXM filters by a relatively small number of eigen filters derived by the KL transform of the basic EXM filter set. This is a novel application of the KL transform, which is usually employed to represent signals and not impulse responses as in our present work.

Finding shape axes using magnetic fields.

This paper presents a novel method, based on magnetic field principles, for obtaining the axes of shapes. The method is based on directional information of the shape's boundary. By simulating a parallel algorithm, we are able to generate the inner as well as the outer axes (axes of concavities) of the shape. The preprocessing phase for this algorithm involves obtaining the shape's gradient. Each point of the gradient is substituted by a minute magnetic dipole. The cumulative magnetic field due to these dipoles is accumulated at all points in the image in a one-pass algorithm. The magnitude of the final magnetic vector field has valleys that are created from mutual and directionally balanced cancellations of opposing boundary segments. These valleys signify the axes of the shape. The axes are obtained by performing a valley search. The magnetic field modeling (MFM) method has an advantage over previous approaches since it utilizes not only the location information of the boundary, but also its directional information. As demonstrated, experimental results of the MFM method are much improved, compared to other skeletonization algorithms which tend to generate spurious and noisy axes.

An auditory localization model based on high-frequency spectral cues.

We present in this paper a connectionist model that extracts interaural intensity differences (IID) from head-related transfer functions (HRTF) in the form of spectral cues to localize broadband high-frequency auditory stimuli, in both azimuth and elevation. A novel discriminative matching measure (DMM) is defined and optimized to characterize matching this IID spectrum. The optimal DMM approach and a novel back-propagation-based fuzzy model of localization are shown to be capable of localizing sources in azimuth, using only spectral IID cues. The fuzzy neural network model is extended to include localization in elevation. The use of training data with additive noise provides robustness to input errors. Outputs are modeled as two-dimensional Gaussians that act as membership functions for the fuzzy sets of sound locations. Error back-propagation is used to train the network to correlate input patterns and the desired output patterns. The fuzzy outputs are used to estimate the location of the source by detecting Gaussians using the max-energy paradigm. The proposed model shows that HRTF-based spectral IID patterns can provide sufficient information for extracting localization cues using a connectionist paradigm. Successful recognition in the presence of additive noise in the inputs indicates that the computational framework of this model is robust to errors made in estimating the IID patterns. The localization errors for such noisy patterns at various elevations and azimuths are compared and found to be within limits of localization blurs observed in humans.

Optimal head related transfer functions for hearing and monaural localization in elevation: a signal processing design perspective.

Localization of sound sources by human listeners has been widely studied and theories and various models of the localization and hearing mechanism have been constructed. In the classical "duplex" theory, sound localization in azimuth is explained by interaural time or equivalently, phase differences at low frequencies, and by interaural amplitude differences at higher frequencies. Head related transfer functions (HRTF's) present a linear system approach to modeling localization by representing the direction-dependent transformation the sound undergoes at each ear. Localization in elevation is explained by directional differences in the HRTF's, which also explains monaural localization. We conjecture that the HRTF's evolved during the course of nature (due to the evolution of the shape and structure of the ear etc.) are optimal with respect to several physically realizable criteria. In this paper, we investigate the problem of defining the design constraints which when optimized yield a set of HRTF's for hearing and monaural vertical localization in an attempt to better understand, and if possible, duplicate nature's design. We pursue an engineer's design perspective and formulate a constrained optimization problem, where the desired set of HRTF's is optimized according to a cost function based on several criteria for localization, hearing and smoothness, and also by imposing physically realizable constraints on the HRTF's such as nonnegativity, energy etc. The value of the cost function for a candidate set of HRTF's is an indication of the similarity of that set of HRTF's with respect to the ideal solution (measured HRTF data). The final optimization results we present are similar to the actual HRTF's measured in human subjects, and the associated cost function values are found to be almost equal. This points to the fact that the optimization criteria defined are quite relevant. The significant outcome of this research is the identification of a relevant set of mathematical criteria that could be optimized in the human auditory system to facilitate good hearing and localization. These criteria along with the associated constraints represent the desirable characteristics of the HRTF's in an HRTF-based localization system, and could lead to a better understanding and modeling of the auditory system.

Ataxia telangiectasia and lymphoma: an indication for individualized chemotherapy dosing--report of treatment in a highly inbred Arab family.

Ten of 18 children in a highly inbred Arab kindred suffered from either ataxia telangiectasia (AT) or a variant syndrome consisting of ataxia, microcephaly, and congenital cataract (AMC). Four of the nine afflicted children were treated in our unit when they developed lymphomas (both Hodgkin's and non-Hodgkin's including Burkitt's). They were given chemotherapy (either standard COMP or low-dose ABV/CVPP). The children with non-Hodgkin's lymphomas died of sepsis after receiving full-dose COMP. Low-dose ABV/CVPP brought about a 20-month remission in one child with nodular sclerosing Hodgkin's lymphoma and both AT and AMC, but she developed a preleukemic syndrome and her parents refused further treatment; she too died. A fourth child, also with nodular sclerosing Hodgkin's lymphoma, is currently in complete remission after ABV/CVPP. Treatment of lymphomas in patients with AT is extraordinarily difficult and has potential side effects so grave as to necessitate careful monitoring and individualized protocols.

Hodgkin's disease in childhood: treatment modalities, outcome and epidemiological aspects. The Northern Israel Cancer Center experience.

PURPOSE: A retrospective analysis of Hodgkin's disease (HD) in children treated at the Northern Israel Cancer Center between 1971 and 1990 was conducted. PATIENTS AND METHODS: Records of 102 patients < 18 years of age at diagnosis were reviewed. Patient characteristics were similar to those previously reported. There were 54 boys and 48 girls (1.1:1 boy:girl ratio), with more boys < 10 years of age. Forty-four patients were of Arab ancestry and 58 were Jewish; incidence rates were similar in both groups. The most common histological types were nodular sclerosing and mixed cellularity, the latter being more commonly diagnosed in the younger age group. RESULTS: The outcome of various treatment modalities in childhood HD were evaluated. Sixty-five patients (64%) had stage I or II and 37 (34%) had stage III or IV at diagnosis. Patients with stage I-II received radiotherapy alone (20 patients), chemotherapy alone (10 patients), or a combined approach of chemotherapy plus radiotherapy (35 patients). Survival rates and median disease-free intervals were statistically similar in all three modalities. However, relapse rates were higher among patients receiving radiotherapy alone or chemotherapy alone (35% and 38%, respectively) compared with patients receiving chemotherapy plus radiotherapy (14%). CONCLUSIONS: We conclude that a combined approach of chemotherapy plus radiotherapy is advantageous over radiotherapy alone or chemotherapy alone.

[Hodgkin's disease in childhood--northern Israel cancer center experience, 1971-1990].

Records of children treated for Hodgkin's disease between 1971-1990 were analyzed retrospectively, and 102 cases of children younger than 18 years when diagnosed were reviewed. There were 54 boys and 48 girls, with male predominance in those younger than 10 years. 44 patients were of Arab ancestry and 58 were Jewish; the incidence was similar in both groups. Most common were the nodular sclerosing and mixed cellularity types, the latter more common in the younger age group. The outcome of various treatments was evaluated. At diagnosis, 64% were in stages I or II and 34% in stages III or IV. 20 of those in stages I-II received radiotherapy (RT) alone, 10 chemotherapy (CT) alone, and 35 combined CT and RT. Survival rates and median disease-free intervals were statistically similar with all 3 modalities. However, relapse rates were higher with either RT or CT alone (35% and 38%, respectively) than with combined therapy (14%). We conclude that combined CT and RT is superior to RT or CT alone.

[Primary malignant lymphoma of the testis].

Among 715 patients with non-Hodgkin's lymphoma referred to this oncology center between 1973-1989, there were 7 (0.9%) with primary lymphoma of the testis. The mean age was 55 (range 23-78). The presenting symptom in all the 7 was testicular mass or swelling. 6 of them underwent inguinal orchiectomy and 1 had a testicular biopsy. The pathological subtype of the lymphoma in all 7 was the diffuse large cell type. The disease stage was IE in 3, IIE in 3 and IVE in 1. Only 1 had bilateral testicular involvement. 6 were treated primarily with various combinations of chemotherapy and 1 by radiotherapy. 3 developed recurrent disease (2 treated primarily with chemotherapy and 1 by radiotherapy) and died 14, 19 and 27 months, respectively, after diagnosis. 3 are alive with no evidence of recurrent lymphoma. Follow-up in an additional patient has been too short for evaluation. Primary testicular lymphoma is a relatively rare form of extranodal lymphoma. It has a tendency to early systemic progression and is potentially curable by aggressive combination chemotherapy.

Plastipore in reconstruction of the laryngo-tracheal complex.

The feasibility of using plastic material in the reconstruction of the damaged laryngotracheal complex was studied. In 6 dogs, large segments of larynx and trachea were removed and replaced by Plastipore sheets covered on the inside by free skin or mucosa grafts. It was demonstrated that the skin and mucosa covering the Plastipore--and resurfacing the airway's lumen--remained viable. However, there appeared to be inadequate incorporation of the Plastipore into the tissue, probably due to the small pore size. Larger size pores may increase the likelihood that the Plastipore will become firmly embedded in the tissue environment.

Spinal intradural metastases of extraneural origin.

Three cases of intradural extramedullary (spinal subdural) metastases, originating from outside the central nervous system, are presented. Two were located at the cervico-dorsal junction, and one was mid-dorsal. A breast ductus carcinoma, a cervical neuroblastoma, and an apoduma of unknown origin, were the primary neoplasms. All presented clinically with a short history typical of cancerous spinal cord compression. Plain X-rays of the spine did not show areas of destruction. Myelography in two cases clearly suggested that intradural location of the tumour. The relative frequency of these tumours and their pathogenesis are briefly reviewed. It is stressed that primary tumours are mainly in the breast or lung. Their metastases are mainly found in the cervico-dorsal region. It is assumed that they really are metastases of the dura mater itself, growing inward. The importance of the lymphatic and venous pathways in their spread into the dura mater is emphasized.