Ekaterini Solomou

Improving brain tumor characterization on MRI by probabilistic neural networks and non-linear transformation of textural features

The aim of the present study was to design, implement and evaluate a software system for discriminating between metastatic and primary brain tumors (gliomas and meningiomas) on MRI, employing textural features from routinely taken T1 post-contrast images. The proposed classifier is a modified probabilistic neural network (PNN), incorporating a non-linear least squares features transformation (LSFT) into the PNN classifier. Thirty-six textural features were extracted from each one of 67 T1-weighted post-contrast MR images (21 metastases, 19 meningiomas and 27 gliomas). LSFT enhanced the performance of the PNN, achieving classification accuracies of 95.24% for discriminating between metastatic and primary tumors and 93.48% for distinguishing gliomas from meningiomas. To improve the generalization of the proposed classification system, the external cross-validation method was also used, resulting in 71.43% and 81.25% accuracies in distinguishing metastatic from primary tumors and gliomas from meningiomas, respectively. LSFT improved PNN performance, increased class separability and resulted in dimensionality reduction.

Asymptomatic adult cystic lymphangioma of the spleen: case report and review of the literature

In the present report we describe a case of an asymptomatic splenic cystic lymphangioma in a 43 year-old female. Only a few cases of this benign tumor have been reported in adult patients so far. Clinical examination revealed a tender mass in the upper left quadrant of the abdomen. Abdominal ultrasound and CT-scan revealed a large well-defined splenic cystic mass surrounded by multiple peripheral cysts, all divided by thin septa. MRI confirmed these findings and excluded the possibility of malignant degeneration. Histologic examination permitted the accurate diagnosis to be made. Different imaging findings of this tumor have been described but only a few reports have focused on the value of MRI imaging.

Non-linear least squares features transformation for improving the performance of probabilistic neural networks in classifying human brain tumors on MRI

The aim of the present study was to design, implement, and evaluate a software system for discriminating between metastases, meningiomas, and gliomas on MRI. The proposed classifier is a modified probabilistic neural network (PNN), incorporating a second degree least squares features transformation (LSFT) into the PNN classifier. Thirty-six textural features were extracted from each one of 75 T1-weighted post-contrast MR images (24 metastases, 21 meningiomas, and 30 gliomas). Classification performance was evaluated employing the leave-one-out method and for all possible textural feature combinations. LSFT enhanced the performance of the PNN, achieving 93.33%in discriminating between the three major types of human brain tumors, against 89.33% scored by the PNN alone. Best feature combination for achieving highest discrimination power included the mean value and entropy, which reflect specific properties of texture, i.e. signal strength and inhomogeneity. LSFT improved PNN performance, increased class separability, and resulted in dimensionality reduction.

Intracerebral haemorrhage as a rare complication of HSV-1 meningoencephalitis: case report and review of the literature.

We present a case of herpetic meningoencephalitis confirmed by PCR in a 22-y-old male, with accompanying appearance of a large intracerebral haematoma as a complication. Despite the impressive imaging findings, the final outcome of the patients progress was favourable.

A case of a giant pseudoangiomatous stromal hyperplasia of the breast: magnetic resonance imaging findings

Pseudoangiomatous stromal hyperplasia (PASH) of the breast is a benign myofibroblastic process. We present the case of a 17-year-old girl who underwent diagnostic work-up due to an enlargement of her left breast. She was submitted to ultrasounds and magnetic resonance imaging (MRI) which depicted a 14 cm lesion in her left breast. The patient was later operated and histology revealed PASH. Although PASH may range from 0.6–12 cm, a few lesions over 12 cm have been described, the largest being 20 cm. Large series present mammographic and ultrasonographic features of PASH in the literature, but little has been reported on the MR characteristics of PASH up to today. Signal on the T1-weighted image (T1WI) and T2-weighted image (T2WI) may vary. Curves generated from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) studies are mainly type I or less frequently type II. There are no reports about diffusion-weighted imaging and corresponding apparent diffusion coefficient (ADC) values for PASH in the literature. ADC values in our case lie within the range of values reported for other benign breast lesions. The presence of slit-like spaces within the lesion on MR imaging along with DCE-MRI type I curve and ADC values consistent with a benign lesion may favour the diagnosis of PASH. Tissue biopsy is necessary, however for the final diagnosis. This case report will further contribute to the understanding of MR imaging features of PASH, especially in cases where mammography is not indicated.

PDA-based system with teleradiology and image analysis capabilities

The aim of the present study was to design and implement a Personal Digital Assistant (PDA)-based teleradiology system incorporating image processing and analysis facilities for use in emergency situations within a hospital environment. The system comprised a DICOM-server, connected to an MRI unit, 3 wireless access points, and 3 PDAs (HP iPaq rx3715). PDA application software was developed in MS Embedded Visual C++ 4.0. Each PDA can receive, load, process and analyze hi-quality static MR images. Image processing includes gray-scale manipulation and spatial filtering techniques while image analysis incorporates a probabilistic neural network (PNN) classifier, which was optimally designed employing a suitable combination of textural features and was evaluated using the leave-one-out method. The PNN is capable of discriminating between three major types of human brain tumors with accuracy of 86.66%. The developed application may be useful as a mobile medical teleconsultation tool.

Brucellosis presenting as piriformis myositis: a case report

IntroductionMyositis is a rare bacterial muscle infection. Involvement of the piriformis muscle has been rarely reported in the literature. In this report we describe a case of piriformis myositis due to Brucella melitensis, which to the best of our knowledge is the first such case presented in the literature.Case presentationWe report the case of a 19-year-old Caucasian man who presented to our institution with fever and right hip pain. Brucellosis was suspected, but the clinical suspicion was for spondylodiscitis. A pelvic magnetic resonance imaging scan allowed prompt diagnosis of inflammatory involvement of the right piriformis muscle. Blood culture results were positive for B. melitensis. Our patient was treated with antibiotics, and follow-up magnetic resonance imaging scans showed resolution of the inflammation.ConclusionBrucellosis can present as piriformis myositis. The clinical diagnosis of piriformis myositis is difficult, as it can mimic other common entities such as referred back pain from spondylodiscitis. Magnetic resonance imaging is the method of choice for establishing the diagnosis in the early stages of the disease, as late diagnosis can lead to abscess formation and the need for drainage.

Designing a pattern recognition system on GPU for discriminating between patients with micro-ischaemic and multiple sclerosis lesions, using MRI images

The aim of this study was to employ state-of-art graphics processing unit (GPU) technology and CUDA parallel programming to design and implement a stand-alone pattern recognition (PR) system to discriminate between patients with micro-ischaemic (mIS) and multiple sclerosis (MS) lesions. The dataset comprised MRI image series of 32 patients with mIS and 19 with MS lesions. The probabilistic neural network classifier and 40 textural features, calculated from lesions in the magnetic resonance imaging (MRI) images, were used to design the PR system. The highest classification accuracy was 90.2%, employing six textural features. It took about 135 minutes to design the PR system on a desktop CPU (Intel Core 2 Quad Q9550), using sequential programming, against 250 seconds on the Nvidia 8800GT GPU card, using parallel programming. The proposed PR system may be redesigned on site, when new verified data are incorporated in its depository, and it may serve as a second opinion tool in a clinical environment.

Magnetic Resonance Imaging Of Metastatic Bone Disease

Early diagnosis of bone metastases is crucial in order to determine the prognosis and optimize therapy. Traditional methods, such as plain radiography or bone scintigraphy, lack either sensitivity or specificity. Computed tomography (CT) is quite sensitive, however, its ability to detect early deposits is limited. FDG {PET – CT} scan detects metastatic bone disease before occurrence of osteoblastic activity. Magnetic resonance imaging (MRI) has been shown to be the most sensitive imaging technique, with a sensitivity of up to 100% reported and its specificity was reported to reach 97%.

Combination therapy with mitoxantrone fluorouracil, leucovorin, and granulocyte colony-stimulating factor in patients with advanced breast cancer: a phase II study

Abstract The purpose of this study was to assess the efficacy of combination therapy with mitoxantrone, fluorouracil, and high-dose leucovorin (MFL) and granulocyte colony-stimulating factor (G-CSF) in patients with metastatic breast cancer. From May 1994 to June 1997, 46 patients with stage IV breast cancer were treated with MFl chemotherapy plus G-CSF, as needed. World Health Organization criteria were used to define objective responses and toxicity. Thirty-eight (82.6%) of 46 patients were assessable for response to therapy and toxicity. Median follow-up was 15.2 months (range, 4 to 34 months). Twelve (31.6%) of these patients demonstrated a complete response and 16 (42.1%) a partial response, giving an overall response rate of 73.7%. A small number of patients had grade 3 and 4 neutropenia and/or anemia. The results of this study using combination therapy with MFL plus G-CSF are encouraging. The overall response rate of 73.7% in the 38 assessable patients is noteworthy, because all had stage IV breast cancer and 24 (63.2%) of 38 patients had visceral-dominant disease. We conclude that the combination of MFL and G-CSF is a well-tolerated, active regimen for use as first-line therapy in patients with metastatic breast cancer.